From 8e761289a2446367267c6c0d9a26e734f0f78306 Mon Sep 17 00:00:00 2001 From: Karen Arutyunov Date: Wed, 16 Dec 2020 20:29:05 +0300 Subject: Get rid of legacy build systems and rename cutl/ to libcutl/ --- cutl/details/boost/regex/v4/basic_regex.hpp | 782 ------ .../details/boost/regex/v4/basic_regex_creator.hpp | 1571 ----------- cutl/details/boost/regex/v4/basic_regex_parser.hpp | 2874 -------------------- cutl/details/boost/regex/v4/c_regex_traits.hpp | 211 -- cutl/details/boost/regex/v4/char_regex_traits.hpp | 81 - cutl/details/boost/regex/v4/cpp_regex_traits.hpp | 1099 -------- cutl/details/boost/regex/v4/cregex.hpp | 330 --- cutl/details/boost/regex/v4/error_type.hpp | 59 - cutl/details/boost/regex/v4/fileiter.hpp | 455 ---- cutl/details/boost/regex/v4/instances.hpp | 222 -- cutl/details/boost/regex/v4/iterator_category.hpp | 91 - cutl/details/boost/regex/v4/iterator_traits.hpp | 135 - cutl/details/boost/regex/v4/match_flags.hpp | 138 - cutl/details/boost/regex/v4/match_results.hpp | 702 ----- cutl/details/boost/regex/v4/mem_block_cache.hpp | 99 - cutl/details/boost/regex/v4/perl_matcher.hpp | 587 ---- .../details/boost/regex/v4/perl_matcher_common.hpp | 996 ------- .../boost/regex/v4/perl_matcher_non_recursive.hpp | 1642 ----------- .../boost/regex/v4/perl_matcher_recursive.hpp | 991 ------- cutl/details/boost/regex/v4/primary_transform.hpp | 146 - cutl/details/boost/regex/v4/protected_call.hpp | 81 - cutl/details/boost/regex/v4/regbase.hpp | 180 -- cutl/details/boost/regex/v4/regex.hpp | 202 -- cutl/details/boost/regex/v4/regex_format.hpp | 1156 -------- cutl/details/boost/regex/v4/regex_fwd.hpp | 73 - cutl/details/boost/regex/v4/regex_grep.hpp | 155 -- cutl/details/boost/regex/v4/regex_iterator.hpp | 201 -- cutl/details/boost/regex/v4/regex_match.hpp | 382 --- cutl/details/boost/regex/v4/regex_merge.hpp | 93 - cutl/details/boost/regex/v4/regex_raw_buffer.hpp | 210 -- cutl/details/boost/regex/v4/regex_replace.hpp | 99 - cutl/details/boost/regex/v4/regex_search.hpp | 217 -- cutl/details/boost/regex/v4/regex_split.hpp | 172 -- .../boost/regex/v4/regex_token_iterator.hpp | 342 --- cutl/details/boost/regex/v4/regex_traits.hpp | 189 -- .../boost/regex/v4/regex_traits_defaults.hpp | 371 --- cutl/details/boost/regex/v4/regex_workaround.hpp | 232 -- cutl/details/boost/regex/v4/states.hpp | 301 -- cutl/details/boost/regex/v4/sub_match.hpp | 512 ---- cutl/details/boost/regex/v4/syntax_type.hpp | 105 - cutl/details/boost/regex/v4/u32regex_iterator.hpp | 193 -- .../boost/regex/v4/u32regex_token_iterator.hpp | 377 --- cutl/details/boost/regex/v4/w32_regex_traits.hpp | 741 ----- 43 files changed, 19795 deletions(-) delete mode 100644 cutl/details/boost/regex/v4/basic_regex.hpp delete mode 100644 cutl/details/boost/regex/v4/basic_regex_creator.hpp delete mode 100644 cutl/details/boost/regex/v4/basic_regex_parser.hpp delete mode 100644 cutl/details/boost/regex/v4/c_regex_traits.hpp delete mode 100644 cutl/details/boost/regex/v4/char_regex_traits.hpp delete mode 100644 cutl/details/boost/regex/v4/cpp_regex_traits.hpp delete mode 100644 cutl/details/boost/regex/v4/cregex.hpp delete mode 100644 cutl/details/boost/regex/v4/error_type.hpp delete mode 100644 cutl/details/boost/regex/v4/fileiter.hpp delete mode 100644 cutl/details/boost/regex/v4/instances.hpp delete mode 100644 cutl/details/boost/regex/v4/iterator_category.hpp delete mode 100644 cutl/details/boost/regex/v4/iterator_traits.hpp delete mode 100644 cutl/details/boost/regex/v4/match_flags.hpp delete mode 100644 cutl/details/boost/regex/v4/match_results.hpp delete mode 100644 cutl/details/boost/regex/v4/mem_block_cache.hpp delete mode 100644 cutl/details/boost/regex/v4/perl_matcher.hpp delete mode 100644 cutl/details/boost/regex/v4/perl_matcher_common.hpp delete mode 100644 cutl/details/boost/regex/v4/perl_matcher_non_recursive.hpp delete mode 100644 cutl/details/boost/regex/v4/perl_matcher_recursive.hpp delete mode 100644 cutl/details/boost/regex/v4/primary_transform.hpp delete mode 100644 cutl/details/boost/regex/v4/protected_call.hpp delete mode 100644 cutl/details/boost/regex/v4/regbase.hpp delete mode 100644 cutl/details/boost/regex/v4/regex.hpp delete mode 100644 cutl/details/boost/regex/v4/regex_format.hpp delete mode 100644 cutl/details/boost/regex/v4/regex_fwd.hpp delete mode 100644 cutl/details/boost/regex/v4/regex_grep.hpp delete mode 100644 cutl/details/boost/regex/v4/regex_iterator.hpp delete mode 100644 cutl/details/boost/regex/v4/regex_match.hpp delete mode 100644 cutl/details/boost/regex/v4/regex_merge.hpp delete mode 100644 cutl/details/boost/regex/v4/regex_raw_buffer.hpp delete mode 100644 cutl/details/boost/regex/v4/regex_replace.hpp delete mode 100644 cutl/details/boost/regex/v4/regex_search.hpp delete mode 100644 cutl/details/boost/regex/v4/regex_split.hpp delete mode 100644 cutl/details/boost/regex/v4/regex_token_iterator.hpp delete mode 100644 cutl/details/boost/regex/v4/regex_traits.hpp delete mode 100644 cutl/details/boost/regex/v4/regex_traits_defaults.hpp delete mode 100644 cutl/details/boost/regex/v4/regex_workaround.hpp delete mode 100644 cutl/details/boost/regex/v4/states.hpp delete mode 100644 cutl/details/boost/regex/v4/sub_match.hpp delete mode 100644 cutl/details/boost/regex/v4/syntax_type.hpp delete mode 100644 cutl/details/boost/regex/v4/u32regex_iterator.hpp delete mode 100644 cutl/details/boost/regex/v4/u32regex_token_iterator.hpp delete mode 100644 cutl/details/boost/regex/v4/w32_regex_traits.hpp (limited to 'cutl/details/boost/regex/v4') diff --git a/cutl/details/boost/regex/v4/basic_regex.hpp b/cutl/details/boost/regex/v4/basic_regex.hpp deleted file mode 100644 index e238d25..0000000 --- a/cutl/details/boost/regex/v4/basic_regex.hpp +++ /dev/null @@ -1,782 +0,0 @@ -/* - * - * Copyright (c) 1998-2004 John Maddock - * Copyright 2011 Garmin Ltd. or its subsidiaries - * - * Distributed under the Boost Software License, Version 1.0. - * (See accompanying file LICENSE_1_0.txt or copy at - * http://www.boost.org/LICENSE_1_0.txt) - * - */ - - /* - * LOCATION: see http://www.boost.org/ for most recent version. - * FILE basic_regex.cpp - * VERSION see - * DESCRIPTION: Declares template class basic_regex. - */ - -#ifndef BOOST_REGEX_V4_BASIC_REGEX_HPP -#define BOOST_REGEX_V4_BASIC_REGEX_HPP - -#include -#include - -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable: 4103) -#endif -#ifdef BOOST_HAS_ABI_HEADERS -# include BOOST_ABI_PREFIX -#endif -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif - -namespace cutl_details_boost{ -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable : 4251 4231 4800) -#if BOOST_MSVC < 1600 -#pragma warning(disable : 4660) -#endif -#endif - -namespace re_detail{ - -// -// forward declaration, we will need this one later: -// -template -class basic_regex_parser; - -template -void bubble_down_one(I first, I last) -{ - if(first != last) - { - I next = last - 1; - while((next != first) && (*next < *(next-1))) - { - (next-1)->swap(*next); - --next; - } - } -} - -template -inline int hash_value_from_capture_name(Iterator i, Iterator j) -{ - std::size_t r = cutl_details_boost::hash_range(i, j); - r %= ((std::numeric_limits::max)() - 10001); - r += 10000; - return static_cast(r); -} - -class named_subexpressions -{ -public: - struct name - { - template - name(const charT* i, const charT* j, int idx) - : index(idx) - { - hash = hash_value_from_capture_name(i, j); - } - name(int h, int idx) - : index(idx), hash(h) - { - } - int index; - int hash; - bool operator < (const name& other)const - { - return hash < other.hash; - } - bool operator == (const name& other)const - { - return hash == other.hash; - } - void swap(name& other) - { - std::swap(index, other.index); - std::swap(hash, other.hash); - } - }; - - typedef std::vector::const_iterator const_iterator; - typedef std::pair range_type; - - named_subexpressions(){} - - template - void set_name(const charT* i, const charT* j, int index) - { - m_sub_names.push_back(name(i, j, index)); - bubble_down_one(m_sub_names.begin(), m_sub_names.end()); - } - template - int get_id(const charT* i, const charT* j)const - { - name t(i, j, 0); - typename std::vector::const_iterator pos = std::lower_bound(m_sub_names.begin(), m_sub_names.end(), t); - if((pos != m_sub_names.end()) && (*pos == t)) - { - return pos->index; - } - return -1; - } - template - range_type equal_range(const charT* i, const charT* j)const - { - name t(i, j, 0); - return std::equal_range(m_sub_names.begin(), m_sub_names.end(), t); - } - int get_id(int h)const - { - name t(h, 0); - std::vector::const_iterator pos = std::lower_bound(m_sub_names.begin(), m_sub_names.end(), t); - if((pos != m_sub_names.end()) && (*pos == t)) - { - return pos->index; - } - return -1; - } - range_type equal_range(int h)const - { - name t(h, 0); - return std::equal_range(m_sub_names.begin(), m_sub_names.end(), t); - } -private: - std::vector m_sub_names; -}; - -// -// class regex_data: -// represents the data we wish to expose to the matching algorithms. -// -template -struct regex_data : public named_subexpressions -{ - typedef regex_constants::syntax_option_type flag_type; - typedef std::size_t size_type; - - regex_data(const ::cutl_details_boost::shared_ptr< - ::cutl_details_boost::regex_traits_wrapper >& t) - : m_ptraits(t), m_expression(0), m_expression_len(0) {} - regex_data() - : m_ptraits(new ::cutl_details_boost::regex_traits_wrapper()), m_expression(0), m_expression_len(0) {} - - ::cutl_details_boost::shared_ptr< - ::cutl_details_boost::regex_traits_wrapper - > m_ptraits; // traits class instance - flag_type m_flags; // flags with which we were compiled - int m_status; // error code (0 implies OK). - const charT* m_expression; // the original expression - std::ptrdiff_t m_expression_len; // the length of the original expression - size_type m_mark_count; // the number of marked sub-expressions - re_detail::re_syntax_base* m_first_state; // the first state of the machine - unsigned m_restart_type; // search optimisation type - unsigned char m_startmap[1 << CHAR_BIT]; // which characters can start a match - unsigned int m_can_be_null; // whether we can match a null string - re_detail::raw_storage m_data; // the buffer in which our states are constructed - typename traits::char_class_type m_word_mask; // mask used to determine if a character is a word character - std::vector< - std::pair< - std::size_t, std::size_t> > m_subs; // Position of sub-expressions within the *string*. - bool m_has_recursions; // whether we have recursive expressions; -}; -// -// class basic_regex_implementation -// pimpl implementation class for basic_regex. -// -template -class basic_regex_implementation - : public regex_data -{ -public: - typedef regex_constants::syntax_option_type flag_type; - typedef std::ptrdiff_t difference_type; - typedef std::size_t size_type; - typedef typename traits::locale_type locale_type; - typedef const charT* const_iterator; - - basic_regex_implementation(){} - basic_regex_implementation(const ::cutl_details_boost::shared_ptr< - ::cutl_details_boost::regex_traits_wrapper >& t) - : regex_data(t) {} - void assign(const charT* arg_first, - const charT* arg_last, - flag_type f) - { - regex_data* pdat = this; - basic_regex_parser parser(pdat); - parser.parse(arg_first, arg_last, f); - } - - locale_type BOOST_REGEX_CALL imbue(locale_type l) - { - return this->m_ptraits->imbue(l); - } - locale_type BOOST_REGEX_CALL getloc()const - { - return this->m_ptraits->getloc(); - } - std::basic_string BOOST_REGEX_CALL str()const - { - std::basic_string result; - if(this->m_status == 0) - result = std::basic_string(this->m_expression, this->m_expression_len); - return result; - } - const_iterator BOOST_REGEX_CALL expression()const - { - return this->m_expression; - } - std::pair BOOST_REGEX_CALL subexpression(std::size_t n)const - { - if(n == 0) - cutl_details_boost::throw_exception(std::out_of_range("0 is not a valid subexpression index.")); - const std::pair& pi = this->m_subs.at(n - 1); - std::pair p(expression() + pi.first, expression() + pi.second); - return p; - } - // - // begin, end: - const_iterator BOOST_REGEX_CALL begin()const - { - return (this->m_status ? 0 : this->m_expression); - } - const_iterator BOOST_REGEX_CALL end()const - { - return (this->m_status ? 0 : this->m_expression + this->m_expression_len); - } - flag_type BOOST_REGEX_CALL flags()const - { - return this->m_flags; - } - size_type BOOST_REGEX_CALL size()const - { - return this->m_expression_len; - } - int BOOST_REGEX_CALL status()const - { - return this->m_status; - } - size_type BOOST_REGEX_CALL mark_count()const - { - return this->m_mark_count; - } - const re_detail::re_syntax_base* get_first_state()const - { - return this->m_first_state; - } - unsigned get_restart_type()const - { - return this->m_restart_type; - } - const unsigned char* get_map()const - { - return this->m_startmap; - } - const ::cutl_details_boost::regex_traits_wrapper& get_traits()const - { - return *(this->m_ptraits); - } - bool can_be_null()const - { - return this->m_can_be_null; - } - const regex_data& get_data()const - { - basic_regex_implementation const* p = this; - return *static_cast*>(p); - } -}; - -} // namespace re_detail -// -// class basic_regex: -// represents the compiled -// regular expression: -// - -#ifdef BOOST_REGEX_NO_FWD -template > -#else -template -#endif -class basic_regex : public regbase -{ -public: - // typedefs: - typedef std::size_t traits_size_type; - typedef typename traits::string_type traits_string_type; - typedef charT char_type; - typedef traits traits_type; - - typedef charT value_type; - typedef charT& reference; - typedef const charT& const_reference; - typedef const charT* const_iterator; - typedef const_iterator iterator; - typedef std::ptrdiff_t difference_type; - typedef std::size_t size_type; - typedef regex_constants::syntax_option_type flag_type; - // locale_type - // placeholder for actual locale type used by the - // traits class to localise *this. - typedef typename traits::locale_type locale_type; - -public: - explicit basic_regex(){} - explicit basic_regex(const charT* p, flag_type f = regex_constants::normal) - { - assign(p, f); - } - basic_regex(const charT* p1, const charT* p2, flag_type f = regex_constants::normal) - { - assign(p1, p2, f); - } - basic_regex(const charT* p, size_type len, flag_type f) - { - assign(p, len, f); - } - basic_regex(const basic_regex& that) - : m_pimpl(that.m_pimpl) {} - ~basic_regex(){} - basic_regex& BOOST_REGEX_CALL operator=(const basic_regex& that) - { - return assign(that); - } - basic_regex& BOOST_REGEX_CALL operator=(const charT* ptr) - { - return assign(ptr); - } - - // - // assign: - basic_regex& assign(const basic_regex& that) - { - m_pimpl = that.m_pimpl; - return *this; - } - basic_regex& assign(const charT* p, flag_type f = regex_constants::normal) - { - return assign(p, p + traits::length(p), f); - } - basic_regex& assign(const charT* p, size_type len, flag_type f) - { - return assign(p, p + len, f); - } -private: - basic_regex& do_assign(const charT* p1, - const charT* p2, - flag_type f); -public: - basic_regex& assign(const charT* p1, - const charT* p2, - flag_type f = regex_constants::normal) - { - return do_assign(p1, p2, f); - } -#if !defined(BOOST_NO_MEMBER_TEMPLATES) - - template - unsigned int BOOST_REGEX_CALL set_expression(const std::basic_string& p, flag_type f = regex_constants::normal) - { - return set_expression(p.data(), p.data() + p.size(), f); - } - - template - explicit basic_regex(const std::basic_string& p, flag_type f = regex_constants::normal) - { - assign(p, f); - } - - template - basic_regex(InputIterator arg_first, InputIterator arg_last, flag_type f = regex_constants::normal) - { - typedef typename traits::string_type seq_type; - seq_type a(arg_first, arg_last); - if(a.size()) - assign(static_cast(&*a.begin()), static_cast(&*a.begin() + a.size()), f); - else - assign(static_cast(0), static_cast(0), f); - } - - template - basic_regex& BOOST_REGEX_CALL operator=(const std::basic_string& p) - { - return assign(p.data(), p.data() + p.size(), regex_constants::normal); - } - - template - basic_regex& BOOST_REGEX_CALL assign( - const std::basic_string& s, - flag_type f = regex_constants::normal) - { - return assign(s.data(), s.data() + s.size(), f); - } - - template - basic_regex& BOOST_REGEX_CALL assign(InputIterator arg_first, - InputIterator arg_last, - flag_type f = regex_constants::normal) - { - typedef typename traits::string_type seq_type; - seq_type a(arg_first, arg_last); - if(a.size()) - { - const charT* p1 = &*a.begin(); - const charT* p2 = &*a.begin() + a.size(); - return assign(p1, p2, f); - } - return assign(static_cast(0), static_cast(0), f); - } -#else - unsigned int BOOST_REGEX_CALL set_expression(const std::basic_string& p, flag_type f = regex_constants::normal) - { - return set_expression(p.data(), p.data() + p.size(), f); - } - - basic_regex(const std::basic_string& p, flag_type f = regex_constants::normal) - { - assign(p, f); - } - - basic_regex& BOOST_REGEX_CALL operator=(const std::basic_string& p) - { - return assign(p.data(), p.data() + p.size(), regex_constants::normal); - } - - basic_regex& BOOST_REGEX_CALL assign( - const std::basic_string& s, - flag_type f = regex_constants::normal) - { - return assign(s.data(), s.data() + s.size(), f); - } - -#endif - - // - // locale: - locale_type BOOST_REGEX_CALL imbue(locale_type l); - locale_type BOOST_REGEX_CALL getloc()const - { - return m_pimpl.get() ? m_pimpl->getloc() : locale_type(); - } - // - // getflags: - // retained for backwards compatibility only, "flags" - // is now the preferred name: - flag_type BOOST_REGEX_CALL getflags()const - { - return flags(); - } - flag_type BOOST_REGEX_CALL flags()const - { - return m_pimpl.get() ? m_pimpl->flags() : 0; - } - // - // str: - std::basic_string BOOST_REGEX_CALL str()const - { - return m_pimpl.get() ? m_pimpl->str() : std::basic_string(); - } - // - // begin, end, subexpression: - std::pair BOOST_REGEX_CALL subexpression(std::size_t n)const - { - if(!m_pimpl.get()) - cutl_details_boost::throw_exception(std::logic_error("Can't access subexpressions in an invalid regex.")); - return m_pimpl->subexpression(n); - } - const_iterator BOOST_REGEX_CALL begin()const - { - return (m_pimpl.get() ? m_pimpl->begin() : 0); - } - const_iterator BOOST_REGEX_CALL end()const - { - return (m_pimpl.get() ? m_pimpl->end() : 0); - } - // - // swap: - void BOOST_REGEX_CALL swap(basic_regex& that)throw() - { - m_pimpl.swap(that.m_pimpl); - } - // - // size: - size_type BOOST_REGEX_CALL size()const - { - return (m_pimpl.get() ? m_pimpl->size() : 0); - } - // - // max_size: - size_type BOOST_REGEX_CALL max_size()const - { - return UINT_MAX; - } - // - // empty: - bool BOOST_REGEX_CALL empty()const - { - return (m_pimpl.get() ? 0 != m_pimpl->status() : true); - } - - size_type BOOST_REGEX_CALL mark_count()const - { - return (m_pimpl.get() ? m_pimpl->mark_count() : 0); - } - - int status()const - { - return (m_pimpl.get() ? m_pimpl->status() : regex_constants::error_empty); - } - - int BOOST_REGEX_CALL compare(const basic_regex& that) const - { - if(m_pimpl.get() == that.m_pimpl.get()) - return 0; - if(!m_pimpl.get()) - return -1; - if(!that.m_pimpl.get()) - return 1; - if(status() != that.status()) - return status() - that.status(); - if(flags() != that.flags()) - return flags() - that.flags(); - return str().compare(that.str()); - } - bool BOOST_REGEX_CALL operator==(const basic_regex& e)const - { - return compare(e) == 0; - } - bool BOOST_REGEX_CALL operator != (const basic_regex& e)const - { - return compare(e) != 0; - } - bool BOOST_REGEX_CALL operator<(const basic_regex& e)const - { - return compare(e) < 0; - } - bool BOOST_REGEX_CALL operator>(const basic_regex& e)const - { - return compare(e) > 0; - } - bool BOOST_REGEX_CALL operator<=(const basic_regex& e)const - { - return compare(e) <= 0; - } - bool BOOST_REGEX_CALL operator>=(const basic_regex& e)const - { - return compare(e) >= 0; - } - - // - // The following are deprecated as public interfaces - // but are available for compatibility with earlier versions. - const charT* BOOST_REGEX_CALL expression()const - { - return (m_pimpl.get() && !m_pimpl->status() ? m_pimpl->expression() : 0); - } - unsigned int BOOST_REGEX_CALL set_expression(const charT* p1, const charT* p2, flag_type f = regex_constants::normal) - { - assign(p1, p2, f | regex_constants::no_except); - return status(); - } - unsigned int BOOST_REGEX_CALL set_expression(const charT* p, flag_type f = regex_constants::normal) - { - assign(p, f | regex_constants::no_except); - return status(); - } - unsigned int BOOST_REGEX_CALL error_code()const - { - return status(); - } - // - // private access methods: - // - const re_detail::re_syntax_base* get_first_state()const - { - BOOST_ASSERT(0 != m_pimpl.get()); - return m_pimpl->get_first_state(); - } - unsigned get_restart_type()const - { - BOOST_ASSERT(0 != m_pimpl.get()); - return m_pimpl->get_restart_type(); - } - const unsigned char* get_map()const - { - BOOST_ASSERT(0 != m_pimpl.get()); - return m_pimpl->get_map(); - } - const ::cutl_details_boost::regex_traits_wrapper& get_traits()const - { - BOOST_ASSERT(0 != m_pimpl.get()); - return m_pimpl->get_traits(); - } - bool can_be_null()const - { - BOOST_ASSERT(0 != m_pimpl.get()); - return m_pimpl->can_be_null(); - } - const re_detail::regex_data& get_data()const - { - BOOST_ASSERT(0 != m_pimpl.get()); - return m_pimpl->get_data(); - } - cutl_details_boost::shared_ptr get_named_subs()const - { - return m_pimpl; - } - -private: - shared_ptr > m_pimpl; -}; - -// -// out of line members; -// these are the only members that mutate the basic_regex object, -// and are designed to provide the strong exception guarentee -// (in the event of a throw, the state of the object remains unchanged). -// -template -basic_regex& basic_regex::do_assign(const charT* p1, - const charT* p2, - flag_type f) -{ - shared_ptr > temp; - if(!m_pimpl.get()) - { - temp = shared_ptr >(new re_detail::basic_regex_implementation()); - } - else - { - temp = shared_ptr >(new re_detail::basic_regex_implementation(m_pimpl->m_ptraits)); - } - temp->assign(p1, p2, f); - temp.swap(m_pimpl); - return *this; -} - -template -typename basic_regex::locale_type BOOST_REGEX_CALL basic_regex::imbue(locale_type l) -{ - shared_ptr > temp(new re_detail::basic_regex_implementation()); - locale_type result = temp->imbue(l); - temp.swap(m_pimpl); - return result; -} - -// -// non-members: -// -template -void swap(basic_regex& e1, basic_regex& e2) -{ - e1.swap(e2); -} - -#ifndef BOOST_NO_STD_LOCALE -template -std::basic_ostream& - operator << (std::basic_ostream& os, - const basic_regex& e) -{ - return (os << e.str()); -} -#else -template -std::ostream& operator << (std::ostream& os, const basic_regex& e) -{ - return (os << e.str()); -} -#endif - -// -// class reg_expression: -// this is provided for backwards compatibility only, -// it is deprecated, no not use! -// -#ifdef BOOST_REGEX_NO_FWD -template > -#else -template -#endif -class reg_expression : public basic_regex -{ -public: - typedef typename basic_regex::flag_type flag_type; - typedef typename basic_regex::size_type size_type; - explicit reg_expression(){} - explicit reg_expression(const charT* p, flag_type f = regex_constants::normal) - : basic_regex(p, f){} - reg_expression(const charT* p1, const charT* p2, flag_type f = regex_constants::normal) - : basic_regex(p1, p2, f){} - reg_expression(const charT* p, size_type len, flag_type f) - : basic_regex(p, len, f){} - reg_expression(const reg_expression& that) - : basic_regex(that) {} - ~reg_expression(){} - reg_expression& BOOST_REGEX_CALL operator=(const reg_expression& that) - { - return this->assign(that); - } - -#if !defined(BOOST_NO_MEMBER_TEMPLATES) - template - explicit reg_expression(const std::basic_string& p, flag_type f = regex_constants::normal) - : basic_regex(p, f) - { - } - - template - reg_expression(InputIterator arg_first, InputIterator arg_last, flag_type f = regex_constants::normal) - : basic_regex(arg_first, arg_last, f) - { - } - - template - reg_expression& BOOST_REGEX_CALL operator=(const std::basic_string& p) - { - this->assign(p); - return *this; - } -#else - explicit reg_expression(const std::basic_string& p, flag_type f = regex_constants::normal) - : basic_regex(p, f) - { - } - - reg_expression& BOOST_REGEX_CALL operator=(const std::basic_string& p) - { - this->assign(p); - return *this; - } -#endif - -}; - -#ifdef BOOST_MSVC -#pragma warning (pop) -#endif - -} // namespace cutl_details_boost - -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable: 4103) -#endif -#ifdef BOOST_HAS_ABI_HEADERS -# include BOOST_ABI_SUFFIX -#endif -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif - -#endif - diff --git a/cutl/details/boost/regex/v4/basic_regex_creator.hpp b/cutl/details/boost/regex/v4/basic_regex_creator.hpp deleted file mode 100644 index 375d29b..0000000 --- a/cutl/details/boost/regex/v4/basic_regex_creator.hpp +++ /dev/null @@ -1,1571 +0,0 @@ -/* - * - * Copyright (c) 2004 - * John Maddock - * - * Use, modification and distribution are subject to the - * Boost Software License, Version 1.0. (See accompanying file - * LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt) - * - */ - - /* - * LOCATION: see http://www.boost.org for most recent version. - * FILE basic_regex_creator.cpp - * VERSION see - * DESCRIPTION: Declares template class basic_regex_creator which fills in - * the data members of a regex_data object. - */ - -#ifndef BOOST_REGEX_V4_BASIC_REGEX_CREATOR_HPP -#define BOOST_REGEX_V4_BASIC_REGEX_CREATOR_HPP - -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable: 4103) -#endif -#ifdef BOOST_HAS_ABI_HEADERS -# include BOOST_ABI_PREFIX -#endif -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif - -#ifdef BOOST_MSVC -# pragma warning(push) -# pragma warning(disable: 4800) -#endif - -namespace cutl_details_boost{ - -namespace re_detail{ - -template -struct digraph : public std::pair -{ - digraph() : std::pair(0, 0){} - digraph(charT c1) : std::pair(c1, 0){} - digraph(charT c1, charT c2) : std::pair(c1, c2) - {} -#if !BOOST_WORKAROUND(BOOST_MSVC, < 1300) - digraph(const digraph& d) : std::pair(d.first, d.second){} -#endif - template - digraph(const Seq& s) : std::pair() - { - BOOST_ASSERT(s.size() <= 2); - BOOST_ASSERT(s.size()); - this->first = s[0]; - this->second = (s.size() > 1) ? s[1] : 0; - } -}; - -template -class basic_char_set -{ -public: - typedef digraph digraph_type; - typedef typename traits::string_type string_type; - typedef typename traits::char_class_type m_type; - - basic_char_set() - { - m_negate = false; - m_has_digraphs = false; - m_classes = 0; - m_negated_classes = 0; - m_empty = true; - } - - void add_single(const digraph_type& s) - { - m_singles.insert(m_singles.end(), s); - if(s.second) - m_has_digraphs = true; - m_empty = false; - } - void add_range(const digraph_type& first, const digraph_type& end) - { - m_ranges.insert(m_ranges.end(), first); - m_ranges.insert(m_ranges.end(), end); - if(first.second) - { - m_has_digraphs = true; - add_single(first); - } - if(end.second) - { - m_has_digraphs = true; - add_single(end); - } - m_empty = false; - } - void add_class(m_type m) - { - m_classes |= m; - m_empty = false; - } - void add_negated_class(m_type m) - { - m_negated_classes |= m; - m_empty = false; - } - void add_equivalent(const digraph_type& s) - { - m_equivalents.insert(m_equivalents.end(), s); - if(s.second) - { - m_has_digraphs = true; - add_single(s); - } - m_empty = false; - } - void negate() - { - m_negate = true; - //m_empty = false; - } - - // - // accessor functions: - // - bool has_digraphs()const - { - return m_has_digraphs; - } - bool is_negated()const - { - return m_negate; - } - typedef typename std::vector::const_iterator list_iterator; - list_iterator singles_begin()const - { - return m_singles.begin(); - } - list_iterator singles_end()const - { - return m_singles.end(); - } - list_iterator ranges_begin()const - { - return m_ranges.begin(); - } - list_iterator ranges_end()const - { - return m_ranges.end(); - } - list_iterator equivalents_begin()const - { - return m_equivalents.begin(); - } - list_iterator equivalents_end()const - { - return m_equivalents.end(); - } - m_type classes()const - { - return m_classes; - } - m_type negated_classes()const - { - return m_negated_classes; - } - bool empty()const - { - return m_empty; - } -private: - std::vector m_singles; // a list of single characters to match - std::vector m_ranges; // a list of end points of our ranges - bool m_negate; // true if the set is to be negated - bool m_has_digraphs; // true if we have digraphs present - m_type m_classes; // character classes to match - m_type m_negated_classes; // negated character classes to match - bool m_empty; // whether we've added anything yet - std::vector m_equivalents; // a list of equivalence classes -}; - -template -class basic_regex_creator -{ -public: - basic_regex_creator(regex_data* data); - std::ptrdiff_t getoffset(void* addr) - { - return getoffset(addr, m_pdata->m_data.data()); - } - std::ptrdiff_t getoffset(const void* addr, const void* base) - { - return static_cast(addr) - static_cast(base); - } - re_syntax_base* getaddress(std::ptrdiff_t off) - { - return getaddress(off, m_pdata->m_data.data()); - } - re_syntax_base* getaddress(std::ptrdiff_t off, void* base) - { - return static_cast(static_cast(static_cast(base) + off)); - } - void init(unsigned l_flags) - { - m_pdata->m_flags = l_flags; - m_icase = l_flags & regex_constants::icase; - } - regbase::flag_type flags() - { - return m_pdata->m_flags; - } - void flags(regbase::flag_type f) - { - m_pdata->m_flags = f; - if(m_icase != static_cast(f & regbase::icase)) - { - m_icase = static_cast(f & regbase::icase); - } - } - re_syntax_base* append_state(syntax_element_type t, std::size_t s = sizeof(re_syntax_base)); - re_syntax_base* insert_state(std::ptrdiff_t pos, syntax_element_type t, std::size_t s = sizeof(re_syntax_base)); - re_literal* append_literal(charT c); - re_syntax_base* append_set(const basic_char_set& char_set); - re_syntax_base* append_set(const basic_char_set& char_set, mpl::false_*); - re_syntax_base* append_set(const basic_char_set& char_set, mpl::true_*); - void finalize(const charT* p1, const charT* p2); -protected: - regex_data* m_pdata; // pointer to the basic_regex_data struct we are filling in - const ::cutl_details_boost::regex_traits_wrapper& - m_traits; // convenience reference to traits class - re_syntax_base* m_last_state; // the last state we added - bool m_icase; // true for case insensitive matches - unsigned m_repeater_id; // the state_id of the next repeater - bool m_has_backrefs; // true if there are actually any backrefs - unsigned m_backrefs; // bitmask of permitted backrefs - cutl_details_boost::uintmax_t m_bad_repeats; // bitmask of repeats we can't deduce a startmap for; - bool m_has_recursions; // set when we have recursive expresisons to fixup - std::vector m_recursion_checks; // notes which recursions we've followed while analysing this expression - typename traits::char_class_type m_word_mask; // mask used to determine if a character is a word character - typename traits::char_class_type m_mask_space; // mask used to determine if a character is a word character - typename traits::char_class_type m_lower_mask; // mask used to determine if a character is a lowercase character - typename traits::char_class_type m_upper_mask; // mask used to determine if a character is an uppercase character - typename traits::char_class_type m_alpha_mask; // mask used to determine if a character is an alphabetic character -private: - basic_regex_creator& operator=(const basic_regex_creator&); - basic_regex_creator(const basic_regex_creator&); - - void fixup_pointers(re_syntax_base* state); - void fixup_recursions(re_syntax_base* state); - void create_startmaps(re_syntax_base* state); - int calculate_backstep(re_syntax_base* state); - void create_startmap(re_syntax_base* state, unsigned char* l_map, unsigned int* pnull, unsigned char mask); - unsigned get_restart_type(re_syntax_base* state); - void set_all_masks(unsigned char* bits, unsigned char); - bool is_bad_repeat(re_syntax_base* pt); - void set_bad_repeat(re_syntax_base* pt); - syntax_element_type get_repeat_type(re_syntax_base* state); - void probe_leading_repeat(re_syntax_base* state); -}; - -template -basic_regex_creator::basic_regex_creator(regex_data* data) - : m_pdata(data), m_traits(*(data->m_ptraits)), m_last_state(0), m_repeater_id(0), m_has_backrefs(false), m_backrefs(0), m_has_recursions(false) -{ - m_pdata->m_data.clear(); - m_pdata->m_status = ::cutl_details_boost::regex_constants::error_ok; - static const charT w = 'w'; - static const charT s = 's'; - static const charT l[5] = { 'l', 'o', 'w', 'e', 'r', }; - static const charT u[5] = { 'u', 'p', 'p', 'e', 'r', }; - static const charT a[5] = { 'a', 'l', 'p', 'h', 'a', }; - m_word_mask = m_traits.lookup_classname(&w, &w +1); - m_mask_space = m_traits.lookup_classname(&s, &s +1); - m_lower_mask = m_traits.lookup_classname(l, l + 5); - m_upper_mask = m_traits.lookup_classname(u, u + 5); - m_alpha_mask = m_traits.lookup_classname(a, a + 5); - m_pdata->m_word_mask = m_word_mask; - BOOST_ASSERT(m_word_mask != 0); - BOOST_ASSERT(m_mask_space != 0); - BOOST_ASSERT(m_lower_mask != 0); - BOOST_ASSERT(m_upper_mask != 0); - BOOST_ASSERT(m_alpha_mask != 0); -} - -template -re_syntax_base* basic_regex_creator::append_state(syntax_element_type t, std::size_t s) -{ - // if the state is a backref then make a note of it: - if(t == syntax_element_backref) - this->m_has_backrefs = true; - // append a new state, start by aligning our last one: - m_pdata->m_data.align(); - // set the offset to the next state in our last one: - if(m_last_state) - m_last_state->next.i = m_pdata->m_data.size() - getoffset(m_last_state); - // now actually extent our data: - m_last_state = static_cast(m_pdata->m_data.extend(s)); - // fill in boilerplate options in the new state: - m_last_state->next.i = 0; - m_last_state->type = t; - return m_last_state; -} - -template -re_syntax_base* basic_regex_creator::insert_state(std::ptrdiff_t pos, syntax_element_type t, std::size_t s) -{ - // append a new state, start by aligning our last one: - m_pdata->m_data.align(); - // set the offset to the next state in our last one: - if(m_last_state) - m_last_state->next.i = m_pdata->m_data.size() - getoffset(m_last_state); - // remember the last state position: - std::ptrdiff_t off = getoffset(m_last_state) + s; - // now actually insert our data: - re_syntax_base* new_state = static_cast(m_pdata->m_data.insert(pos, s)); - // fill in boilerplate options in the new state: - new_state->next.i = s; - new_state->type = t; - m_last_state = getaddress(off); - return new_state; -} - -template -re_literal* basic_regex_creator::append_literal(charT c) -{ - re_literal* result; - // start by seeing if we have an existing re_literal we can extend: - if((0 == m_last_state) || (m_last_state->type != syntax_element_literal)) - { - // no existing re_literal, create a new one: - result = static_cast(append_state(syntax_element_literal, sizeof(re_literal) + sizeof(charT))); - result->length = 1; - *static_cast(static_cast(result+1)) = m_traits.translate(c, m_icase); - } - else - { - // we have an existing re_literal, extend it: - std::ptrdiff_t off = getoffset(m_last_state); - m_pdata->m_data.extend(sizeof(charT)); - m_last_state = result = static_cast(getaddress(off)); - charT* characters = static_cast(static_cast(result+1)); - characters[result->length] = m_traits.translate(c, m_icase); - ++(result->length); - } - return result; -} - -template -inline re_syntax_base* basic_regex_creator::append_set( - const basic_char_set& char_set) -{ - typedef mpl::bool_< (sizeof(charT) == 1) > truth_type; - return char_set.has_digraphs() - ? append_set(char_set, static_cast(0)) - : append_set(char_set, static_cast(0)); -} - -template -re_syntax_base* basic_regex_creator::append_set( - const basic_char_set& char_set, mpl::false_*) -{ - typedef typename traits::string_type string_type; - typedef typename basic_char_set::list_iterator item_iterator; - typedef typename traits::char_class_type m_type; - - re_set_long* result = static_cast*>(append_state(syntax_element_long_set, sizeof(re_set_long))); - // - // fill in the basics: - // - result->csingles = static_cast(::cutl_details_boost::re_detail::distance(char_set.singles_begin(), char_set.singles_end())); - result->cranges = static_cast(::cutl_details_boost::re_detail::distance(char_set.ranges_begin(), char_set.ranges_end())) / 2; - result->cequivalents = static_cast(::cutl_details_boost::re_detail::distance(char_set.equivalents_begin(), char_set.equivalents_end())); - result->cclasses = char_set.classes(); - result->cnclasses = char_set.negated_classes(); - if(flags() & regbase::icase) - { - // adjust classes as needed: - if(((result->cclasses & m_lower_mask) == m_lower_mask) || ((result->cclasses & m_upper_mask) == m_upper_mask)) - result->cclasses |= m_alpha_mask; - if(((result->cnclasses & m_lower_mask) == m_lower_mask) || ((result->cnclasses & m_upper_mask) == m_upper_mask)) - result->cnclasses |= m_alpha_mask; - } - - result->isnot = char_set.is_negated(); - result->singleton = !char_set.has_digraphs(); - // - // remember where the state is for later: - // - std::ptrdiff_t offset = getoffset(result); - // - // now extend with all the singles: - // - item_iterator first, last; - first = char_set.singles_begin(); - last = char_set.singles_end(); - while(first != last) - { - charT* p = static_cast(this->m_pdata->m_data.extend(sizeof(charT) * (first->second ? 3 : 2))); - p[0] = m_traits.translate(first->first, m_icase); - if(first->second) - { - p[1] = m_traits.translate(first->second, m_icase); - p[2] = 0; - } - else - p[1] = 0; - ++first; - } - // - // now extend with all the ranges: - // - first = char_set.ranges_begin(); - last = char_set.ranges_end(); - while(first != last) - { - // first grab the endpoints of the range: - digraph c1 = *first; - c1.first = this->m_traits.translate(c1.first, this->m_icase); - c1.second = this->m_traits.translate(c1.second, this->m_icase); - ++first; - digraph c2 = *first; - c2.first = this->m_traits.translate(c2.first, this->m_icase); - c2.second = this->m_traits.translate(c2.second, this->m_icase); - ++first; - string_type s1, s2; - // different actions now depending upon whether collation is turned on: - if(flags() & regex_constants::collate) - { - // we need to transform our range into sort keys: -#if BOOST_WORKAROUND(__GNUC__, < 3) - string_type in(3, charT(0)); - in[0] = c1.first; - in[1] = c1.second; - s1 = this->m_traits.transform(in.c_str(), (in[1] ? in.c_str()+2 : in.c_str()+1)); - in[0] = c2.first; - in[1] = c2.second; - s2 = this->m_traits.transform(in.c_str(), (in[1] ? in.c_str()+2 : in.c_str()+1)); -#else - charT a1[3] = { c1.first, c1.second, charT(0), }; - charT a2[3] = { c2.first, c2.second, charT(0), }; - s1 = this->m_traits.transform(a1, (a1[1] ? a1+2 : a1+1)); - s2 = this->m_traits.transform(a2, (a2[1] ? a2+2 : a2+1)); -#endif - if(s1.size() == 0) - s1 = string_type(1, charT(0)); - if(s2.size() == 0) - s2 = string_type(1, charT(0)); - } - else - { - if(c1.second) - { - s1.insert(s1.end(), c1.first); - s1.insert(s1.end(), c1.second); - } - else - s1 = string_type(1, c1.first); - if(c2.second) - { - s2.insert(s2.end(), c2.first); - s2.insert(s2.end(), c2.second); - } - else - s2.insert(s2.end(), c2.first); - } - if(s1 > s2) - { - // Oops error: - return 0; - } - charT* p = static_cast(this->m_pdata->m_data.extend(sizeof(charT) * (s1.size() + s2.size() + 2) ) ); - re_detail::copy(s1.begin(), s1.end(), p); - p[s1.size()] = charT(0); - p += s1.size() + 1; - re_detail::copy(s2.begin(), s2.end(), p); - p[s2.size()] = charT(0); - } - // - // now process the equivalence classes: - // - first = char_set.equivalents_begin(); - last = char_set.equivalents_end(); - while(first != last) - { - string_type s; - if(first->second) - { -#if BOOST_WORKAROUND(__GNUC__, < 3) - string_type in(3, charT(0)); - in[0] = first->first; - in[1] = first->second; - s = m_traits.transform_primary(in.c_str(), in.c_str()+2); -#else - charT cs[3] = { first->first, first->second, charT(0), }; - s = m_traits.transform_primary(cs, cs+2); -#endif - } - else - s = m_traits.transform_primary(&first->first, &first->first+1); - if(s.empty()) - return 0; // invalid or unsupported equivalence class - charT* p = static_cast(this->m_pdata->m_data.extend(sizeof(charT) * (s.size()+1) ) ); - re_detail::copy(s.begin(), s.end(), p); - p[s.size()] = charT(0); - ++first; - } - // - // finally reset the address of our last state: - // - m_last_state = result = static_cast*>(getaddress(offset)); - return result; -} - -template -inline bool char_less(T t1, T t2) -{ - return t1 < t2; -} -inline bool char_less(char t1, char t2) -{ - return static_cast(t1) < static_cast(t2); -} -inline bool char_less(signed char t1, signed char t2) -{ - return static_cast(t1) < static_cast(t2); -} - -template -re_syntax_base* basic_regex_creator::append_set( - const basic_char_set& char_set, mpl::true_*) -{ - typedef typename traits::string_type string_type; - typedef typename basic_char_set::list_iterator item_iterator; - - re_set* result = static_cast(append_state(syntax_element_set, sizeof(re_set))); - bool negate = char_set.is_negated(); - std::memset(result->_map, 0, sizeof(result->_map)); - // - // handle singles first: - // - item_iterator first, last; - first = char_set.singles_begin(); - last = char_set.singles_end(); - while(first != last) - { - for(unsigned int i = 0; i < (1 << CHAR_BIT); ++i) - { - if(this->m_traits.translate(static_cast(i), this->m_icase) - == this->m_traits.translate(first->first, this->m_icase)) - result->_map[i] = true; - } - ++first; - } - // - // OK now handle ranges: - // - first = char_set.ranges_begin(); - last = char_set.ranges_end(); - while(first != last) - { - // first grab the endpoints of the range: - charT c1 = this->m_traits.translate(first->first, this->m_icase); - ++first; - charT c2 = this->m_traits.translate(first->first, this->m_icase); - ++first; - // different actions now depending upon whether collation is turned on: - if(flags() & regex_constants::collate) - { - // we need to transform our range into sort keys: - charT c3[2] = { c1, charT(0), }; - string_type s1 = this->m_traits.transform(c3, c3+1); - c3[0] = c2; - string_type s2 = this->m_traits.transform(c3, c3+1); - if(s1 > s2) - { - // Oops error: - return 0; - } - BOOST_ASSERT(c3[1] == charT(0)); - for(unsigned i = 0; i < (1u << CHAR_BIT); ++i) - { - c3[0] = static_cast(i); - string_type s3 = this->m_traits.transform(c3, c3 +1); - if((s1 <= s3) && (s3 <= s2)) - result->_map[i] = true; - } - } - else - { - if(char_less(c2, c1)) - { - // Oops error: - return 0; - } - // everything in range matches: - std::memset(result->_map + static_cast(c1), true, 1 + static_cast(c2) - static_cast(c1)); - } - } - // - // and now the classes: - // - typedef typename traits::char_class_type m_type; - m_type m = char_set.classes(); - if(flags() & regbase::icase) - { - // adjust m as needed: - if(((m & m_lower_mask) == m_lower_mask) || ((m & m_upper_mask) == m_upper_mask)) - m |= m_alpha_mask; - } - if(m != 0) - { - for(unsigned i = 0; i < (1u << CHAR_BIT); ++i) - { - if(this->m_traits.isctype(static_cast(i), m)) - result->_map[i] = true; - } - } - // - // and now the negated classes: - // - m = char_set.negated_classes(); - if(flags() & regbase::icase) - { - // adjust m as needed: - if(((m & m_lower_mask) == m_lower_mask) || ((m & m_upper_mask) == m_upper_mask)) - m |= m_alpha_mask; - } - if(m != 0) - { - for(unsigned i = 0; i < (1u << CHAR_BIT); ++i) - { - if(0 == this->m_traits.isctype(static_cast(i), m)) - result->_map[i] = true; - } - } - // - // now process the equivalence classes: - // - first = char_set.equivalents_begin(); - last = char_set.equivalents_end(); - while(first != last) - { - string_type s; - BOOST_ASSERT(static_cast(0) == first->second); - s = m_traits.transform_primary(&first->first, &first->first+1); - if(s.empty()) - return 0; // invalid or unsupported equivalence class - for(unsigned i = 0; i < (1u << CHAR_BIT); ++i) - { - charT c[2] = { (static_cast(i)), charT(0), }; - string_type s2 = this->m_traits.transform_primary(c, c+1); - if(s == s2) - result->_map[i] = true; - } - ++first; - } - if(negate) - { - for(unsigned i = 0; i < (1u << CHAR_BIT); ++i) - { - result->_map[i] = !(result->_map[i]); - } - } - return result; -} - -template -void basic_regex_creator::finalize(const charT* p1, const charT* p2) -{ - if(this->m_pdata->m_status) - return; - // we've added all the states we need, now finish things off. - // start by adding a terminating state: - append_state(syntax_element_match); - // extend storage to store original expression: - std::ptrdiff_t len = p2 - p1; - m_pdata->m_expression_len = len; - charT* ps = static_cast(m_pdata->m_data.extend(sizeof(charT) * (1 + (p2 - p1)))); - m_pdata->m_expression = ps; - re_detail::copy(p1, p2, ps); - ps[p2 - p1] = 0; - // fill in our other data... - // successful parsing implies a zero status: - m_pdata->m_status = 0; - // get the first state of the machine: - m_pdata->m_first_state = static_cast(m_pdata->m_data.data()); - // fixup pointers in the machine: - fixup_pointers(m_pdata->m_first_state); - if(m_has_recursions) - { - m_pdata->m_has_recursions = true; - fixup_recursions(m_pdata->m_first_state); - if(this->m_pdata->m_status) - return; - } - else - m_pdata->m_has_recursions = false; - // create nested startmaps: - create_startmaps(m_pdata->m_first_state); - // create main startmap: - std::memset(m_pdata->m_startmap, 0, sizeof(m_pdata->m_startmap)); - m_pdata->m_can_be_null = 0; - - m_bad_repeats = 0; - if(m_has_recursions) - m_recursion_checks.assign(1 + m_pdata->m_mark_count, false); - create_startmap(m_pdata->m_first_state, m_pdata->m_startmap, &(m_pdata->m_can_be_null), mask_all); - // get the restart type: - m_pdata->m_restart_type = get_restart_type(m_pdata->m_first_state); - // optimise a leading repeat if there is one: - probe_leading_repeat(m_pdata->m_first_state); -} - -template -void basic_regex_creator::fixup_pointers(re_syntax_base* state) -{ - while(state) - { - switch(state->type) - { - case syntax_element_recurse: - m_has_recursions = true; - if(state->next.i) - state->next.p = getaddress(state->next.i, state); - else - state->next.p = 0; - break; - case syntax_element_rep: - case syntax_element_dot_rep: - case syntax_element_char_rep: - case syntax_element_short_set_rep: - case syntax_element_long_set_rep: - // set the state_id of this repeat: - static_cast(state)->state_id = m_repeater_id++; - BOOST_FALLTHROUGH; - case syntax_element_alt: - std::memset(static_cast(state)->_map, 0, sizeof(static_cast(state)->_map)); - static_cast(state)->can_be_null = 0; - BOOST_FALLTHROUGH; - case syntax_element_jump: - static_cast(state)->alt.p = getaddress(static_cast(state)->alt.i, state); - BOOST_FALLTHROUGH; - default: - if(state->next.i) - state->next.p = getaddress(state->next.i, state); - else - state->next.p = 0; - } - state = state->next.p; - } -} - -template -void basic_regex_creator::fixup_recursions(re_syntax_base* state) -{ - re_syntax_base* base = state; - while(state) - { - switch(state->type) - { - case syntax_element_assert_backref: - { - // just check that the index is valid: - int idx = static_cast(state)->index; - if(idx < 0) - { - idx = -idx-1; - if(idx >= 10000) - { - idx = m_pdata->get_id(idx); - if(idx <= 0) - { - // check of sub-expression that doesn't exist: - if(0 == this->m_pdata->m_status) // update the error code if not already set - this->m_pdata->m_status = cutl_details_boost::regex_constants::error_bad_pattern; - // - // clear the expression, we should be empty: - // - this->m_pdata->m_expression = 0; - this->m_pdata->m_expression_len = 0; - // - // and throw if required: - // - if(0 == (this->flags() & regex_constants::no_except)) - { - std::string message = "Encountered a forward reference to a marked sub-expression that does not exist."; - cutl_details_boost::regex_error e(message, cutl_details_boost::regex_constants::error_bad_pattern, 0); - e.raise(); - } - } - } - } - } - break; - case syntax_element_recurse: - { - bool ok = false; - re_syntax_base* p = base; - std::ptrdiff_t idx = static_cast(state)->alt.i; - if(idx > 10000) - { - // - // There may be more than one capture group with this hash, just do what Perl - // does and recurse to the leftmost: - // - idx = m_pdata->get_id(static_cast(idx)); - } - while(p) - { - if((p->type == syntax_element_startmark) && (static_cast(p)->index == idx)) - { - // - // We've found the target of the recursion, set the jump target: - // - static_cast(state)->alt.p = p; - ok = true; - // - // Now scan the target for nested repeats: - // - p = p->next.p; - int next_rep_id = 0; - while(p) - { - switch(p->type) - { - case syntax_element_rep: - case syntax_element_dot_rep: - case syntax_element_char_rep: - case syntax_element_short_set_rep: - case syntax_element_long_set_rep: - next_rep_id = static_cast(p)->state_id; - break; - case syntax_element_endmark: - if(static_cast(p)->index == idx) - next_rep_id = -1; - break; - default: - break; - } - if(next_rep_id) - break; - p = p->next.p; - } - if(next_rep_id > 0) - { - static_cast(state)->state_id = next_rep_id - 1; - } - - break; - } - p = p->next.p; - } - if(!ok) - { - // recursion to sub-expression that doesn't exist: - if(0 == this->m_pdata->m_status) // update the error code if not already set - this->m_pdata->m_status = cutl_details_boost::regex_constants::error_bad_pattern; - // - // clear the expression, we should be empty: - // - this->m_pdata->m_expression = 0; - this->m_pdata->m_expression_len = 0; - // - // and throw if required: - // - if(0 == (this->flags() & regex_constants::no_except)) - { - std::string message = "Encountered a forward reference to a recursive sub-expression that does not exist."; - cutl_details_boost::regex_error e(message, cutl_details_boost::regex_constants::error_bad_pattern, 0); - e.raise(); - } - } - } - break; - default: - break; - } - state = state->next.p; - } -} - -template -void basic_regex_creator::create_startmaps(re_syntax_base* state) -{ - // non-recursive implementation: - // create the last map in the machine first, so that earlier maps - // can make use of the result... - // - // This was originally a recursive implementation, but that caused stack - // overflows with complex expressions on small stacks (think COM+). - - // start by saving the case setting: - bool l_icase = m_icase; - std::vector > v; - - while(state) - { - switch(state->type) - { - case syntax_element_toggle_case: - // we need to track case changes here: - m_icase = static_cast(state)->icase; - state = state->next.p; - continue; - case syntax_element_alt: - case syntax_element_rep: - case syntax_element_dot_rep: - case syntax_element_char_rep: - case syntax_element_short_set_rep: - case syntax_element_long_set_rep: - // just push the state onto our stack for now: - v.push_back(std::pair(m_icase, state)); - state = state->next.p; - break; - case syntax_element_backstep: - // we need to calculate how big the backstep is: - static_cast(state)->index - = this->calculate_backstep(state->next.p); - if(static_cast(state)->index < 0) - { - // Oops error: - if(0 == this->m_pdata->m_status) // update the error code if not already set - this->m_pdata->m_status = cutl_details_boost::regex_constants::error_bad_pattern; - // - // clear the expression, we should be empty: - // - this->m_pdata->m_expression = 0; - this->m_pdata->m_expression_len = 0; - // - // and throw if required: - // - if(0 == (this->flags() & regex_constants::no_except)) - { - std::string message = "Invalid lookbehind assertion encountered in the regular expression."; - cutl_details_boost::regex_error e(message, cutl_details_boost::regex_constants::error_bad_pattern, 0); - e.raise(); - } - } - BOOST_FALLTHROUGH; - default: - state = state->next.p; - } - } - - // now work through our list, building all the maps as we go: - while(v.size()) - { - // Initialize m_recursion_checks if we need it: - if(m_has_recursions) - m_recursion_checks.assign(1 + m_pdata->m_mark_count, false); - - const std::pair& p = v.back(); - m_icase = p.first; - state = p.second; - v.pop_back(); - - // Build maps: - m_bad_repeats = 0; - create_startmap(state->next.p, static_cast(state)->_map, &static_cast(state)->can_be_null, mask_take); - m_bad_repeats = 0; - - if(m_has_recursions) - m_recursion_checks.assign(1 + m_pdata->m_mark_count, false); - create_startmap(static_cast(state)->alt.p, static_cast(state)->_map, &static_cast(state)->can_be_null, mask_skip); - // adjust the type of the state to allow for faster matching: - state->type = this->get_repeat_type(state); - } - // restore case sensitivity: - m_icase = l_icase; -} - -template -int basic_regex_creator::calculate_backstep(re_syntax_base* state) -{ - typedef typename traits::char_class_type m_type; - int result = 0; - while(state) - { - switch(state->type) - { - case syntax_element_startmark: - if((static_cast(state)->index == -1) - || (static_cast(state)->index == -2)) - { - state = static_cast(state->next.p)->alt.p->next.p; - continue; - } - else if(static_cast(state)->index == -3) - { - state = state->next.p->next.p; - continue; - } - break; - case syntax_element_endmark: - if((static_cast(state)->index == -1) - || (static_cast(state)->index == -2)) - return result; - break; - case syntax_element_literal: - result += static_cast(state)->length; - break; - case syntax_element_wild: - case syntax_element_set: - result += 1; - break; - case syntax_element_dot_rep: - case syntax_element_char_rep: - case syntax_element_short_set_rep: - case syntax_element_backref: - case syntax_element_rep: - case syntax_element_combining: - case syntax_element_long_set_rep: - case syntax_element_backstep: - { - re_repeat* rep = static_cast(state); - // adjust the type of the state to allow for faster matching: - state->type = this->get_repeat_type(state); - if((state->type == syntax_element_dot_rep) - || (state->type == syntax_element_char_rep) - || (state->type == syntax_element_short_set_rep)) - { - if(rep->max != rep->min) - return -1; - result += static_cast(rep->min); - state = rep->alt.p; - continue; - } - else if(state->type == syntax_element_long_set_rep) - { - BOOST_ASSERT(rep->next.p->type == syntax_element_long_set); - if(static_cast*>(rep->next.p)->singleton == 0) - return -1; - if(rep->max != rep->min) - return -1; - result += static_cast(rep->min); - state = rep->alt.p; - continue; - } - } - return -1; - case syntax_element_long_set: - if(static_cast*>(state)->singleton == 0) - return -1; - result += 1; - break; - case syntax_element_jump: - state = static_cast(state)->alt.p; - continue; - case syntax_element_alt: - { - int r1 = calculate_backstep(state->next.p); - int r2 = calculate_backstep(static_cast(state)->alt.p); - if((r1 < 0) || (r1 != r2)) - return -1; - return result + r1; - } - default: - break; - } - state = state->next.p; - } - return -1; -} - -template -void basic_regex_creator::create_startmap(re_syntax_base* state, unsigned char* l_map, unsigned int* pnull, unsigned char mask) -{ - int not_last_jump = 1; - re_syntax_base* recursion_start = 0; - int recursion_sub = 0; - re_syntax_base* recursion_restart = 0; - - // track case sensitivity: - bool l_icase = m_icase; - - while(state) - { - switch(state->type) - { - case syntax_element_toggle_case: - l_icase = static_cast(state)->icase; - state = state->next.p; - break; - case syntax_element_literal: - { - // don't set anything in *pnull, set each element in l_map - // that could match the first character in the literal: - if(l_map) - { - l_map[0] |= mask_init; - charT first_char = *static_cast(static_cast(static_cast(state) + 1)); - for(unsigned int i = 0; i < (1u << CHAR_BIT); ++i) - { - if(m_traits.translate(static_cast(i), l_icase) == first_char) - l_map[i] |= mask; - } - } - return; - } - case syntax_element_end_line: - { - // next character must be a line separator (if there is one): - if(l_map) - { - l_map[0] |= mask_init; - l_map[static_cast('\n')] |= mask; - l_map[static_cast('\r')] |= mask; - l_map[static_cast('\f')] |= mask; - l_map[0x85] |= mask; - } - // now figure out if we can match a NULL string at this point: - if(pnull) - create_startmap(state->next.p, 0, pnull, mask); - return; - } - case syntax_element_recurse: - { - if(state->type == syntax_element_startmark) - recursion_sub = static_cast(state)->index; - else - recursion_sub = 0; - if(m_recursion_checks[recursion_sub]) - { - // Infinite recursion!! - if(0 == this->m_pdata->m_status) // update the error code if not already set - this->m_pdata->m_status = cutl_details_boost::regex_constants::error_bad_pattern; - // - // clear the expression, we should be empty: - // - this->m_pdata->m_expression = 0; - this->m_pdata->m_expression_len = 0; - // - // and throw if required: - // - if(0 == (this->flags() & regex_constants::no_except)) - { - std::string message = "Encountered an infinite recursion."; - cutl_details_boost::regex_error e(message, cutl_details_boost::regex_constants::error_bad_pattern, 0); - e.raise(); - } - } - else if(recursion_start == 0) - { - recursion_start = state; - recursion_restart = state->next.p; - state = static_cast(state)->alt.p; - m_recursion_checks[recursion_sub] = true; - break; - } - m_recursion_checks[recursion_sub] = true; - // can't handle nested recursion here... - BOOST_FALLTHROUGH; - } - case syntax_element_backref: - // can be null, and any character can match: - if(pnull) - *pnull |= mask; - BOOST_FALLTHROUGH; - case syntax_element_wild: - { - // can't be null, any character can match: - set_all_masks(l_map, mask); - return; - } - case syntax_element_match: - { - // must be null, any character can match: - set_all_masks(l_map, mask); - if(pnull) - *pnull |= mask; - return; - } - case syntax_element_word_start: - { - // recurse, then AND with all the word characters: - create_startmap(state->next.p, l_map, pnull, mask); - if(l_map) - { - l_map[0] |= mask_init; - for(unsigned int i = 0; i < (1u << CHAR_BIT); ++i) - { - if(!m_traits.isctype(static_cast(i), m_word_mask)) - l_map[i] &= static_cast(~mask); - } - } - return; - } - case syntax_element_word_end: - { - // recurse, then AND with all the word characters: - create_startmap(state->next.p, l_map, pnull, mask); - if(l_map) - { - l_map[0] |= mask_init; - for(unsigned int i = 0; i < (1u << CHAR_BIT); ++i) - { - if(m_traits.isctype(static_cast(i), m_word_mask)) - l_map[i] &= static_cast(~mask); - } - } - return; - } - case syntax_element_buffer_end: - { - // we *must be null* : - if(pnull) - *pnull |= mask; - return; - } - case syntax_element_long_set: - if(l_map) - { - typedef typename traits::char_class_type m_type; - if(static_cast*>(state)->singleton) - { - l_map[0] |= mask_init; - for(unsigned int i = 0; i < (1u << CHAR_BIT); ++i) - { - charT c = static_cast(i); - if(&c != re_is_set_member(&c, &c + 1, static_cast*>(state), *m_pdata, l_icase)) - l_map[i] |= mask; - } - } - else - set_all_masks(l_map, mask); - } - return; - case syntax_element_set: - if(l_map) - { - l_map[0] |= mask_init; - for(unsigned int i = 0; i < (1u << CHAR_BIT); ++i) - { - if(static_cast(state)->_map[ - static_cast(m_traits.translate(static_cast(i), l_icase))]) - l_map[i] |= mask; - } - } - return; - case syntax_element_jump: - // take the jump: - state = static_cast(state)->alt.p; - not_last_jump = -1; - break; - case syntax_element_alt: - case syntax_element_rep: - case syntax_element_dot_rep: - case syntax_element_char_rep: - case syntax_element_short_set_rep: - case syntax_element_long_set_rep: - { - re_alt* rep = static_cast(state); - if(rep->_map[0] & mask_init) - { - if(l_map) - { - // copy previous results: - l_map[0] |= mask_init; - for(unsigned int i = 0; i <= UCHAR_MAX; ++i) - { - if(rep->_map[i] & mask_any) - l_map[i] |= mask; - } - } - if(pnull) - { - if(rep->can_be_null & mask_any) - *pnull |= mask; - } - } - else - { - // we haven't created a startmap for this alternative yet - // so take the union of the two options: - if(is_bad_repeat(state)) - { - set_all_masks(l_map, mask); - if(pnull) - *pnull |= mask; - return; - } - set_bad_repeat(state); - create_startmap(state->next.p, l_map, pnull, mask); - if((state->type == syntax_element_alt) - || (static_cast(state)->min == 0) - || (not_last_jump == 0)) - create_startmap(rep->alt.p, l_map, pnull, mask); - } - } - return; - case syntax_element_soft_buffer_end: - // match newline or null: - if(l_map) - { - l_map[0] |= mask_init; - l_map[static_cast('\n')] |= mask; - l_map[static_cast('\r')] |= mask; - } - if(pnull) - *pnull |= mask; - return; - case syntax_element_endmark: - // need to handle independent subs as a special case: - if(static_cast(state)->index < 0) - { - // can be null, any character can match: - set_all_masks(l_map, mask); - if(pnull) - *pnull |= mask; - return; - } - else if(recursion_start && (recursion_sub != 0) && (recursion_sub == static_cast(state)->index)) - { - // recursion termination: - recursion_start = 0; - state = recursion_restart; - break; - } - - // - // Normally we just go to the next state... but if this sub-expression is - // the target of a recursion, then we might be ending a recursion, in which - // case we should check whatever follows that recursion, as well as whatever - // follows this state: - // - if(m_pdata->m_has_recursions && static_cast(state)->index) - { - bool ok = false; - re_syntax_base* p = m_pdata->m_first_state; - while(p) - { - if(p->type == syntax_element_recurse) - { - re_brace* p2 = static_cast(static_cast(p)->alt.p); - if((p2->type == syntax_element_startmark) && (p2->index == static_cast(state)->index)) - { - ok = true; - break; - } - } - p = p->next.p; - } - if(ok) - { - create_startmap(p->next.p, l_map, pnull, mask); - } - } - state = state->next.p; - break; - - case syntax_element_startmark: - // need to handle independent subs as a special case: - if(static_cast(state)->index == -3) - { - state = state->next.p->next.p; - break; - } - BOOST_FALLTHROUGH; - default: - state = state->next.p; - } - ++not_last_jump; - } -} - -template -unsigned basic_regex_creator::get_restart_type(re_syntax_base* state) -{ - // - // find out how the machine starts, so we can optimise the search: - // - while(state) - { - switch(state->type) - { - case syntax_element_startmark: - case syntax_element_endmark: - state = state->next.p; - continue; - case syntax_element_start_line: - return regbase::restart_line; - case syntax_element_word_start: - return regbase::restart_word; - case syntax_element_buffer_start: - return regbase::restart_buf; - case syntax_element_restart_continue: - return regbase::restart_continue; - default: - state = 0; - continue; - } - } - return regbase::restart_any; -} - -template -void basic_regex_creator::set_all_masks(unsigned char* bits, unsigned char mask) -{ - // - // set mask in all of bits elements, - // if bits[0] has mask_init not set then we can - // optimise this to a call to memset: - // - if(bits) - { - if(bits[0] == 0) - (std::memset)(bits, mask, 1u << CHAR_BIT); - else - { - for(unsigned i = 0; i < (1u << CHAR_BIT); ++i) - bits[i] |= mask; - } - bits[0] |= mask_init; - } -} - -template -bool basic_regex_creator::is_bad_repeat(re_syntax_base* pt) -{ - switch(pt->type) - { - case syntax_element_rep: - case syntax_element_dot_rep: - case syntax_element_char_rep: - case syntax_element_short_set_rep: - case syntax_element_long_set_rep: - { - unsigned state_id = static_cast(pt)->state_id; - if(state_id > sizeof(m_bad_repeats) * CHAR_BIT) - return true; // run out of bits, assume we can't traverse this one. - static const cutl_details_boost::uintmax_t one = 1uL; - return m_bad_repeats & (one << state_id); - } - default: - return false; - } -} - -template -void basic_regex_creator::set_bad_repeat(re_syntax_base* pt) -{ - switch(pt->type) - { - case syntax_element_rep: - case syntax_element_dot_rep: - case syntax_element_char_rep: - case syntax_element_short_set_rep: - case syntax_element_long_set_rep: - { - unsigned state_id = static_cast(pt)->state_id; - static const cutl_details_boost::uintmax_t one = 1uL; - if(state_id <= sizeof(m_bad_repeats) * CHAR_BIT) - m_bad_repeats |= (one << state_id); - } - break; - default: - break; - } -} - -template -syntax_element_type basic_regex_creator::get_repeat_type(re_syntax_base* state) -{ - typedef typename traits::char_class_type m_type; - if(state->type == syntax_element_rep) - { - // check to see if we are repeating a single state: - if(state->next.p->next.p->next.p == static_cast(state)->alt.p) - { - switch(state->next.p->type) - { - case re_detail::syntax_element_wild: - return re_detail::syntax_element_dot_rep; - case re_detail::syntax_element_literal: - return re_detail::syntax_element_char_rep; - case re_detail::syntax_element_set: - return re_detail::syntax_element_short_set_rep; - case re_detail::syntax_element_long_set: - if(static_cast*>(state->next.p)->singleton) - return re_detail::syntax_element_long_set_rep; - break; - default: - break; - } - } - } - return state->type; -} - -template -void basic_regex_creator::probe_leading_repeat(re_syntax_base* state) -{ - // enumerate our states, and see if we have a leading repeat - // for which failed search restarts can be optimised; - do - { - switch(state->type) - { - case syntax_element_startmark: - if(static_cast(state)->index >= 0) - { - state = state->next.p; - continue; - } - if((static_cast(state)->index == -1) - || (static_cast(state)->index == -2)) - { - // skip past the zero width assertion: - state = static_cast(state->next.p)->alt.p->next.p; - continue; - } - if(static_cast(state)->index == -3) - { - // Have to skip the leading jump state: - state = state->next.p->next.p; - continue; - } - return; - case syntax_element_endmark: - case syntax_element_start_line: - case syntax_element_end_line: - case syntax_element_word_boundary: - case syntax_element_within_word: - case syntax_element_word_start: - case syntax_element_word_end: - case syntax_element_buffer_start: - case syntax_element_buffer_end: - case syntax_element_restart_continue: - state = state->next.p; - break; - case syntax_element_dot_rep: - case syntax_element_char_rep: - case syntax_element_short_set_rep: - case syntax_element_long_set_rep: - if(this->m_has_backrefs == 0) - static_cast(state)->leading = true; - BOOST_FALLTHROUGH; - default: - return; - } - }while(state); -} - - -} // namespace re_detail - -} // namespace cutl_details_boost - -#ifdef BOOST_MSVC -# pragma warning(pop) -#endif - -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable: 4103) -#endif -#ifdef BOOST_HAS_ABI_HEADERS -# include BOOST_ABI_SUFFIX -#endif -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif - -#endif - diff --git a/cutl/details/boost/regex/v4/basic_regex_parser.hpp b/cutl/details/boost/regex/v4/basic_regex_parser.hpp deleted file mode 100644 index 9182da2..0000000 --- a/cutl/details/boost/regex/v4/basic_regex_parser.hpp +++ /dev/null @@ -1,2874 +0,0 @@ -/* - * - * Copyright (c) 2004 - * John Maddock - * - * Use, modification and distribution are subject to the - * Boost Software License, Version 1.0. (See accompanying file - * LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt) - * - */ - - /* - * LOCATION: see http://www.boost.org for most recent version. - * FILE basic_regex_parser.cpp - * VERSION see - * DESCRIPTION: Declares template class basic_regex_parser. - */ - -#ifndef BOOST_REGEX_V4_BASIC_REGEX_PARSER_HPP -#define BOOST_REGEX_V4_BASIC_REGEX_PARSER_HPP - -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable: 4103) -#endif -#ifdef BOOST_HAS_ABI_HEADERS -# include BOOST_ABI_PREFIX -#endif -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif - -namespace cutl_details_boost{ -namespace re_detail{ - -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable:4244 4800) -#endif - -template -class basic_regex_parser : public basic_regex_creator -{ -public: - basic_regex_parser(regex_data* data); - void parse(const charT* p1, const charT* p2, unsigned flags); - void fail(regex_constants::error_type error_code, std::ptrdiff_t position); - void fail(regex_constants::error_type error_code, std::ptrdiff_t position, std::string message, std::ptrdiff_t start_pos); - void fail(regex_constants::error_type error_code, std::ptrdiff_t position, const std::string& message) - { - fail(error_code, position, message, position); - } - - bool parse_all(); - bool parse_basic(); - bool parse_extended(); - bool parse_literal(); - bool parse_open_paren(); - bool parse_basic_escape(); - bool parse_extended_escape(); - bool parse_match_any(); - bool parse_repeat(std::size_t low = 0, std::size_t high = (std::numeric_limits::max)()); - bool parse_repeat_range(bool isbasic); - bool parse_alt(); - bool parse_set(); - bool parse_backref(); - void parse_set_literal(basic_char_set& char_set); - bool parse_inner_set(basic_char_set& char_set); - bool parse_QE(); - bool parse_perl_extension(); - bool add_emacs_code(bool negate); - bool unwind_alts(std::ptrdiff_t last_paren_start); - digraph get_next_set_literal(basic_char_set& char_set); - charT unescape_character(); - regex_constants::syntax_option_type parse_options(); - -private: - typedef bool (basic_regex_parser::*parser_proc_type)(); - typedef typename traits::string_type string_type; - typedef typename traits::char_class_type char_class_type; - parser_proc_type m_parser_proc; // the main parser to use - const charT* m_base; // the start of the string being parsed - const charT* m_end; // the end of the string being parsed - const charT* m_position; // our current parser position - unsigned m_mark_count; // how many sub-expressions we have - int m_mark_reset; // used to indicate that we're inside a (?|...) block. - unsigned m_max_mark; // largest mark count seen inside a (?|...) block. - std::ptrdiff_t m_paren_start; // where the last seen ')' began (where repeats are inserted). - std::ptrdiff_t m_alt_insert_point; // where to insert the next alternative - bool m_has_case_change; // true if somewhere in the current block the case has changed -#if defined(BOOST_MSVC) && defined(_M_IX86) - // This is an ugly warning suppression workaround (for warnings *inside* std::vector - // that can not otherwise be suppressed)... - BOOST_STATIC_ASSERT(sizeof(long) >= sizeof(void*)); - std::vector m_alt_jumps; // list of alternative in the current scope. -#else - std::vector m_alt_jumps; // list of alternative in the current scope. -#endif - - basic_regex_parser& operator=(const basic_regex_parser&); - basic_regex_parser(const basic_regex_parser&); -}; - -template -basic_regex_parser::basic_regex_parser(regex_data* data) - : basic_regex_creator(data), m_mark_count(0), m_mark_reset(-1), m_max_mark(0), m_paren_start(0), m_alt_insert_point(0), m_has_case_change(false) -{ -} - -template -void basic_regex_parser::parse(const charT* p1, const charT* p2, unsigned l_flags) -{ - // pass l_flags on to base class: - this->init(l_flags); - // set up pointers: - m_position = m_base = p1; - m_end = p2; - // empty strings are errors: - if((p1 == p2) && - ( - ((l_flags & regbase::main_option_type) != regbase::perl_syntax_group) - || (l_flags & regbase::no_empty_expressions) - ) - ) - { - fail(regex_constants::error_empty, 0); - return; - } - // select which parser to use: - switch(l_flags & regbase::main_option_type) - { - case regbase::perl_syntax_group: - { - m_parser_proc = &basic_regex_parser::parse_extended; - // - // Add a leading paren with index zero to give recursions a target: - // - re_brace* br = static_cast(this->append_state(syntax_element_startmark, sizeof(re_brace))); - br->index = 0; - br->icase = this->flags() & regbase::icase; - break; - } - case regbase::basic_syntax_group: - m_parser_proc = &basic_regex_parser::parse_basic; - break; - case regbase::literal: - m_parser_proc = &basic_regex_parser::parse_literal; - break; - default: - // Ooops, someone has managed to set more than one of the main option flags, - // so this must be an error: - fail(regex_constants::error_unknown, 0, "An invalid combination of regular expression syntax flags was used."); - return; - } - - // parse all our characters: - bool result = parse_all(); - // - // Unwind our alternatives: - // - unwind_alts(-1); - // reset l_flags as a global scope (?imsx) may have altered them: - this->flags(l_flags); - // if we haven't gobbled up all the characters then we must - // have had an unexpected ')' : - if(!result) - { - fail(regex_constants::error_paren, ::cutl_details_boost::re_detail::distance(m_base, m_position), "Found a closing ) with no corresponding openening parenthesis."); - return; - } - // if an error has been set then give up now: - if(this->m_pdata->m_status) - return; - // fill in our sub-expression count: - this->m_pdata->m_mark_count = 1 + m_mark_count; - this->finalize(p1, p2); -} - -template -void basic_regex_parser::fail(regex_constants::error_type error_code, std::ptrdiff_t position) -{ - // get the error message: - std::string message = this->m_pdata->m_ptraits->error_string(error_code); - fail(error_code, position, message); -} - -template -void basic_regex_parser::fail(regex_constants::error_type error_code, std::ptrdiff_t position, std::string message, std::ptrdiff_t start_pos) -{ - if(0 == this->m_pdata->m_status) // update the error code if not already set - this->m_pdata->m_status = error_code; - m_position = m_end; // don't bother parsing anything else - -#ifndef BOOST_NO_TEMPLATED_ITERATOR_CONSTRUCTORS - // - // Augment error message with the regular expression text: - // - if(start_pos == position) - start_pos = (std::max)(static_cast(0), position - static_cast(10)); - std::ptrdiff_t end_pos = (std::min)(position + static_cast(10), static_cast(m_end - m_base)); - if(error_code != regex_constants::error_empty) - { - if((start_pos != 0) || (end_pos != (m_end - m_base))) - message += " The error occurred while parsing the regular expression fragment: '"; - else - message += " The error occurred while parsing the regular expression: '"; - if(start_pos != end_pos) - { - message += std::string(m_base + start_pos, m_base + position); - message += ">>>HERE>>>"; - message += std::string(m_base + position, m_base + end_pos); - } - message += "'."; - } -#endif - -#ifndef BOOST_NO_EXCEPTIONS - if(0 == (this->flags() & regex_constants::no_except)) - { - cutl_details_boost::regex_error e(message, error_code, position); - e.raise(); - } -#else - (void)position; // suppress warnings. -#endif -} - -template -bool basic_regex_parser::parse_all() -{ - bool result = true; - while(result && (m_position != m_end)) - { - result = (this->*m_parser_proc)(); - } - return result; -} - -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable:4702) -#endif -template -bool basic_regex_parser::parse_basic() -{ - switch(this->m_traits.syntax_type(*m_position)) - { - case regex_constants::syntax_escape: - return parse_basic_escape(); - case regex_constants::syntax_dot: - return parse_match_any(); - case regex_constants::syntax_caret: - ++m_position; - this->append_state(syntax_element_start_line); - break; - case regex_constants::syntax_dollar: - ++m_position; - this->append_state(syntax_element_end_line); - break; - case regex_constants::syntax_star: - if(!(this->m_last_state) || (this->m_last_state->type == syntax_element_start_line)) - return parse_literal(); - else - { - ++m_position; - return parse_repeat(); - } - case regex_constants::syntax_plus: - if(!(this->m_last_state) || (this->m_last_state->type == syntax_element_start_line) || !(this->flags() & regbase::emacs_ex)) - return parse_literal(); - else - { - ++m_position; - return parse_repeat(1); - } - case regex_constants::syntax_question: - if(!(this->m_last_state) || (this->m_last_state->type == syntax_element_start_line) || !(this->flags() & regbase::emacs_ex)) - return parse_literal(); - else - { - ++m_position; - return parse_repeat(0, 1); - } - case regex_constants::syntax_open_set: - return parse_set(); - case regex_constants::syntax_newline: - if(this->flags() & regbase::newline_alt) - return parse_alt(); - else - return parse_literal(); - default: - return parse_literal(); - } - return true; -} - -template -bool basic_regex_parser::parse_extended() -{ - bool result = true; - switch(this->m_traits.syntax_type(*m_position)) - { - case regex_constants::syntax_open_mark: - return parse_open_paren(); - case regex_constants::syntax_close_mark: - return false; - case regex_constants::syntax_escape: - return parse_extended_escape(); - case regex_constants::syntax_dot: - return parse_match_any(); - case regex_constants::syntax_caret: - ++m_position; - this->append_state( - (this->flags() & regex_constants::no_mod_m ? syntax_element_buffer_start : syntax_element_start_line)); - break; - case regex_constants::syntax_dollar: - ++m_position; - this->append_state( - (this->flags() & regex_constants::no_mod_m ? syntax_element_buffer_end : syntax_element_end_line)); - break; - case regex_constants::syntax_star: - if(m_position == this->m_base) - { - fail(regex_constants::error_badrepeat, 0, "The repeat operator \"*\" cannot start a regular expression."); - return false; - } - ++m_position; - return parse_repeat(); - case regex_constants::syntax_question: - if(m_position == this->m_base) - { - fail(regex_constants::error_badrepeat, 0, "The repeat operator \"?\" cannot start a regular expression."); - return false; - } - ++m_position; - return parse_repeat(0,1); - case regex_constants::syntax_plus: - if(m_position == this->m_base) - { - fail(regex_constants::error_badrepeat, 0, "The repeat operator \"+\" cannot start a regular expression."); - return false; - } - ++m_position; - return parse_repeat(1); - case regex_constants::syntax_open_brace: - ++m_position; - return parse_repeat_range(false); - case regex_constants::syntax_close_brace: - fail(regex_constants::error_brace, this->m_position - this->m_base, "Found a closing repetition operator } with no corresponding {."); - return false; - case regex_constants::syntax_or: - return parse_alt(); - case regex_constants::syntax_open_set: - return parse_set(); - case regex_constants::syntax_newline: - if(this->flags() & regbase::newline_alt) - return parse_alt(); - else - return parse_literal(); - case regex_constants::syntax_hash: - // - // If we have a mod_x flag set, then skip until - // we get to a newline character: - // - if((this->flags() - & (regbase::no_perl_ex|regbase::mod_x)) - == regbase::mod_x) - { - while((m_position != m_end) && !is_separator(*m_position++)){} - return true; - } - BOOST_FALLTHROUGH; - default: - result = parse_literal(); - break; - } - return result; -} -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif - -template -bool basic_regex_parser::parse_literal() -{ - // append this as a literal provided it's not a space character - // or the perl option regbase::mod_x is not set: - if( - ((this->flags() - & (regbase::main_option_type|regbase::mod_x|regbase::no_perl_ex)) - != regbase::mod_x) - || !this->m_traits.isctype(*m_position, this->m_mask_space)) - this->append_literal(*m_position); - ++m_position; - return true; -} - -template -bool basic_regex_parser::parse_open_paren() -{ - // - // skip the '(' and error check: - // - if(++m_position == m_end) - { - fail(regex_constants::error_paren, m_position - m_base); - return false; - } - // - // begin by checking for a perl-style (?...) extension: - // - if( - ((this->flags() & (regbase::main_option_type | regbase::no_perl_ex)) == 0) - || ((this->flags() & (regbase::main_option_type | regbase::emacs_ex)) == (regbase::basic_syntax_group|regbase::emacs_ex)) - ) - { - if(this->m_traits.syntax_type(*m_position) == regex_constants::syntax_question) - return parse_perl_extension(); - } - // - // update our mark count, and append the required state: - // - unsigned markid = 0; - if(0 == (this->flags() & regbase::nosubs)) - { - markid = ++m_mark_count; -#ifndef BOOST_NO_STD_DISTANCE - if(this->flags() & regbase::save_subexpression_location) - this->m_pdata->m_subs.push_back(std::pair(std::distance(m_base, m_position) - 1, 0)); -#else - if(this->flags() & regbase::save_subexpression_location) - this->m_pdata->m_subs.push_back(std::pair((m_position - m_base) - 1, 0)); -#endif - } - re_brace* pb = static_cast(this->append_state(syntax_element_startmark, sizeof(re_brace))); - pb->index = markid; - pb->icase = this->flags() & regbase::icase; - std::ptrdiff_t last_paren_start = this->getoffset(pb); - // back up insertion point for alternations, and set new point: - std::ptrdiff_t last_alt_point = m_alt_insert_point; - this->m_pdata->m_data.align(); - m_alt_insert_point = this->m_pdata->m_data.size(); - // - // back up the current flags in case we have a nested (?imsx) group: - // - regex_constants::syntax_option_type opts = this->flags(); - bool old_case_change = m_has_case_change; - m_has_case_change = false; // no changes to this scope as yet... - // - // Back up branch reset data in case we have a nested (?|...) - // - int mark_reset = m_mark_reset; - m_mark_reset = -1; - // - // now recursively add more states, this will terminate when we get to a - // matching ')' : - // - parse_all(); - // - // Unwind pushed alternatives: - // - if(0 == unwind_alts(last_paren_start)) - return false; - // - // restore flags: - // - if(m_has_case_change) - { - // the case has changed in one or more of the alternatives - // within the scoped (...) block: we have to add a state - // to reset the case sensitivity: - static_cast( - this->append_state(syntax_element_toggle_case, sizeof(re_case)) - )->icase = opts & regbase::icase; - } - this->flags(opts); - m_has_case_change = old_case_change; - // - // restore branch reset: - // - m_mark_reset = mark_reset; - // - // we either have a ')' or we have run out of characters prematurely: - // - if(m_position == m_end) - { - this->fail(regex_constants::error_paren, ::cutl_details_boost::re_detail::distance(m_base, m_end)); - return false; - } - BOOST_ASSERT(this->m_traits.syntax_type(*m_position) == regex_constants::syntax_close_mark); -#ifndef BOOST_NO_STD_DISTANCE - if(markid && (this->flags() & regbase::save_subexpression_location)) - this->m_pdata->m_subs.at(markid - 1).second = std::distance(m_base, m_position); -#else - if(markid && (this->flags() & regbase::save_subexpression_location)) - this->m_pdata->m_subs.at(markid - 1).second = (m_position - m_base); -#endif - ++m_position; - // - // append closing parenthesis state: - // - pb = static_cast(this->append_state(syntax_element_endmark, sizeof(re_brace))); - pb->index = markid; - pb->icase = this->flags() & regbase::icase; - this->m_paren_start = last_paren_start; - // - // restore the alternate insertion point: - // - this->m_alt_insert_point = last_alt_point; - // - // allow backrefs to this mark: - // - if((markid > 0) && (markid < sizeof(unsigned) * CHAR_BIT)) - this->m_backrefs |= 1u << (markid - 1); - - return true; -} - -template -bool basic_regex_parser::parse_basic_escape() -{ - ++m_position; - bool result = true; - switch(this->m_traits.escape_syntax_type(*m_position)) - { - case regex_constants::syntax_open_mark: - return parse_open_paren(); - case regex_constants::syntax_close_mark: - return false; - case regex_constants::syntax_plus: - if(this->flags() & regex_constants::bk_plus_qm) - { - ++m_position; - return parse_repeat(1); - } - else - return parse_literal(); - case regex_constants::syntax_question: - if(this->flags() & regex_constants::bk_plus_qm) - { - ++m_position; - return parse_repeat(0, 1); - } - else - return parse_literal(); - case regex_constants::syntax_open_brace: - if(this->flags() & regbase::no_intervals) - return parse_literal(); - ++m_position; - return parse_repeat_range(true); - case regex_constants::syntax_close_brace: - if(this->flags() & regbase::no_intervals) - return parse_literal(); - fail(regex_constants::error_brace, this->m_position - this->m_base, "Found a closing repetition operator } with no corresponding {."); - return false; - case regex_constants::syntax_or: - if(this->flags() & regbase::bk_vbar) - return parse_alt(); - else - result = parse_literal(); - break; - case regex_constants::syntax_digit: - return parse_backref(); - case regex_constants::escape_type_start_buffer: - if(this->flags() & regbase::emacs_ex) - { - ++m_position; - this->append_state(syntax_element_buffer_start); - } - else - result = parse_literal(); - break; - case regex_constants::escape_type_end_buffer: - if(this->flags() & regbase::emacs_ex) - { - ++m_position; - this->append_state(syntax_element_buffer_end); - } - else - result = parse_literal(); - break; - case regex_constants::escape_type_word_assert: - if(this->flags() & regbase::emacs_ex) - { - ++m_position; - this->append_state(syntax_element_word_boundary); - } - else - result = parse_literal(); - break; - case regex_constants::escape_type_not_word_assert: - if(this->flags() & regbase::emacs_ex) - { - ++m_position; - this->append_state(syntax_element_within_word); - } - else - result = parse_literal(); - break; - case regex_constants::escape_type_left_word: - if(this->flags() & regbase::emacs_ex) - { - ++m_position; - this->append_state(syntax_element_word_start); - } - else - result = parse_literal(); - break; - case regex_constants::escape_type_right_word: - if(this->flags() & regbase::emacs_ex) - { - ++m_position; - this->append_state(syntax_element_word_end); - } - else - result = parse_literal(); - break; - default: - if(this->flags() & regbase::emacs_ex) - { - bool negate = true; - switch(*m_position) - { - case 'w': - negate = false; - BOOST_FALLTHROUGH; - case 'W': - { - basic_char_set char_set; - if(negate) - char_set.negate(); - char_set.add_class(this->m_word_mask); - if(0 == this->append_set(char_set)) - { - fail(regex_constants::error_ctype, m_position - m_base); - return false; - } - ++m_position; - return true; - } - case 's': - negate = false; - BOOST_FALLTHROUGH; - case 'S': - return add_emacs_code(negate); - case 'c': - case 'C': - // not supported yet: - fail(regex_constants::error_escape, m_position - m_base, "The \\c and \\C escape sequences are not supported by POSIX basic regular expressions: try the Perl syntax instead."); - return false; - default: - break; - } - } - result = parse_literal(); - break; - } - return result; -} - -template -bool basic_regex_parser::parse_extended_escape() -{ - ++m_position; - if(m_position == m_end) - { - fail(regex_constants::error_escape, m_position - m_base, "Incomplete escape sequence found."); - return false; - } - bool negate = false; // in case this is a character class escape: \w \d etc - switch(this->m_traits.escape_syntax_type(*m_position)) - { - case regex_constants::escape_type_not_class: - negate = true; - BOOST_FALLTHROUGH; - case regex_constants::escape_type_class: - { -escape_type_class_jump: - typedef typename traits::char_class_type m_type; - m_type m = this->m_traits.lookup_classname(m_position, m_position+1); - if(m != 0) - { - basic_char_set char_set; - if(negate) - char_set.negate(); - char_set.add_class(m); - if(0 == this->append_set(char_set)) - { - fail(regex_constants::error_ctype, m_position - m_base); - return false; - } - ++m_position; - return true; - } - // - // not a class, just a regular unknown escape: - // - this->append_literal(unescape_character()); - break; - } - case regex_constants::syntax_digit: - return parse_backref(); - case regex_constants::escape_type_left_word: - ++m_position; - this->append_state(syntax_element_word_start); - break; - case regex_constants::escape_type_right_word: - ++m_position; - this->append_state(syntax_element_word_end); - break; - case regex_constants::escape_type_start_buffer: - ++m_position; - this->append_state(syntax_element_buffer_start); - break; - case regex_constants::escape_type_end_buffer: - ++m_position; - this->append_state(syntax_element_buffer_end); - break; - case regex_constants::escape_type_word_assert: - ++m_position; - this->append_state(syntax_element_word_boundary); - break; - case regex_constants::escape_type_not_word_assert: - ++m_position; - this->append_state(syntax_element_within_word); - break; - case regex_constants::escape_type_Z: - ++m_position; - this->append_state(syntax_element_soft_buffer_end); - break; - case regex_constants::escape_type_Q: - return parse_QE(); - case regex_constants::escape_type_C: - return parse_match_any(); - case regex_constants::escape_type_X: - ++m_position; - this->append_state(syntax_element_combining); - break; - case regex_constants::escape_type_G: - ++m_position; - this->append_state(syntax_element_restart_continue); - break; - case regex_constants::escape_type_not_property: - negate = true; - BOOST_FALLTHROUGH; - case regex_constants::escape_type_property: - { - ++m_position; - char_class_type m; - if(m_position == m_end) - { - fail(regex_constants::error_escape, m_position - m_base, "Incomplete property escape found."); - return false; - } - // maybe have \p{ddd} - if(this->m_traits.syntax_type(*m_position) == regex_constants::syntax_open_brace) - { - const charT* base = m_position; - // skip forward until we find enclosing brace: - while((m_position != m_end) && (this->m_traits.syntax_type(*m_position) != regex_constants::syntax_close_brace)) - ++m_position; - if(m_position == m_end) - { - fail(regex_constants::error_escape, m_position - m_base, "Closing } missing from property escape sequence."); - return false; - } - m = this->m_traits.lookup_classname(++base, m_position++); - } - else - { - m = this->m_traits.lookup_classname(m_position, m_position+1); - ++m_position; - } - if(m != 0) - { - basic_char_set char_set; - if(negate) - char_set.negate(); - char_set.add_class(m); - if(0 == this->append_set(char_set)) - { - fail(regex_constants::error_ctype, m_position - m_base); - return false; - } - return true; - } - fail(regex_constants::error_ctype, m_position - m_base, "Escape sequence was neither a valid property nor a valid character class name."); - return false; - } - case regex_constants::escape_type_reset_start_mark: - if(0 == (this->flags() & (regbase::main_option_type | regbase::no_perl_ex))) - { - re_brace* pb = static_cast(this->append_state(syntax_element_startmark, sizeof(re_brace))); - pb->index = -5; - pb->icase = this->flags() & regbase::icase; - this->m_pdata->m_data.align(); - ++m_position; - return true; - } - goto escape_type_class_jump; - case regex_constants::escape_type_line_ending: - if(0 == (this->flags() & (regbase::main_option_type | regbase::no_perl_ex))) - { - const charT* e = get_escape_R_string(); - const charT* old_position = m_position; - const charT* old_end = m_end; - const charT* old_base = m_base; - m_position = e; - m_base = e; - m_end = e + traits::length(e); - bool r = parse_all(); - m_position = ++old_position; - m_end = old_end; - m_base = old_base; - return r; - } - goto escape_type_class_jump; - case regex_constants::escape_type_extended_backref: - if(0 == (this->flags() & (regbase::main_option_type | regbase::no_perl_ex))) - { - bool have_brace = false; - bool negative = false; - static const char* incomplete_message = "Incomplete \\g escape found."; - if(++m_position == m_end) - { - fail(regex_constants::error_escape, m_position - m_base, incomplete_message); - return false; - } - // maybe have \g{ddd} - regex_constants::syntax_type syn = this->m_traits.syntax_type(*m_position); - regex_constants::syntax_type syn_end = 0; - if((syn == regex_constants::syntax_open_brace) - || (syn == regex_constants::escape_type_left_word) - || (syn == regex_constants::escape_type_end_buffer)) - { - if(++m_position == m_end) - { - fail(regex_constants::error_escape, m_position - m_base, incomplete_message); - return false; - } - have_brace = true; - switch(syn) - { - case regex_constants::syntax_open_brace: - syn_end = regex_constants::syntax_close_brace; - break; - case regex_constants::escape_type_left_word: - syn_end = regex_constants::escape_type_right_word; - break; - default: - syn_end = regex_constants::escape_type_end_buffer; - break; - } - } - negative = (*m_position == static_cast('-')); - if((negative) && (++m_position == m_end)) - { - fail(regex_constants::error_escape, m_position - m_base, incomplete_message); - return false; - } - const charT* pc = m_position; - int i = this->m_traits.toi(pc, m_end, 10); - if((i < 0) && syn_end) - { - // Check for a named capture, get the leftmost one if there is more than one: - const charT* base = m_position; - while((m_position != m_end) && (this->m_traits.syntax_type(*m_position) != syn_end)) - { - ++m_position; - } - i = hash_value_from_capture_name(base, m_position); - pc = m_position; - } - if(negative) - i = 1 + m_mark_count - i; - if(((i > 0) && (this->m_backrefs & (1u << (i-1)))) || ((i > 10000) && (this->m_pdata->get_id(i) > 0) && (this->m_backrefs & (1u << (this->m_pdata->get_id(i)-1))))) - { - m_position = pc; - re_brace* pb = static_cast(this->append_state(syntax_element_backref, sizeof(re_brace))); - pb->index = i; - pb->icase = this->flags() & regbase::icase; - } - else - { - fail(regex_constants::error_backref, m_position - m_base); - return false; - } - m_position = pc; - if(have_brace) - { - if((m_position == m_end) || (this->m_traits.syntax_type(*m_position) != syn_end)) - { - fail(regex_constants::error_escape, m_position - m_base, incomplete_message); - return false; - } - ++m_position; - } - return true; - } - goto escape_type_class_jump; - case regex_constants::escape_type_control_v: - if(0 == (this->flags() & (regbase::main_option_type | regbase::no_perl_ex))) - goto escape_type_class_jump; - BOOST_FALLTHROUGH; - default: - this->append_literal(unescape_character()); - break; - } - return true; -} - -template -bool basic_regex_parser::parse_match_any() -{ - // - // we have a '.' that can match any character: - // - ++m_position; - static_cast( - this->append_state(syntax_element_wild, sizeof(re_dot)) - )->mask = static_cast(this->flags() & regbase::no_mod_s - ? re_detail::force_not_newline - : this->flags() & regbase::mod_s ? - re_detail::force_newline : re_detail::dont_care); - return true; -} - -template -bool basic_regex_parser::parse_repeat(std::size_t low, std::size_t high) -{ - bool greedy = true; - bool pocessive = false; - std::size_t insert_point; - // - // when we get to here we may have a non-greedy ? mark still to come: - // - if((m_position != m_end) - && ( - (0 == (this->flags() & (regbase::main_option_type | regbase::no_perl_ex))) - || ((regbase::basic_syntax_group|regbase::emacs_ex) == (this->flags() & (regbase::main_option_type | regbase::emacs_ex))) - ) - ) - { - // OK we have a perl or emacs regex, check for a '?': - if(this->m_traits.syntax_type(*m_position) == regex_constants::syntax_question) - { - greedy = false; - ++m_position; - } - // for perl regexes only check for pocessive ++ repeats. - if((m_position != m_end) - && (0 == (this->flags() & regbase::main_option_type)) - && (this->m_traits.syntax_type(*m_position) == regex_constants::syntax_plus)) - { - pocessive = true; - ++m_position; - } - } - if(0 == this->m_last_state) - { - fail(regex_constants::error_badrepeat, ::cutl_details_boost::re_detail::distance(m_base, m_position), "Nothing to repeat."); - return false; - } - if(this->m_last_state->type == syntax_element_endmark) - { - // insert a repeat before the '(' matching the last ')': - insert_point = this->m_paren_start; - } - else if((this->m_last_state->type == syntax_element_literal) && (static_cast(this->m_last_state)->length > 1)) - { - // the last state was a literal with more than one character, split it in two: - re_literal* lit = static_cast(this->m_last_state); - charT c = (static_cast(static_cast(lit+1)))[lit->length - 1]; - --(lit->length); - // now append new state: - lit = static_cast(this->append_state(syntax_element_literal, sizeof(re_literal) + sizeof(charT))); - lit->length = 1; - (static_cast(static_cast(lit+1)))[0] = c; - insert_point = this->getoffset(this->m_last_state); - } - else - { - // repeat the last state whatever it was, need to add some error checking here: - switch(this->m_last_state->type) - { - case syntax_element_start_line: - case syntax_element_end_line: - case syntax_element_word_boundary: - case syntax_element_within_word: - case syntax_element_word_start: - case syntax_element_word_end: - case syntax_element_buffer_start: - case syntax_element_buffer_end: - case syntax_element_alt: - case syntax_element_soft_buffer_end: - case syntax_element_restart_continue: - case syntax_element_jump: - case syntax_element_startmark: - case syntax_element_backstep: - // can't legally repeat any of the above: - fail(regex_constants::error_badrepeat, m_position - m_base); - return false; - default: - // do nothing... - break; - } - insert_point = this->getoffset(this->m_last_state); - } - // - // OK we now know what to repeat, so insert the repeat around it: - // - re_repeat* rep = static_cast(this->insert_state(insert_point, syntax_element_rep, re_repeater_size)); - rep->min = low; - rep->max = high; - rep->greedy = greedy; - rep->leading = false; - // store our repeater position for later: - std::ptrdiff_t rep_off = this->getoffset(rep); - // and append a back jump to the repeat: - re_jump* jmp = static_cast(this->append_state(syntax_element_jump, sizeof(re_jump))); - jmp->alt.i = rep_off - this->getoffset(jmp); - this->m_pdata->m_data.align(); - // now fill in the alt jump for the repeat: - rep = static_cast(this->getaddress(rep_off)); - rep->alt.i = this->m_pdata->m_data.size() - rep_off; - // - // If the repeat is pocessive then bracket the repeat with a (?>...) - // independent sub-expression construct: - // - if(pocessive) - { - if(m_position != m_end) - { - // - // Check for illegal following quantifier, we have to do this here, because - // the extra states we insert below circumvents our usual error checking :-( - // - switch(this->m_traits.syntax_type(*m_position)) - { - case regex_constants::syntax_star: - case regex_constants::syntax_plus: - case regex_constants::syntax_question: - case regex_constants::syntax_open_brace: - fail(regex_constants::error_badrepeat, m_position - m_base); - return false; - } - } - re_brace* pb = static_cast(this->insert_state(insert_point, syntax_element_startmark, sizeof(re_brace))); - pb->index = -3; - pb->icase = this->flags() & regbase::icase; - jmp = static_cast(this->insert_state(insert_point + sizeof(re_brace), syntax_element_jump, sizeof(re_jump))); - this->m_pdata->m_data.align(); - jmp->alt.i = this->m_pdata->m_data.size() - this->getoffset(jmp); - pb = static_cast(this->append_state(syntax_element_endmark, sizeof(re_brace))); - pb->index = -3; - pb->icase = this->flags() & regbase::icase; - } - return true; -} - -template -bool basic_regex_parser::parse_repeat_range(bool isbasic) -{ - static const char* incomplete_message = "Missing } in quantified repetition."; - // - // parse a repeat-range: - // - std::size_t min, max; - int v; - // skip whitespace: - while((m_position != m_end) && this->m_traits.isctype(*m_position, this->m_mask_space)) - ++m_position; - if(this->m_position == this->m_end) - { - if(this->flags() & (regbase::main_option_type | regbase::no_perl_ex)) - { - fail(regex_constants::error_brace, this->m_position - this->m_base, incomplete_message); - return false; - } - // Treat the opening '{' as a literal character, rewind to start of error: - --m_position; - while(this->m_traits.syntax_type(*m_position) != regex_constants::syntax_open_brace) --m_position; - return parse_literal(); - } - // get min: - v = this->m_traits.toi(m_position, m_end, 10); - // skip whitespace: - if(v < 0) - { - if(this->flags() & (regbase::main_option_type | regbase::no_perl_ex)) - { - fail(regex_constants::error_brace, this->m_position - this->m_base, incomplete_message); - return false; - } - // Treat the opening '{' as a literal character, rewind to start of error: - --m_position; - while(this->m_traits.syntax_type(*m_position) != regex_constants::syntax_open_brace) --m_position; - return parse_literal(); - } - while((m_position != m_end) && this->m_traits.isctype(*m_position, this->m_mask_space)) - ++m_position; - if(this->m_position == this->m_end) - { - if(this->flags() & (regbase::main_option_type | regbase::no_perl_ex)) - { - fail(regex_constants::error_brace, this->m_position - this->m_base, incomplete_message); - return false; - } - // Treat the opening '{' as a literal character, rewind to start of error: - --m_position; - while(this->m_traits.syntax_type(*m_position) != regex_constants::syntax_open_brace) --m_position; - return parse_literal(); - } - min = v; - // see if we have a comma: - if(this->m_traits.syntax_type(*m_position) == regex_constants::syntax_comma) - { - // move on and error check: - ++m_position; - // skip whitespace: - while((m_position != m_end) && this->m_traits.isctype(*m_position, this->m_mask_space)) - ++m_position; - if(this->m_position == this->m_end) - { - if(this->flags() & (regbase::main_option_type | regbase::no_perl_ex)) - { - fail(regex_constants::error_brace, this->m_position - this->m_base, incomplete_message); - return false; - } - // Treat the opening '{' as a literal character, rewind to start of error: - --m_position; - while(this->m_traits.syntax_type(*m_position) != regex_constants::syntax_open_brace) --m_position; - return parse_literal(); - } - // get the value if any: - v = this->m_traits.toi(m_position, m_end, 10); - max = (v >= 0) ? (std::size_t)v : (std::numeric_limits::max)(); - } - else - { - // no comma, max = min: - max = min; - } - // skip whitespace: - while((m_position != m_end) && this->m_traits.isctype(*m_position, this->m_mask_space)) - ++m_position; - // OK now check trailing }: - if(this->m_position == this->m_end) - { - if(this->flags() & (regbase::main_option_type | regbase::no_perl_ex)) - { - fail(regex_constants::error_brace, this->m_position - this->m_base, incomplete_message); - return false; - } - // Treat the opening '{' as a literal character, rewind to start of error: - --m_position; - while(this->m_traits.syntax_type(*m_position) != regex_constants::syntax_open_brace) --m_position; - return parse_literal(); - } - if(isbasic) - { - if(this->m_traits.syntax_type(*m_position) == regex_constants::syntax_escape) - { - ++m_position; - if(this->m_position == this->m_end) - { - fail(regex_constants::error_brace, this->m_position - this->m_base, incomplete_message); - return false; - } - } - else - { - fail(regex_constants::error_brace, this->m_position - this->m_base, incomplete_message); - return false; - } - } - if(this->m_traits.syntax_type(*m_position) == regex_constants::syntax_close_brace) - ++m_position; - else - { - // Treat the opening '{' as a literal character, rewind to start of error: - --m_position; - while(this->m_traits.syntax_type(*m_position) != regex_constants::syntax_open_brace) --m_position; - return parse_literal(); - } - // - // finally go and add the repeat, unless error: - // - if(min > max) - { - // Backtrack to error location: - m_position -= 2; - while(this->m_traits.isctype(*m_position, this->m_word_mask)) --m_position; - ++m_position; - fail(regex_constants::error_badbrace, m_position - m_base); - return false; - } - return parse_repeat(min, max); -} - -template -bool basic_regex_parser::parse_alt() -{ - // - // error check: if there have been no previous states, - // or if the last state was a '(' then error: - // - if( - ((this->m_last_state == 0) || (this->m_last_state->type == syntax_element_startmark)) - && - !( - ((this->flags() & regbase::main_option_type) == regbase::perl_syntax_group) - && - ((this->flags() & regbase::no_empty_expressions) == 0) - ) - ) - { - fail(regex_constants::error_empty, this->m_position - this->m_base, "A regular expression can start with the alternation operator |."); - return false; - } - // - // Reset mark count if required: - // - if(m_max_mark < m_mark_count) - m_max_mark = m_mark_count; - if(m_mark_reset >= 0) - m_mark_count = m_mark_reset; - - ++m_position; - // - // we need to append a trailing jump: - // - re_syntax_base* pj = this->append_state(re_detail::syntax_element_jump, sizeof(re_jump)); - std::ptrdiff_t jump_offset = this->getoffset(pj); - // - // now insert the alternative: - // - re_alt* palt = static_cast(this->insert_state(this->m_alt_insert_point, syntax_element_alt, re_alt_size)); - jump_offset += re_alt_size; - this->m_pdata->m_data.align(); - palt->alt.i = this->m_pdata->m_data.size() - this->getoffset(palt); - // - // update m_alt_insert_point so that the next alternate gets - // inserted at the start of the second of the two we've just created: - // - this->m_alt_insert_point = this->m_pdata->m_data.size(); - // - // the start of this alternative must have a case changes state - // if the current block has messed around with case changes: - // - if(m_has_case_change) - { - static_cast( - this->append_state(syntax_element_toggle_case, sizeof(re_case)) - )->icase = this->m_icase; - } - // - // push the alternative onto our stack, a recursive - // implementation here is easier to understand (and faster - // as it happens), but causes all kinds of stack overflow problems - // on programs with small stacks (COM+). - // - m_alt_jumps.push_back(jump_offset); - return true; -} - -template -bool basic_regex_parser::parse_set() -{ - static const char* incomplete_message = "Character set declaration starting with [ terminated prematurely - either no ] was found or the set had no content."; - ++m_position; - if(m_position == m_end) - { - fail(regex_constants::error_brack, m_position - m_base, incomplete_message); - return false; - } - basic_char_set char_set; - - const charT* base = m_position; // where the '[' was - const charT* item_base = m_position; // where the '[' or '^' was - - while(m_position != m_end) - { - switch(this->m_traits.syntax_type(*m_position)) - { - case regex_constants::syntax_caret: - if(m_position == base) - { - char_set.negate(); - ++m_position; - item_base = m_position; - } - else - parse_set_literal(char_set); - break; - case regex_constants::syntax_close_set: - if(m_position == item_base) - { - parse_set_literal(char_set); - break; - } - else - { - ++m_position; - if(0 == this->append_set(char_set)) - { - fail(regex_constants::error_ctype, m_position - m_base); - return false; - } - } - return true; - case regex_constants::syntax_open_set: - if(parse_inner_set(char_set)) - break; - return true; - case regex_constants::syntax_escape: - { - // - // look ahead and see if this is a character class shortcut - // \d \w \s etc... - // - ++m_position; - if(this->m_traits.escape_syntax_type(*m_position) - == regex_constants::escape_type_class) - { - char_class_type m = this->m_traits.lookup_classname(m_position, m_position+1); - if(m != 0) - { - char_set.add_class(m); - ++m_position; - break; - } - } - else if(this->m_traits.escape_syntax_type(*m_position) - == regex_constants::escape_type_not_class) - { - // negated character class: - char_class_type m = this->m_traits.lookup_classname(m_position, m_position+1); - if(m != 0) - { - char_set.add_negated_class(m); - ++m_position; - break; - } - } - // not a character class, just a regular escape: - --m_position; - parse_set_literal(char_set); - break; - } - default: - parse_set_literal(char_set); - break; - } - } - return m_position != m_end; -} - -template -bool basic_regex_parser::parse_inner_set(basic_char_set& char_set) -{ - static const char* incomplete_message = "Character class declaration starting with [ terminated prematurely - either no ] was found or the set had no content."; - // - // we have either a character class [:name:] - // a collating element [.name.] - // or an equivalence class [=name=] - // - if(m_end == ++m_position) - { - fail(regex_constants::error_brack, m_position - m_base, incomplete_message); - return false; - } - switch(this->m_traits.syntax_type(*m_position)) - { - case regex_constants::syntax_dot: - // - // a collating element is treated as a literal: - // - --m_position; - parse_set_literal(char_set); - return true; - case regex_constants::syntax_colon: - { - // check that character classes are actually enabled: - if((this->flags() & (regbase::main_option_type | regbase::no_char_classes)) - == (regbase::basic_syntax_group | regbase::no_char_classes)) - { - --m_position; - parse_set_literal(char_set); - return true; - } - // skip the ':' - if(m_end == ++m_position) - { - fail(regex_constants::error_brack, m_position - m_base, incomplete_message); - return false; - } - const charT* name_first = m_position; - // skip at least one character, then find the matching ':]' - if(m_end == ++m_position) - { - fail(regex_constants::error_brack, m_position - m_base, incomplete_message); - return false; - } - while((m_position != m_end) - && (this->m_traits.syntax_type(*m_position) != regex_constants::syntax_colon)) - ++m_position; - const charT* name_last = m_position; - if(m_end == m_position) - { - fail(regex_constants::error_brack, m_position - m_base, incomplete_message); - return false; - } - if((m_end == ++m_position) - || (this->m_traits.syntax_type(*m_position) != regex_constants::syntax_close_set)) - { - fail(regex_constants::error_brack, m_position - m_base, incomplete_message); - return false; - } - // - // check for negated class: - // - bool negated = false; - if(this->m_traits.syntax_type(*name_first) == regex_constants::syntax_caret) - { - ++name_first; - negated = true; - } - typedef typename traits::char_class_type m_type; - m_type m = this->m_traits.lookup_classname(name_first, name_last); - if(m == 0) - { - if(char_set.empty() && (name_last - name_first == 1)) - { - // maybe a special case: - ++m_position; - if( (m_position != m_end) - && (this->m_traits.syntax_type(*m_position) - == regex_constants::syntax_close_set)) - { - if(this->m_traits.escape_syntax_type(*name_first) - == regex_constants::escape_type_left_word) - { - ++m_position; - this->append_state(syntax_element_word_start); - return false; - } - if(this->m_traits.escape_syntax_type(*name_first) - == regex_constants::escape_type_right_word) - { - ++m_position; - this->append_state(syntax_element_word_end); - return false; - } - } - } - fail(regex_constants::error_ctype, name_first - m_base); - return false; - } - if(negated == false) - char_set.add_class(m); - else - char_set.add_negated_class(m); - ++m_position; - break; - } - case regex_constants::syntax_equal: - { - // skip the '=' - if(m_end == ++m_position) - { - fail(regex_constants::error_brack, m_position - m_base, incomplete_message); - return false; - } - const charT* name_first = m_position; - // skip at least one character, then find the matching '=]' - if(m_end == ++m_position) - { - fail(regex_constants::error_brack, m_position - m_base, incomplete_message); - return false; - } - while((m_position != m_end) - && (this->m_traits.syntax_type(*m_position) != regex_constants::syntax_equal)) - ++m_position; - const charT* name_last = m_position; - if(m_end == m_position) - { - fail(regex_constants::error_brack, m_position - m_base, incomplete_message); - return false; - } - if((m_end == ++m_position) - || (this->m_traits.syntax_type(*m_position) != regex_constants::syntax_close_set)) - { - fail(regex_constants::error_brack, m_position - m_base, incomplete_message); - return false; - } - string_type m = this->m_traits.lookup_collatename(name_first, name_last); - if((0 == m.size()) || (m.size() > 2)) - { - fail(regex_constants::error_collate, name_first - m_base); - return false; - } - digraph d; - d.first = m[0]; - if(m.size() > 1) - d.second = m[1]; - else - d.second = 0; - char_set.add_equivalent(d); - ++m_position; - break; - } - default: - --m_position; - parse_set_literal(char_set); - break; - } - return true; -} - -template -void basic_regex_parser::parse_set_literal(basic_char_set& char_set) -{ - digraph start_range(get_next_set_literal(char_set)); - if(m_end == m_position) - { - fail(regex_constants::error_brack, m_position - m_base); - return; - } - if(this->m_traits.syntax_type(*m_position) == regex_constants::syntax_dash) - { - // we have a range: - if(m_end == ++m_position) - { - fail(regex_constants::error_brack, m_position - m_base); - return; - } - if(this->m_traits.syntax_type(*m_position) != regex_constants::syntax_close_set) - { - digraph end_range = get_next_set_literal(char_set); - char_set.add_range(start_range, end_range); - if(this->m_traits.syntax_type(*m_position) == regex_constants::syntax_dash) - { - if(m_end == ++m_position) - { - fail(regex_constants::error_brack, m_position - m_base); - return; - } - if(this->m_traits.syntax_type(*m_position) == regex_constants::syntax_close_set) - { - // trailing - : - --m_position; - return; - } - fail(regex_constants::error_range, m_position - m_base); - return; - } - return; - } - --m_position; - } - char_set.add_single(start_range); -} - -template -digraph basic_regex_parser::get_next_set_literal(basic_char_set& char_set) -{ - digraph result; - switch(this->m_traits.syntax_type(*m_position)) - { - case regex_constants::syntax_dash: - if(!char_set.empty()) - { - // see if we are at the end of the set: - if((++m_position == m_end) || (this->m_traits.syntax_type(*m_position) != regex_constants::syntax_close_set)) - { - fail(regex_constants::error_range, m_position - m_base); - return result; - } - --m_position; - } - result.first = *m_position++; - return result; - case regex_constants::syntax_escape: - // check to see if escapes are supported first: - if(this->flags() & regex_constants::no_escape_in_lists) - { - result = *m_position++; - break; - } - ++m_position; - result = unescape_character(); - break; - case regex_constants::syntax_open_set: - { - if(m_end == ++m_position) - { - fail(regex_constants::error_collate, m_position - m_base); - return result; - } - if(this->m_traits.syntax_type(*m_position) != regex_constants::syntax_dot) - { - --m_position; - result.first = *m_position; - ++m_position; - return result; - } - if(m_end == ++m_position) - { - fail(regex_constants::error_collate, m_position - m_base); - return result; - } - const charT* name_first = m_position; - // skip at least one character, then find the matching ':]' - if(m_end == ++m_position) - { - fail(regex_constants::error_collate, name_first - m_base); - return result; - } - while((m_position != m_end) - && (this->m_traits.syntax_type(*m_position) != regex_constants::syntax_dot)) - ++m_position; - const charT* name_last = m_position; - if(m_end == m_position) - { - fail(regex_constants::error_collate, name_first - m_base); - return result; - } - if((m_end == ++m_position) - || (this->m_traits.syntax_type(*m_position) != regex_constants::syntax_close_set)) - { - fail(regex_constants::error_collate, name_first - m_base); - return result; - } - ++m_position; - string_type s = this->m_traits.lookup_collatename(name_first, name_last); - if(s.empty() || (s.size() > 2)) - { - fail(regex_constants::error_collate, name_first - m_base); - return result; - } - result.first = s[0]; - if(s.size() > 1) - result.second = s[1]; - else - result.second = 0; - return result; - } - default: - result = *m_position++; - } - return result; -} - -// -// does a value fit in the specified charT type? -// -template -bool valid_value(charT, int v, const mpl::true_&) -{ - return (v >> (sizeof(charT) * CHAR_BIT)) == 0; -} -template -bool valid_value(charT, int, const mpl::false_&) -{ - return true; // v will alsways fit in a charT -} -template -bool valid_value(charT c, int v) -{ - return valid_value(c, v, mpl::bool_<(sizeof(charT) < sizeof(int))>()); -} - -template -charT basic_regex_parser::unescape_character() -{ -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable:4127) -#endif - charT result(0); - if(m_position == m_end) - { - fail(regex_constants::error_escape, m_position - m_base, "Escape sequence terminated prematurely."); - return false; - } - switch(this->m_traits.escape_syntax_type(*m_position)) - { - case regex_constants::escape_type_control_a: - result = charT('\a'); - break; - case regex_constants::escape_type_e: - result = charT(27); - break; - case regex_constants::escape_type_control_f: - result = charT('\f'); - break; - case regex_constants::escape_type_control_n: - result = charT('\n'); - break; - case regex_constants::escape_type_control_r: - result = charT('\r'); - break; - case regex_constants::escape_type_control_t: - result = charT('\t'); - break; - case regex_constants::escape_type_control_v: - result = charT('\v'); - break; - case regex_constants::escape_type_word_assert: - result = charT('\b'); - break; - case regex_constants::escape_type_ascii_control: - ++m_position; - if(m_position == m_end) - { - // Rewind to start of escape: - --m_position; - while(this->m_traits.syntax_type(*m_position) != regex_constants::syntax_escape) --m_position; - fail(regex_constants::error_escape, m_position - m_base, "ASCII escape sequence terminated prematurely."); - return result; - } - result = static_cast(*m_position % 32); - break; - case regex_constants::escape_type_hex: - ++m_position; - if(m_position == m_end) - { - // Rewind to start of escape: - --m_position; - while(this->m_traits.syntax_type(*m_position) != regex_constants::syntax_escape) --m_position; - fail(regex_constants::error_escape, m_position - m_base, "Hexadecimal escape sequence terminated prematurely."); - return result; - } - // maybe have \x{ddd} - if(this->m_traits.syntax_type(*m_position) == regex_constants::syntax_open_brace) - { - ++m_position; - if(m_position == m_end) - { - // Rewind to start of escape: - --m_position; - while(this->m_traits.syntax_type(*m_position) != regex_constants::syntax_escape) --m_position; - fail(regex_constants::error_escape, m_position - m_base, "Missing } in hexadecimal escape sequence."); - return result; - } - int i = this->m_traits.toi(m_position, m_end, 16); - if((m_position == m_end) - || (i < 0) - || ((std::numeric_limits::is_specialized) && (i > (int)(std::numeric_limits::max)())) - || (this->m_traits.syntax_type(*m_position) != regex_constants::syntax_close_brace)) - { - // Rewind to start of escape: - --m_position; - while(this->m_traits.syntax_type(*m_position) != regex_constants::syntax_escape) --m_position; - fail(regex_constants::error_badbrace, m_position - m_base, "Hexadecimal escape sequence was invalid."); - return result; - } - ++m_position; - result = charT(i); - } - else - { - std::ptrdiff_t len = (std::min)(static_cast(2), static_cast(m_end - m_position)); - int i = this->m_traits.toi(m_position, m_position + len, 16); - if((i < 0) - || !valid_value(charT(0), i)) - { - // Rewind to start of escape: - --m_position; - while(this->m_traits.syntax_type(*m_position) != regex_constants::syntax_escape) --m_position; - fail(regex_constants::error_escape, m_position - m_base, "Escape sequence did not encode a valid character."); - return result; - } - result = charT(i); - } - return result; - case regex_constants::syntax_digit: - { - // an octal escape sequence, the first character must be a zero - // followed by up to 3 octal digits: - std::ptrdiff_t len = (std::min)(::cutl_details_boost::re_detail::distance(m_position, m_end), static_cast(4)); - const charT* bp = m_position; - int val = this->m_traits.toi(bp, bp + 1, 8); - if(val != 0) - { - // Rewind to start of escape: - --m_position; - while(this->m_traits.syntax_type(*m_position) != regex_constants::syntax_escape) --m_position; - // Oops not an octal escape after all: - fail(regex_constants::error_escape, m_position - m_base, "Invalid octal escape sequence."); - return result; - } - val = this->m_traits.toi(m_position, m_position + len, 8); - if(val < 0) - { - // Rewind to start of escape: - --m_position; - while(this->m_traits.syntax_type(*m_position) != regex_constants::syntax_escape) --m_position; - fail(regex_constants::error_escape, m_position - m_base, "Octal escape sequence is invalid."); - return result; - } - return static_cast(val); - } - case regex_constants::escape_type_named_char: - { - ++m_position; - if(m_position == m_end) - { - // Rewind to start of escape: - --m_position; - while(this->m_traits.syntax_type(*m_position) != regex_constants::syntax_escape) --m_position; - fail(regex_constants::error_escape, m_position - m_base); - return false; - } - // maybe have \N{name} - if(this->m_traits.syntax_type(*m_position) == regex_constants::syntax_open_brace) - { - const charT* base = m_position; - // skip forward until we find enclosing brace: - while((m_position != m_end) && (this->m_traits.syntax_type(*m_position) != regex_constants::syntax_close_brace)) - ++m_position; - if(m_position == m_end) - { - // Rewind to start of escape: - --m_position; - while(this->m_traits.syntax_type(*m_position) != regex_constants::syntax_escape) --m_position; - fail(regex_constants::error_escape, m_position - m_base); - return false; - } - string_type s = this->m_traits.lookup_collatename(++base, m_position++); - if(s.empty()) - { - // Rewind to start of escape: - --m_position; - while(this->m_traits.syntax_type(*m_position) != regex_constants::syntax_escape) --m_position; - fail(regex_constants::error_collate, m_position - m_base); - return false; - } - if(s.size() == 1) - { - return s[0]; - } - } - // fall through is a failure: - // Rewind to start of escape: - --m_position; - while(this->m_traits.syntax_type(*m_position) != regex_constants::syntax_escape) --m_position; - fail(regex_constants::error_escape, m_position - m_base); - return false; - } - default: - result = *m_position; - break; - } - ++m_position; - return result; -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif -} - -template -bool basic_regex_parser::parse_backref() -{ - BOOST_ASSERT(m_position != m_end); - const charT* pc = m_position; - int i = this->m_traits.toi(pc, pc + 1, 10); - if((i == 0) || (((this->flags() & regbase::main_option_type) == regbase::perl_syntax_group) && (this->flags() & regbase::no_bk_refs))) - { - // not a backref at all but an octal escape sequence: - charT c = unescape_character(); - this->append_literal(c); - } - else if((i > 0) && (this->m_backrefs & (1u << (i-1)))) - { - m_position = pc; - re_brace* pb = static_cast(this->append_state(syntax_element_backref, sizeof(re_brace))); - pb->index = i; - pb->icase = this->flags() & regbase::icase; - } - else - { - // Rewind to start of escape: - --m_position; - while(this->m_traits.syntax_type(*m_position) != regex_constants::syntax_escape) --m_position; - fail(regex_constants::error_backref, m_position - m_base); - return false; - } - return true; -} - -template -bool basic_regex_parser::parse_QE() -{ -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable:4127) -#endif - // - // parse a \Q...\E sequence: - // - ++m_position; // skip the Q - const charT* start = m_position; - const charT* end; - do - { - while((m_position != m_end) - && (this->m_traits.syntax_type(*m_position) != regex_constants::syntax_escape)) - ++m_position; - if(m_position == m_end) - { - // a \Q...\E sequence may terminate with the end of the expression: - end = m_position; - break; - } - if(++m_position == m_end) // skip the escape - { - fail(regex_constants::error_escape, m_position - m_base, "Unterminated \\Q...\\E sequence."); - return false; - } - // check to see if it's a \E: - if(this->m_traits.escape_syntax_type(*m_position) == regex_constants::escape_type_E) - { - ++m_position; - end = m_position - 2; - break; - } - // otherwise go round again: - }while(true); - // - // now add all the character between the two escapes as literals: - // - while(start != end) - { - this->append_literal(*start); - ++start; - } - return true; -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif -} - -template -bool basic_regex_parser::parse_perl_extension() -{ - if(++m_position == m_end) - { - // Rewind to start of (? sequence: - --m_position; - while(this->m_traits.syntax_type(*m_position) != regex_constants::syntax_open_mark) --m_position; - fail(regex_constants::error_perl_extension, m_position - m_base); - return false; - } - // - // treat comments as a special case, as these - // are the only ones that don't start with a leading - // startmark state: - // - if(this->m_traits.syntax_type(*m_position) == regex_constants::syntax_hash) - { - while((m_position != m_end) - && (this->m_traits.syntax_type(*m_position++) != regex_constants::syntax_close_mark)) - {} - return true; - } - // - // backup some state, and prepare the way: - // - int markid = 0; - std::ptrdiff_t jump_offset = 0; - re_brace* pb = static_cast(this->append_state(syntax_element_startmark, sizeof(re_brace))); - pb->icase = this->flags() & regbase::icase; - std::ptrdiff_t last_paren_start = this->getoffset(pb); - // back up insertion point for alternations, and set new point: - std::ptrdiff_t last_alt_point = m_alt_insert_point; - this->m_pdata->m_data.align(); - m_alt_insert_point = this->m_pdata->m_data.size(); - std::ptrdiff_t expected_alt_point = m_alt_insert_point; - bool restore_flags = true; - regex_constants::syntax_option_type old_flags = this->flags(); - bool old_case_change = m_has_case_change; - m_has_case_change = false; - charT name_delim; - int mark_reset = m_mark_reset; - int max_mark = m_max_mark; - m_mark_reset = -1; - m_max_mark = m_mark_count; - int v; - // - // select the actual extension used: - // - switch(this->m_traits.syntax_type(*m_position)) - { - case regex_constants::syntax_or: - m_mark_reset = m_mark_count; - BOOST_FALLTHROUGH; - case regex_constants::syntax_colon: - // - // a non-capturing mark: - // - pb->index = markid = 0; - ++m_position; - break; - case regex_constants::syntax_digit: - { - // - // a recursive subexpression: - // - v = this->m_traits.toi(m_position, m_end, 10); - if((v < 0) || (this->m_traits.syntax_type(*m_position) != regex_constants::syntax_close_mark)) - { - // Rewind to start of (? sequence: - --m_position; - while(this->m_traits.syntax_type(*m_position) != regex_constants::syntax_open_mark) --m_position; - fail(regex_constants::error_perl_extension, m_position - m_base, "The recursive sub-expression refers to an invalid marking group, or is unterminated."); - return false; - } -insert_recursion: - pb->index = markid = 0; - re_recurse* pr = static_cast(this->append_state(syntax_element_recurse, sizeof(re_recurse))); - pr->alt.i = v; - pr->state_id = 0; - static_cast( - this->append_state(syntax_element_toggle_case, sizeof(re_case)) - )->icase = this->flags() & regbase::icase; - break; - } - case regex_constants::syntax_plus: - // - // A forward-relative recursive subexpression: - // - ++m_position; - v = this->m_traits.toi(m_position, m_end, 10); - if((v <= 0) || (this->m_traits.syntax_type(*m_position) != regex_constants::syntax_close_mark)) - { - // Rewind to start of (? sequence: - --m_position; - while(this->m_traits.syntax_type(*m_position) != regex_constants::syntax_open_mark) --m_position; - fail(regex_constants::error_perl_extension, m_position - m_base, "An invalid or unterminated recursive sub-expression."); - return false; - } - v += m_mark_count; - goto insert_recursion; - case regex_constants::syntax_dash: - // - // Possibly a backward-relative recursive subexpression: - // - ++m_position; - v = this->m_traits.toi(m_position, m_end, 10); - if(v <= 0) - { - --m_position; - // Oops not a relative recursion at all, but a (?-imsx) group: - goto option_group_jump; - } - v = m_mark_count + 1 - v; - if(v <= 0) - { - // Rewind to start of (? sequence: - --m_position; - while(this->m_traits.syntax_type(*m_position) != regex_constants::syntax_open_mark) --m_position; - fail(regex_constants::error_perl_extension, m_position - m_base, "An invalid or unterminated recursive sub-expression."); - return false; - } - goto insert_recursion; - case regex_constants::syntax_equal: - pb->index = markid = -1; - ++m_position; - jump_offset = this->getoffset(this->append_state(syntax_element_jump, sizeof(re_jump))); - this->m_pdata->m_data.align(); - m_alt_insert_point = this->m_pdata->m_data.size(); - break; - case regex_constants::syntax_not: - pb->index = markid = -2; - ++m_position; - jump_offset = this->getoffset(this->append_state(syntax_element_jump, sizeof(re_jump))); - this->m_pdata->m_data.align(); - m_alt_insert_point = this->m_pdata->m_data.size(); - break; - case regex_constants::escape_type_left_word: - { - // a lookbehind assertion: - if(++m_position == m_end) - { - // Rewind to start of (? sequence: - --m_position; - while(this->m_traits.syntax_type(*m_position) != regex_constants::syntax_open_mark) --m_position; - fail(regex_constants::error_perl_extension, m_position - m_base); - return false; - } - regex_constants::syntax_type t = this->m_traits.syntax_type(*m_position); - if(t == regex_constants::syntax_not) - pb->index = markid = -2; - else if(t == regex_constants::syntax_equal) - pb->index = markid = -1; - else - { - // Probably a named capture which also starts (?< : - name_delim = '>'; - --m_position; - goto named_capture_jump; - } - ++m_position; - jump_offset = this->getoffset(this->append_state(syntax_element_jump, sizeof(re_jump))); - this->append_state(syntax_element_backstep, sizeof(re_brace)); - this->m_pdata->m_data.align(); - m_alt_insert_point = this->m_pdata->m_data.size(); - break; - } - case regex_constants::escape_type_right_word: - // - // an independent sub-expression: - // - pb->index = markid = -3; - ++m_position; - jump_offset = this->getoffset(this->append_state(syntax_element_jump, sizeof(re_jump))); - this->m_pdata->m_data.align(); - m_alt_insert_point = this->m_pdata->m_data.size(); - break; - case regex_constants::syntax_open_mark: - { - // a conditional expression: - pb->index = markid = -4; - if(++m_position == m_end) - { - // Rewind to start of (? sequence: - --m_position; - while(this->m_traits.syntax_type(*m_position) != regex_constants::syntax_open_mark) --m_position; - fail(regex_constants::error_perl_extension, m_position - m_base); - return false; - } - v = this->m_traits.toi(m_position, m_end, 10); - if(m_position == m_end) - { - // Rewind to start of (? sequence: - --m_position; - while(this->m_traits.syntax_type(*m_position) != regex_constants::syntax_open_mark) --m_position; - fail(regex_constants::error_perl_extension, m_position - m_base); - return false; - } - if(*m_position == charT('R')) - { - if(++m_position == m_end) - { - // Rewind to start of (? sequence: - --m_position; - while(this->m_traits.syntax_type(*m_position) != regex_constants::syntax_open_mark) --m_position; - fail(regex_constants::error_perl_extension, m_position - m_base); - return false; - } - if(*m_position == charT('&')) - { - const charT* base = ++m_position; - while((m_position != m_end) && (this->m_traits.syntax_type(*m_position) != regex_constants::syntax_close_mark)) - ++m_position; - if(m_position == m_end) - { - // Rewind to start of (? sequence: - --m_position; - while(this->m_traits.syntax_type(*m_position) != regex_constants::syntax_open_mark) --m_position; - fail(regex_constants::error_perl_extension, m_position - m_base); - return false; - } - v = -static_cast(hash_value_from_capture_name(base, m_position)); - } - else - { - v = -this->m_traits.toi(m_position, m_end, 10); - } - re_brace* br = static_cast(this->append_state(syntax_element_assert_backref, sizeof(re_brace))); - br->index = v < 0 ? (v - 1) : 0; - if(this->m_traits.syntax_type(*m_position) != regex_constants::syntax_close_mark) - { - // Rewind to start of (? sequence: - --m_position; - while(this->m_traits.syntax_type(*m_position) != regex_constants::syntax_open_mark) --m_position; - fail(regex_constants::error_perl_extension, m_position - m_base); - return false; - } - if(++m_position == m_end) - { - // Rewind to start of (? sequence: - --m_position; - while(this->m_traits.syntax_type(*m_position) != regex_constants::syntax_open_mark) --m_position; - fail(regex_constants::error_perl_extension, m_position - m_base); - return false; - } - } - else if((*m_position == charT('\'')) || (*m_position == charT('<'))) - { - const charT* base = ++m_position; - while((m_position != m_end) && (*m_position != charT('>')) && (*m_position != charT('\''))) - ++m_position; - if(m_position == m_end) - { - // Rewind to start of (? sequence: - --m_position; - while(this->m_traits.syntax_type(*m_position) != regex_constants::syntax_open_mark) --m_position; - fail(regex_constants::error_perl_extension, m_position - m_base); - return false; - } - v = static_cast(hash_value_from_capture_name(base, m_position)); - re_brace* br = static_cast(this->append_state(syntax_element_assert_backref, sizeof(re_brace))); - br->index = v; - if(((*m_position != charT('>')) && (*m_position != charT('\''))) || (++m_position == m_end)) - { - // Rewind to start of (? sequence: - --m_position; - while(this->m_traits.syntax_type(*m_position) != regex_constants::syntax_open_mark) --m_position; - fail(regex_constants::error_perl_extension, m_position - m_base, "Unterminated named capture."); - return false; - } - if(this->m_traits.syntax_type(*m_position) != regex_constants::syntax_close_mark) - { - // Rewind to start of (? sequence: - --m_position; - while(this->m_traits.syntax_type(*m_position) != regex_constants::syntax_open_mark) --m_position; - fail(regex_constants::error_perl_extension, m_position - m_base); - return false; - } - if(++m_position == m_end) - { - // Rewind to start of (? sequence: - --m_position; - while(this->m_traits.syntax_type(*m_position) != regex_constants::syntax_open_mark) --m_position; - fail(regex_constants::error_perl_extension, m_position - m_base); - return false; - } - } - else if(*m_position == charT('D')) - { - const char* def = "DEFINE"; - while(*def && (m_position != m_end) && (*m_position == charT(*def))) - ++m_position, ++def; - if((m_position == m_end) || *def) - { - // Rewind to start of (? sequence: - --m_position; - while(this->m_traits.syntax_type(*m_position) != regex_constants::syntax_open_mark) --m_position; - fail(regex_constants::error_perl_extension, m_position - m_base); - return false; - } - re_brace* br = static_cast(this->append_state(syntax_element_assert_backref, sizeof(re_brace))); - br->index = 9999; // special magic value! - if(this->m_traits.syntax_type(*m_position) != regex_constants::syntax_close_mark) - { - // Rewind to start of (? sequence: - --m_position; - while(this->m_traits.syntax_type(*m_position) != regex_constants::syntax_open_mark) --m_position; - fail(regex_constants::error_perl_extension, m_position - m_base); - return false; - } - if(++m_position == m_end) - { - // Rewind to start of (? sequence: - --m_position; - while(this->m_traits.syntax_type(*m_position) != regex_constants::syntax_open_mark) --m_position; - fail(regex_constants::error_perl_extension, m_position - m_base); - return false; - } - } - else if(v > 0) - { - re_brace* br = static_cast(this->append_state(syntax_element_assert_backref, sizeof(re_brace))); - br->index = v; - if(this->m_traits.syntax_type(*m_position) != regex_constants::syntax_close_mark) - { - // Rewind to start of (? sequence: - --m_position; - while(this->m_traits.syntax_type(*m_position) != regex_constants::syntax_open_mark) --m_position; - fail(regex_constants::error_perl_extension, m_position - m_base); - return false; - } - if(++m_position == m_end) - { - // Rewind to start of (? sequence: - --m_position; - while(this->m_traits.syntax_type(*m_position) != regex_constants::syntax_open_mark) --m_position; - fail(regex_constants::error_perl_extension, m_position - m_base); - return false; - } - } - else - { - // verify that we have a lookahead or lookbehind assert: - if(this->m_traits.syntax_type(*m_position) != regex_constants::syntax_question) - { - // Rewind to start of (? sequence: - --m_position; - while(this->m_traits.syntax_type(*m_position) != regex_constants::syntax_open_mark) --m_position; - fail(regex_constants::error_perl_extension, m_position - m_base); - return false; - } - if(++m_position == m_end) - { - // Rewind to start of (? sequence: - --m_position; - while(this->m_traits.syntax_type(*m_position) != regex_constants::syntax_open_mark) --m_position; - fail(regex_constants::error_perl_extension, m_position - m_base); - return false; - } - if(this->m_traits.syntax_type(*m_position) == regex_constants::escape_type_left_word) - { - if(++m_position == m_end) - { - // Rewind to start of (? sequence: - --m_position; - while(this->m_traits.syntax_type(*m_position) != regex_constants::syntax_open_mark) --m_position; - fail(regex_constants::error_perl_extension, m_position - m_base); - return false; - } - if((this->m_traits.syntax_type(*m_position) != regex_constants::syntax_equal) - && (this->m_traits.syntax_type(*m_position) != regex_constants::syntax_not)) - { - // Rewind to start of (? sequence: - --m_position; - while(this->m_traits.syntax_type(*m_position) != regex_constants::syntax_open_mark) --m_position; - fail(regex_constants::error_perl_extension, m_position - m_base); - return false; - } - m_position -= 3; - } - else - { - if((this->m_traits.syntax_type(*m_position) != regex_constants::syntax_equal) - && (this->m_traits.syntax_type(*m_position) != regex_constants::syntax_not)) - { - // Rewind to start of (? sequence: - --m_position; - while(this->m_traits.syntax_type(*m_position) != regex_constants::syntax_open_mark) --m_position; - fail(regex_constants::error_perl_extension, m_position - m_base); - return false; - } - m_position -= 2; - } - } - break; - } - case regex_constants::syntax_close_mark: - // Rewind to start of (? sequence: - --m_position; - while(this->m_traits.syntax_type(*m_position) != regex_constants::syntax_open_mark) --m_position; - fail(regex_constants::error_perl_extension, m_position - m_base); - return false; - case regex_constants::escape_type_end_buffer: - { - name_delim = *m_position; -named_capture_jump: - markid = 0; - if(0 == (this->flags() & regbase::nosubs)) - { - markid = ++m_mark_count; - #ifndef BOOST_NO_STD_DISTANCE - if(this->flags() & regbase::save_subexpression_location) - this->m_pdata->m_subs.push_back(std::pair(std::distance(m_base, m_position) - 2, 0)); - #else - if(this->flags() & regbase::save_subexpression_location) - this->m_pdata->m_subs.push_back(std::pair((m_position - m_base) - 2, 0)); - #endif - } - pb->index = markid; - const charT* base = ++m_position; - if(m_position == m_end) - { - // Rewind to start of (? sequence: - --m_position; - while(this->m_traits.syntax_type(*m_position) != regex_constants::syntax_open_mark) --m_position; - fail(regex_constants::error_perl_extension, m_position - m_base); - return false; - } - while((m_position != m_end) && (*m_position != name_delim)) - ++m_position; - if(m_position == m_end) - { - // Rewind to start of (? sequence: - --m_position; - while(this->m_traits.syntax_type(*m_position) != regex_constants::syntax_open_mark) --m_position; - fail(regex_constants::error_perl_extension, m_position - m_base); - return false; - } - this->m_pdata->set_name(base, m_position, markid); - ++m_position; - break; - } - default: - if(*m_position == charT('R')) - { - ++m_position; - v = 0; - if(this->m_traits.syntax_type(*m_position) != regex_constants::syntax_close_mark) - { - // Rewind to start of (? sequence: - --m_position; - while(this->m_traits.syntax_type(*m_position) != regex_constants::syntax_open_mark) --m_position; - fail(regex_constants::error_perl_extension, m_position - m_base); - return false; - } - goto insert_recursion; - } - if(*m_position == charT('&')) - { - ++m_position; - const charT* base = m_position; - while((m_position != m_end) && (this->m_traits.syntax_type(*m_position) != regex_constants::syntax_close_mark)) - ++m_position; - if(m_position == m_end) - { - // Rewind to start of (? sequence: - --m_position; - while(this->m_traits.syntax_type(*m_position) != regex_constants::syntax_open_mark) --m_position; - fail(regex_constants::error_perl_extension, m_position - m_base); - return false; - } - v = static_cast(hash_value_from_capture_name(base, m_position)); - goto insert_recursion; - } - if(*m_position == charT('P')) - { - ++m_position; - if(m_position == m_end) - { - // Rewind to start of (? sequence: - --m_position; - while(this->m_traits.syntax_type(*m_position) != regex_constants::syntax_open_mark) --m_position; - fail(regex_constants::error_perl_extension, m_position - m_base); - return false; - } - if(*m_position == charT('>')) - { - ++m_position; - const charT* base = m_position; - while((m_position != m_end) && (this->m_traits.syntax_type(*m_position) != regex_constants::syntax_close_mark)) - ++m_position; - if(m_position == m_end) - { - // Rewind to start of (? sequence: - --m_position; - while(this->m_traits.syntax_type(*m_position) != regex_constants::syntax_open_mark) --m_position; - fail(regex_constants::error_perl_extension, m_position - m_base); - return false; - } - v = static_cast(hash_value_from_capture_name(base, m_position)); - goto insert_recursion; - } - } - // - // lets assume that we have a (?imsx) group and try and parse it: - // -option_group_jump: - regex_constants::syntax_option_type opts = parse_options(); - if(m_position == m_end) - { - // Rewind to start of (? sequence: - --m_position; - while(this->m_traits.syntax_type(*m_position) != regex_constants::syntax_open_mark) --m_position; - fail(regex_constants::error_perl_extension, m_position - m_base); - return false; - } - // make a note of whether we have a case change: - m_has_case_change = ((opts & regbase::icase) != (this->flags() & regbase::icase)); - pb->index = markid = 0; - if(this->m_traits.syntax_type(*m_position) == regex_constants::syntax_close_mark) - { - // update flags and carry on as normal: - this->flags(opts); - restore_flags = false; - old_case_change |= m_has_case_change; // defer end of scope by one ')' - } - else if(this->m_traits.syntax_type(*m_position) == regex_constants::syntax_colon) - { - // update flags and carry on until the matching ')' is found: - this->flags(opts); - ++m_position; - } - else - { - // Rewind to start of (? sequence: - --m_position; - while(this->m_traits.syntax_type(*m_position) != regex_constants::syntax_open_mark) --m_position; - fail(regex_constants::error_perl_extension, m_position - m_base); - return false; - } - - // finally append a case change state if we need it: - if(m_has_case_change) - { - static_cast( - this->append_state(syntax_element_toggle_case, sizeof(re_case)) - )->icase = opts & regbase::icase; - } - - } - // - // now recursively add more states, this will terminate when we get to a - // matching ')' : - // - parse_all(); - // - // Unwind alternatives: - // - if(0 == unwind_alts(last_paren_start)) - { - // Rewind to start of (? sequence: - --m_position; - while(this->m_traits.syntax_type(*m_position) != regex_constants::syntax_open_mark) --m_position; - fail(regex_constants::error_perl_extension, m_position - m_base, "Invalid alternation operators within (?...) block."); - return false; - } - // - // we either have a ')' or we have run out of characters prematurely: - // - if(m_position == m_end) - { - // Rewind to start of (? sequence: - --m_position; - while(this->m_traits.syntax_type(*m_position) != regex_constants::syntax_open_mark) --m_position; - this->fail(regex_constants::error_paren, ::cutl_details_boost::re_detail::distance(m_base, m_end)); - return false; - } - BOOST_ASSERT(this->m_traits.syntax_type(*m_position) == regex_constants::syntax_close_mark); - ++m_position; - // - // restore the flags: - // - if(restore_flags) - { - // append a case change state if we need it: - if(m_has_case_change) - { - static_cast( - this->append_state(syntax_element_toggle_case, sizeof(re_case)) - )->icase = old_flags & regbase::icase; - } - this->flags(old_flags); - } - // - // set up the jump pointer if we have one: - // - if(jump_offset) - { - this->m_pdata->m_data.align(); - re_jump* jmp = static_cast(this->getaddress(jump_offset)); - jmp->alt.i = this->m_pdata->m_data.size() - this->getoffset(jmp); - if((this->m_last_state == jmp) && (markid != -2)) - { - // Oops... we didn't have anything inside the assertion. - // Note we don't get here for negated forward lookahead as (?!) - // does have some uses. - // Rewind to start of (? sequence: - --m_position; - while(this->m_traits.syntax_type(*m_position) != regex_constants::syntax_open_mark) --m_position; - fail(regex_constants::error_perl_extension, m_position - m_base, "Invalid or empty zero width assertion."); - return false; - } - } - // - // verify that if this is conditional expression, that we do have - // an alternative, if not add one: - // - if(markid == -4) - { - re_syntax_base* b = this->getaddress(expected_alt_point); - // Make sure we have exactly one alternative following this state: - if(b->type != syntax_element_alt) - { - re_alt* alt = static_cast(this->insert_state(expected_alt_point, syntax_element_alt, sizeof(re_alt))); - alt->alt.i = this->m_pdata->m_data.size() - this->getoffset(alt); - } - else if(this->getaddress(static_cast(b)->alt.i, b)->type == syntax_element_alt) - { - // Can't have seen more than one alternative: - // Rewind to start of (? sequence: - --m_position; - while(this->m_traits.syntax_type(*m_position) != regex_constants::syntax_open_mark) --m_position; - fail(regex_constants::error_bad_pattern, m_position - m_base, "More than one alternation operator | was encountered inside a conditional expression."); - return false; - } - else - { - // We must *not* have seen an alternative inside a (DEFINE) block: - b = this->getaddress(b->next.i, b); - if((b->type == syntax_element_assert_backref) && (static_cast(b)->index == 9999)) - { - // Rewind to start of (? sequence: - --m_position; - while(this->m_traits.syntax_type(*m_position) != regex_constants::syntax_open_mark) --m_position; - fail(regex_constants::error_bad_pattern, m_position - m_base, "Alternation operators are not allowed inside a DEFINE block."); - return false; - } - } - // check for invalid repetition of next state: - b = this->getaddress(expected_alt_point); - b = this->getaddress(static_cast(b)->next.i, b); - if((b->type != syntax_element_assert_backref) - && (b->type != syntax_element_startmark)) - { - // Rewind to start of (? sequence: - --m_position; - while(this->m_traits.syntax_type(*m_position) != regex_constants::syntax_open_mark) --m_position; - fail(regex_constants::error_badrepeat, m_position - m_base, "A repetition operator cannot be applied to a zero-width assertion."); - return false; - } - } - // - // append closing parenthesis state: - // - pb = static_cast(this->append_state(syntax_element_endmark, sizeof(re_brace))); - pb->index = markid; - pb->icase = this->flags() & regbase::icase; - this->m_paren_start = last_paren_start; - // - // restore the alternate insertion point: - // - this->m_alt_insert_point = last_alt_point; - // - // and the case change data: - // - m_has_case_change = old_case_change; - // - // And the mark_reset data: - // - if(m_max_mark > m_mark_count) - { - m_mark_count = m_max_mark; - } - m_mark_reset = mark_reset; - m_max_mark = max_mark; - - - if(markid > 0) - { -#ifndef BOOST_NO_STD_DISTANCE - if(this->flags() & regbase::save_subexpression_location) - this->m_pdata->m_subs.at(markid - 1).second = std::distance(m_base, m_position) - 1; -#else - if(this->flags() & regbase::save_subexpression_location) - this->m_pdata->m_subs.at(markid - 1).second = (m_position - m_base) - 1; -#endif - // - // allow backrefs to this mark: - // - if((markid > 0) && (markid < (int)(sizeof(unsigned) * CHAR_BIT))) - this->m_backrefs |= 1u << (markid - 1); - } - return true; -} - -template -bool basic_regex_parser::add_emacs_code(bool negate) -{ - // - // parses an emacs style \sx or \Sx construct. - // - if(++m_position == m_end) - { - // Rewind to start of sequence: - --m_position; - while(this->m_traits.syntax_type(*m_position) != regex_constants::syntax_escape) --m_position; - fail(regex_constants::error_escape, m_position - m_base); - return false; - } - basic_char_set char_set; - if(negate) - char_set.negate(); - - static const charT s_punct[5] = { 'p', 'u', 'n', 'c', 't', }; - - switch(*m_position) - { - case 's': - case ' ': - char_set.add_class(this->m_mask_space); - break; - case 'w': - char_set.add_class(this->m_word_mask); - break; - case '_': - char_set.add_single(digraph(charT('$'))); - char_set.add_single(digraph(charT('&'))); - char_set.add_single(digraph(charT('*'))); - char_set.add_single(digraph(charT('+'))); - char_set.add_single(digraph(charT('-'))); - char_set.add_single(digraph(charT('_'))); - char_set.add_single(digraph(charT('<'))); - char_set.add_single(digraph(charT('>'))); - break; - case '.': - char_set.add_class(this->m_traits.lookup_classname(s_punct, s_punct+5)); - break; - case '(': - char_set.add_single(digraph(charT('('))); - char_set.add_single(digraph(charT('['))); - char_set.add_single(digraph(charT('{'))); - break; - case ')': - char_set.add_single(digraph(charT(')'))); - char_set.add_single(digraph(charT(']'))); - char_set.add_single(digraph(charT('}'))); - break; - case '"': - char_set.add_single(digraph(charT('"'))); - char_set.add_single(digraph(charT('\''))); - char_set.add_single(digraph(charT('`'))); - break; - case '\'': - char_set.add_single(digraph(charT('\''))); - char_set.add_single(digraph(charT(','))); - char_set.add_single(digraph(charT('#'))); - break; - case '<': - char_set.add_single(digraph(charT(';'))); - break; - case '>': - char_set.add_single(digraph(charT('\n'))); - char_set.add_single(digraph(charT('\f'))); - break; - default: - fail(regex_constants::error_ctype, m_position - m_base); - return false; - } - if(0 == this->append_set(char_set)) - { - fail(regex_constants::error_ctype, m_position - m_base); - return false; - } - ++m_position; - return true; -} - -template -regex_constants::syntax_option_type basic_regex_parser::parse_options() -{ - // we have a (?imsx-imsx) group, convert it into a set of flags: - regex_constants::syntax_option_type f = this->flags(); - bool breakout = false; - do - { - switch(*m_position) - { - case 's': - f |= regex_constants::mod_s; - f &= ~regex_constants::no_mod_s; - break; - case 'm': - f &= ~regex_constants::no_mod_m; - break; - case 'i': - f |= regex_constants::icase; - break; - case 'x': - f |= regex_constants::mod_x; - break; - default: - breakout = true; - continue; - } - if(++m_position == m_end) - { - // Rewind to start of (? sequence: - --m_position; - while(this->m_traits.syntax_type(*m_position) != regex_constants::syntax_open_mark) --m_position; - fail(regex_constants::error_paren, m_position - m_base); - return false; - } - } - while(!breakout); - - breakout = false; - - if(*m_position == static_cast('-')) - { - if(++m_position == m_end) - { - // Rewind to start of (? sequence: - --m_position; - while(this->m_traits.syntax_type(*m_position) != regex_constants::syntax_open_mark) --m_position; - fail(regex_constants::error_paren, m_position - m_base); - return false; - } - do - { - switch(*m_position) - { - case 's': - f &= ~regex_constants::mod_s; - f |= regex_constants::no_mod_s; - break; - case 'm': - f |= regex_constants::no_mod_m; - break; - case 'i': - f &= ~regex_constants::icase; - break; - case 'x': - f &= ~regex_constants::mod_x; - break; - default: - breakout = true; - continue; - } - if(++m_position == m_end) - { - // Rewind to start of (? sequence: - --m_position; - while(this->m_traits.syntax_type(*m_position) != regex_constants::syntax_open_mark) --m_position; - fail(regex_constants::error_paren, m_position - m_base); - return false; - } - } - while(!breakout); - } - return f; -} - -template -bool basic_regex_parser::unwind_alts(std::ptrdiff_t last_paren_start) -{ - // - // If we didn't actually add any states after the last - // alternative then that's an error: - // - if((this->m_alt_insert_point == static_cast(this->m_pdata->m_data.size())) - && m_alt_jumps.size() && (m_alt_jumps.back() > last_paren_start) - && - !( - ((this->flags() & regbase::main_option_type) == regbase::perl_syntax_group) - && - ((this->flags() & regbase::no_empty_expressions) == 0) - ) - ) - { - fail(regex_constants::error_empty, this->m_position - this->m_base, "Can't terminate a sub-expression with an alternation operator |."); - return false; - } - // - // Fix up our alternatives: - // - while(m_alt_jumps.size() && (m_alt_jumps.back() > last_paren_start)) - { - // - // fix up the jump to point to the end of the states - // that we've just added: - // - std::ptrdiff_t jump_offset = m_alt_jumps.back(); - m_alt_jumps.pop_back(); - this->m_pdata->m_data.align(); - re_jump* jmp = static_cast(this->getaddress(jump_offset)); - BOOST_ASSERT(jmp->type == syntax_element_jump); - jmp->alt.i = this->m_pdata->m_data.size() - jump_offset; - } - return true; -} - -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif - -} // namespace re_detail -} // namespace cutl_details_boost - -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable: 4103) -#endif -#ifdef BOOST_HAS_ABI_HEADERS -# include BOOST_ABI_SUFFIX -#endif -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif - -#endif diff --git a/cutl/details/boost/regex/v4/c_regex_traits.hpp b/cutl/details/boost/regex/v4/c_regex_traits.hpp deleted file mode 100644 index d4f4c30..0000000 --- a/cutl/details/boost/regex/v4/c_regex_traits.hpp +++ /dev/null @@ -1,211 +0,0 @@ -/* - * - * Copyright (c) 2004 - * John Maddock - * - * Use, modification and distribution are subject to the - * Boost Software License, Version 1.0. (See accompanying file - * LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt) - * - */ - - /* - * LOCATION: see http://www.boost.org for most recent version. - * FILE c_regex_traits.hpp - * VERSION see - * DESCRIPTION: Declares regular expression traits class that wraps the global C locale. - */ - -#ifndef BOOST_C_REGEX_TRAITS_HPP_INCLUDED -#define BOOST_C_REGEX_TRAITS_HPP_INCLUDED - -#ifndef BOOST_REGEX_CONFIG_HPP -#include -#endif -#ifndef BOOST_REGEX_WORKAROUND_HPP -#include -#endif - -#include - -#ifdef BOOST_NO_STDC_NAMESPACE -namespace std{ - using ::strlen; using ::tolower; -} -#endif - -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable: 4103) -#endif -#ifdef BOOST_HAS_ABI_HEADERS -# include BOOST_ABI_PREFIX -#endif -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif - -namespace cutl_details_boost{ - -template -struct c_regex_traits; - -template<> -struct BOOST_REGEX_DECL c_regex_traits -{ - c_regex_traits(){} - typedef char char_type; - typedef std::size_t size_type; - typedef std::string string_type; - struct locale_type{}; - typedef cutl_details_boost::uint32_t char_class_type; - - static size_type length(const char_type* p) - { - return (std::strlen)(p); - } - - char translate(char c) const - { - return c; - } - char translate_nocase(char c) const - { - return static_cast((std::tolower)(static_cast(c))); - } - - static string_type BOOST_REGEX_CALL transform(const char* p1, const char* p2); - static string_type BOOST_REGEX_CALL transform_primary(const char* p1, const char* p2); - - static char_class_type BOOST_REGEX_CALL lookup_classname(const char* p1, const char* p2); - static string_type BOOST_REGEX_CALL lookup_collatename(const char* p1, const char* p2); - - static bool BOOST_REGEX_CALL isctype(char, char_class_type); - static int BOOST_REGEX_CALL value(char, int); - - locale_type imbue(locale_type l) - { return l; } - locale_type getloc()const - { return locale_type(); } - -private: - // this type is not copyable: - c_regex_traits(const c_regex_traits&); - c_regex_traits& operator=(const c_regex_traits&); -}; - -#ifndef BOOST_NO_WREGEX -template<> -struct BOOST_REGEX_DECL c_regex_traits -{ - c_regex_traits(){} - typedef wchar_t char_type; - typedef std::size_t size_type; - typedef std::wstring string_type; - struct locale_type{}; - typedef cutl_details_boost::uint32_t char_class_type; - - static size_type length(const char_type* p) - { - return (std::wcslen)(p); - } - - wchar_t translate(wchar_t c) const - { - return c; - } - wchar_t translate_nocase(wchar_t c) const - { - return (std::towlower)(c); - } - - static string_type BOOST_REGEX_CALL transform(const wchar_t* p1, const wchar_t* p2); - static string_type BOOST_REGEX_CALL transform_primary(const wchar_t* p1, const wchar_t* p2); - - static char_class_type BOOST_REGEX_CALL lookup_classname(const wchar_t* p1, const wchar_t* p2); - static string_type BOOST_REGEX_CALL lookup_collatename(const wchar_t* p1, const wchar_t* p2); - - static bool BOOST_REGEX_CALL isctype(wchar_t, char_class_type); - static int BOOST_REGEX_CALL value(wchar_t, int); - - locale_type imbue(locale_type l) - { return l; } - locale_type getloc()const - { return locale_type(); } - -private: - // this type is not copyable: - c_regex_traits(const c_regex_traits&); - c_regex_traits& operator=(const c_regex_traits&); -}; - -#ifdef BOOST_REGEX_HAS_OTHER_WCHAR_T -// -// Provide an unsigned short version as well, so the user can link to this -// no matter whether they build with /Zc:wchar_t or not (MSVC specific). -// -template<> -struct BOOST_REGEX_DECL c_regex_traits -{ - c_regex_traits(){} - typedef unsigned short char_type; - typedef std::size_t size_type; - typedef std::basic_string string_type; - struct locale_type{}; - typedef cutl_details_boost::uint32_t char_class_type; - - static size_type length(const char_type* p) - { - return (std::wcslen)((const wchar_t*)p); - } - - unsigned short translate(unsigned short c) const - { - return c; - } - unsigned short translate_nocase(unsigned short c) const - { - return (std::towlower)((wchar_t)c); - } - - static string_type BOOST_REGEX_CALL transform(const unsigned short* p1, const unsigned short* p2); - static string_type BOOST_REGEX_CALL transform_primary(const unsigned short* p1, const unsigned short* p2); - - static char_class_type BOOST_REGEX_CALL lookup_classname(const unsigned short* p1, const unsigned short* p2); - static string_type BOOST_REGEX_CALL lookup_collatename(const unsigned short* p1, const unsigned short* p2); - - static bool BOOST_REGEX_CALL isctype(unsigned short, char_class_type); - static int BOOST_REGEX_CALL value(unsigned short, int); - - locale_type imbue(locale_type l) - { return l; } - locale_type getloc()const - { return locale_type(); } - -private: - // this type is not copyable: - c_regex_traits(const c_regex_traits&); - c_regex_traits& operator=(const c_regex_traits&); -}; - -#endif - -#endif // BOOST_NO_WREGEX - -} - -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable: 4103) -#endif -#ifdef BOOST_HAS_ABI_HEADERS -# include BOOST_ABI_SUFFIX -#endif -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif - -#endif - - - diff --git a/cutl/details/boost/regex/v4/char_regex_traits.hpp b/cutl/details/boost/regex/v4/char_regex_traits.hpp deleted file mode 100644 index 4bae0ec..0000000 --- a/cutl/details/boost/regex/v4/char_regex_traits.hpp +++ /dev/null @@ -1,81 +0,0 @@ -/* - * - * Copyright (c) 2002 - * John Maddock - * - * Use, modification and distribution are subject to the - * Boost Software License, Version 1.0. (See accompanying file - * LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt) - * - */ - - /* - * LOCATION: see http://www.boost.org for most recent version. - * FILE char_regex_traits.cpp - * VERSION see - * DESCRIPTION: Declares deprecated traits classes char_regex_traits<>. - */ - - -#ifndef BOOST_REGEX_V4_CHAR_REGEX_TRAITS_HPP -#define BOOST_REGEX_V4_CHAR_REGEX_TRAITS_HPP - -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable: 4103) -#endif -#ifdef BOOST_HAS_ABI_HEADERS -# include BOOST_ABI_PREFIX -#endif -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif - -namespace cutl_details_boost{ - -namespace deprecated{ -// -// class char_regex_traits_i -// provides case insensitive traits classes (deprecated): -template -class char_regex_traits_i : public regex_traits {}; - -template<> -class char_regex_traits_i : public regex_traits -{ -public: - typedef char char_type; - typedef unsigned char uchar_type; - typedef unsigned int size_type; - typedef regex_traits base_type; - -}; - -#ifndef BOOST_NO_WREGEX -template<> -class char_regex_traits_i : public regex_traits -{ -public: - typedef wchar_t char_type; - typedef unsigned short uchar_type; - typedef unsigned int size_type; - typedef regex_traits base_type; - -}; -#endif -} // namespace deprecated -} // namespace cutl_details_boost - -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable: 4103) -#endif -#ifdef BOOST_HAS_ABI_HEADERS -# include BOOST_ABI_SUFFIX -#endif -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif - -#endif // include - diff --git a/cutl/details/boost/regex/v4/cpp_regex_traits.hpp b/cutl/details/boost/regex/v4/cpp_regex_traits.hpp deleted file mode 100644 index 654f668..0000000 --- a/cutl/details/boost/regex/v4/cpp_regex_traits.hpp +++ /dev/null @@ -1,1099 +0,0 @@ -/* - * - * Copyright (c) 2004 John Maddock - * Copyright 2011 Garmin Ltd. or its subsidiaries - * - * Use, modification and distribution are subject to the - * Boost Software License, Version 1.0. (See accompanying file - * LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt) - * - */ - - /* - * LOCATION: see http://www.boost.org for most recent version. - * FILE cpp_regex_traits.hpp - * VERSION see - * DESCRIPTION: Declares regular expression traits class cpp_regex_traits. - */ - -#ifndef BOOST_CPP_REGEX_TRAITS_HPP_INCLUDED -#define BOOST_CPP_REGEX_TRAITS_HPP_INCLUDED - -#include -#include - -#ifndef BOOST_NO_STD_LOCALE - -#ifndef BOOST_RE_PAT_EXCEPT_HPP -#include -#endif -#ifndef BOOST_REGEX_TRAITS_DEFAULTS_HPP_INCLUDED -#include -#endif -#ifdef BOOST_HAS_THREADS -#include -#endif -#ifndef BOOST_REGEX_PRIMARY_TRANSFORM -#include -#endif -#ifndef BOOST_REGEX_OBJECT_CACHE_HPP -#include -#endif - -#include -#include -#include - -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable: 4103) -#endif -#ifdef BOOST_HAS_ABI_HEADERS -# include BOOST_ABI_PREFIX -#endif -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif - -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable:4786 4251) -#endif - -namespace cutl_details_boost{ - -// -// forward declaration is needed by some compilers: -// -template -class cpp_regex_traits; - -namespace re_detail{ - -// -// class parser_buf: -// acts as a stream buffer which wraps around a pair of pointers: -// -template > -class parser_buf : public ::std::basic_streambuf -{ - typedef ::std::basic_streambuf base_type; - typedef typename base_type::int_type int_type; - typedef typename base_type::char_type char_type; - typedef typename base_type::pos_type pos_type; - typedef ::std::streamsize streamsize; - typedef typename base_type::off_type off_type; -public: - parser_buf() : base_type() { setbuf(0, 0); } - const charT* getnext() { return this->gptr(); } -protected: - std::basic_streambuf* setbuf(char_type* s, streamsize n); - typename parser_buf::pos_type seekpos(pos_type sp, ::std::ios_base::openmode which); - typename parser_buf::pos_type seekoff(off_type off, ::std::ios_base::seekdir way, ::std::ios_base::openmode which); -private: - parser_buf& operator=(const parser_buf&); - parser_buf(const parser_buf&); -}; - -template -std::basic_streambuf* -parser_buf::setbuf(char_type* s, streamsize n) -{ - this->setg(s, s, s + n); - return this; -} - -template -typename parser_buf::pos_type -parser_buf::seekoff(off_type off, ::std::ios_base::seekdir way, ::std::ios_base::openmode which) -{ - typedef typename cutl_details_boost::int_t::least cast_type; - - if(which & ::std::ios_base::out) - return pos_type(off_type(-1)); - std::ptrdiff_t size = this->egptr() - this->eback(); - std::ptrdiff_t pos = this->gptr() - this->eback(); - charT* g = this->eback(); - switch(static_cast(way)) - { - case ::std::ios_base::beg: - if((off < 0) || (off > size)) - return pos_type(off_type(-1)); - else - this->setg(g, g + off, g + size); - break; - case ::std::ios_base::end: - if((off < 0) || (off > size)) - return pos_type(off_type(-1)); - else - this->setg(g, g + size - off, g + size); - break; - case ::std::ios_base::cur: - { - std::ptrdiff_t newpos = static_cast(pos + off); - if((newpos < 0) || (newpos > size)) - return pos_type(off_type(-1)); - else - this->setg(g, g + newpos, g + size); - break; - } - default: ; - } -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable:4244) -#endif - return static_cast(this->gptr() - this->eback()); -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif -} - -template -typename parser_buf::pos_type -parser_buf::seekpos(pos_type sp, ::std::ios_base::openmode which) -{ - if(which & ::std::ios_base::out) - return pos_type(off_type(-1)); - off_type size = static_cast(this->egptr() - this->eback()); - charT* g = this->eback(); - if(off_type(sp) <= size) - { - this->setg(g, g + off_type(sp), g + size); - } - return pos_type(off_type(-1)); -} - -// -// class cpp_regex_traits_base: -// acts as a container for locale and the facets we are using. -// -template -struct cpp_regex_traits_base -{ - cpp_regex_traits_base(const std::locale& l) - { imbue(l); } - std::locale imbue(const std::locale& l); - - std::locale m_locale; - std::ctype const* m_pctype; -#ifndef BOOST_NO_STD_MESSAGES - std::messages const* m_pmessages; -#endif - std::collate const* m_pcollate; - - bool operator<(const cpp_regex_traits_base& b)const - { - if(m_pctype == b.m_pctype) - { -#ifndef BOOST_NO_STD_MESSAGES - if(m_pmessages == b.m_pmessages) - { - return m_pcollate < b.m_pcollate; - } - return m_pmessages < b.m_pmessages; -#else - return m_pcollate < b.m_pcollate; -#endif - } - return m_pctype < b.m_pctype; - } - bool operator==(const cpp_regex_traits_base& b)const - { - return (m_pctype == b.m_pctype) -#ifndef BOOST_NO_STD_MESSAGES - && (m_pmessages == b.m_pmessages) -#endif - && (m_pcollate == b.m_pcollate); - } -}; - -template -std::locale cpp_regex_traits_base::imbue(const std::locale& l) -{ - std::locale result(m_locale); - m_locale = l; - m_pctype = &BOOST_USE_FACET(std::ctype, l); -#ifndef BOOST_NO_STD_MESSAGES - m_pmessages = BOOST_HAS_FACET(std::messages, l) ? &BOOST_USE_FACET(std::messages, l) : 0; -#endif - m_pcollate = &BOOST_USE_FACET(std::collate, l); - return result; -} - -// -// class cpp_regex_traits_char_layer: -// implements methods that require specialisation for narrow characters: -// -template -class cpp_regex_traits_char_layer : public cpp_regex_traits_base -{ - typedef std::basic_string string_type; - typedef std::map map_type; - typedef typename map_type::const_iterator map_iterator_type; -public: - cpp_regex_traits_char_layer(const std::locale& l) - : cpp_regex_traits_base(l) - { - init(); - } - cpp_regex_traits_char_layer(const cpp_regex_traits_base& b) - : cpp_regex_traits_base(b) - { - init(); - } - void init(); - - regex_constants::syntax_type syntax_type(charT c)const - { - map_iterator_type i = m_char_map.find(c); - return ((i == m_char_map.end()) ? 0 : i->second); - } - regex_constants::escape_syntax_type escape_syntax_type(charT c) const - { - map_iterator_type i = m_char_map.find(c); - if(i == m_char_map.end()) - { - if(this->m_pctype->is(std::ctype_base::lower, c)) return regex_constants::escape_type_class; - if(this->m_pctype->is(std::ctype_base::upper, c)) return regex_constants::escape_type_not_class; - return 0; - } - return i->second; - } - -private: - string_type get_default_message(regex_constants::syntax_type); - // TODO: use a hash table when available! - map_type m_char_map; -}; - -template -void cpp_regex_traits_char_layer::init() -{ - // we need to start by initialising our syntax map so we know which - // character is used for which purpose: -#ifndef BOOST_NO_STD_MESSAGES -#ifndef __IBMCPP__ - typename std::messages::catalog cat = static_cast::catalog>(-1); -#else - typename std::messages::catalog cat = reinterpret_cast::catalog>(-1); -#endif - std::string cat_name(cpp_regex_traits::get_catalog_name()); - if(cat_name.size() && (this->m_pmessages != 0)) - { - cat = this->m_pmessages->open( - cat_name, - this->m_locale); - if((int)cat < 0) - { - std::string m("Unable to open message catalog: "); - std::runtime_error err(m + cat_name); - cutl_details_boost::re_detail::raise_runtime_error(err); - } - } - // - // if we have a valid catalog then load our messages: - // - if((int)cat >= 0) - { -#ifndef BOOST_NO_EXCEPTIONS - try{ -#endif - for(regex_constants::syntax_type i = 1; i < regex_constants::syntax_max; ++i) - { - string_type mss = this->m_pmessages->get(cat, 0, i, get_default_message(i)); - for(typename string_type::size_type j = 0; j < mss.size(); ++j) - { - m_char_map[mss[j]] = i; - } - } - this->m_pmessages->close(cat); -#ifndef BOOST_NO_EXCEPTIONS - } - catch(...) - { - if(this->m_pmessages) - this->m_pmessages->close(cat); - throw; - } -#endif - } - else - { -#endif - for(regex_constants::syntax_type i = 1; i < regex_constants::syntax_max; ++i) - { - const char* ptr = get_default_syntax(i); - while(ptr && *ptr) - { - m_char_map[this->m_pctype->widen(*ptr)] = i; - ++ptr; - } - } -#ifndef BOOST_NO_STD_MESSAGES - } -#endif -} - -template -typename cpp_regex_traits_char_layer::string_type - cpp_regex_traits_char_layer::get_default_message(regex_constants::syntax_type i) -{ - const char* ptr = get_default_syntax(i); - string_type result; - while(ptr && *ptr) - { - result.append(1, this->m_pctype->widen(*ptr)); - ++ptr; - } - return result; -} - -// -// specialised version for narrow characters: -// -template <> -class BOOST_REGEX_DECL cpp_regex_traits_char_layer : public cpp_regex_traits_base -{ - typedef std::string string_type; -public: - cpp_regex_traits_char_layer(const std::locale& l) - : cpp_regex_traits_base(l) - { - init(); - } - cpp_regex_traits_char_layer(const cpp_regex_traits_base& l) - : cpp_regex_traits_base(l) - { - init(); - } - - regex_constants::syntax_type syntax_type(char c)const - { - return m_char_map[static_cast(c)]; - } - regex_constants::escape_syntax_type escape_syntax_type(char c) const - { - return m_char_map[static_cast(c)]; - } - -private: - regex_constants::syntax_type m_char_map[1u << CHAR_BIT]; - void init(); -}; - -#ifdef BOOST_REGEX_BUGGY_CTYPE_FACET -enum -{ - char_class_space=1<<0, - char_class_print=1<<1, - char_class_cntrl=1<<2, - char_class_upper=1<<3, - char_class_lower=1<<4, - char_class_alpha=1<<5, - char_class_digit=1<<6, - char_class_punct=1<<7, - char_class_xdigit=1<<8, - char_class_alnum=char_class_alpha|char_class_digit, - char_class_graph=char_class_alnum|char_class_punct, - char_class_blank=1<<9, - char_class_word=1<<10, - char_class_unicode=1<<11, - char_class_horizontal_space=1<<12, - char_class_vertical_space=1<<13 -}; - -#endif - -// -// class cpp_regex_traits_implementation: -// provides pimpl implementation for cpp_regex_traits. -// -template -class cpp_regex_traits_implementation : public cpp_regex_traits_char_layer -{ -public: - typedef typename cpp_regex_traits::char_class_type char_class_type; - typedef typename std::ctype::mask native_mask_type; -#ifndef BOOST_REGEX_BUGGY_CTYPE_FACET - BOOST_STATIC_CONSTANT(char_class_type, mask_blank = 1u << 24); - BOOST_STATIC_CONSTANT(char_class_type, mask_word = 1u << 25); - BOOST_STATIC_CONSTANT(char_class_type, mask_unicode = 1u << 26); - BOOST_STATIC_CONSTANT(char_class_type, mask_horizontal = 1u << 27); - BOOST_STATIC_CONSTANT(char_class_type, mask_vertical = 1u << 28); -#endif - - typedef std::basic_string string_type; - typedef charT char_type; - //cpp_regex_traits_implementation(); - cpp_regex_traits_implementation(const std::locale& l) - : cpp_regex_traits_char_layer(l) - { - init(); - } - cpp_regex_traits_implementation(const cpp_regex_traits_base& l) - : cpp_regex_traits_char_layer(l) - { - init(); - } - std::string error_string(regex_constants::error_type n) const - { - if(!m_error_strings.empty()) - { - std::map::const_iterator p = m_error_strings.find(n); - return (p == m_error_strings.end()) ? std::string(get_default_error_string(n)) : p->second; - } - return get_default_error_string(n); - } - char_class_type lookup_classname(const charT* p1, const charT* p2) const - { - char_class_type result = lookup_classname_imp(p1, p2); - if(result == 0) - { - string_type temp(p1, p2); - this->m_pctype->tolower(&*temp.begin(), &*temp.begin() + temp.size()); - result = lookup_classname_imp(&*temp.begin(), &*temp.begin() + temp.size()); - } - return result; - } - string_type lookup_collatename(const charT* p1, const charT* p2) const; - string_type transform_primary(const charT* p1, const charT* p2) const; - string_type transform(const charT* p1, const charT* p2) const; -private: - std::map m_error_strings; // error messages indexed by numberic ID - std::map m_custom_class_names; // character class names - std::map m_custom_collate_names; // collating element names - unsigned m_collate_type; // the form of the collation string - charT m_collate_delim; // the collation group delimiter - // - // helpers: - // - char_class_type lookup_classname_imp(const charT* p1, const charT* p2) const; - void init(); -#ifdef BOOST_REGEX_BUGGY_CTYPE_FACET -public: - bool isctype(charT c, char_class_type m)const; -#endif -}; - -#ifndef BOOST_REGEX_BUGGY_CTYPE_FACET -#if !defined(BOOST_NO_INCLASS_MEMBER_INITIALIZATION) - -template -typename cpp_regex_traits_implementation::char_class_type const cpp_regex_traits_implementation::mask_blank; -template -typename cpp_regex_traits_implementation::char_class_type const cpp_regex_traits_implementation::mask_word; -template -typename cpp_regex_traits_implementation::char_class_type const cpp_regex_traits_implementation::mask_unicode; -template -typename cpp_regex_traits_implementation::char_class_type const cpp_regex_traits_implementation::mask_vertical; -template -typename cpp_regex_traits_implementation::char_class_type const cpp_regex_traits_implementation::mask_horizontal; - -#endif -#endif - -template -typename cpp_regex_traits_implementation::string_type - cpp_regex_traits_implementation::transform_primary(const charT* p1, const charT* p2) const -{ - // - // PRECONDITIONS: - // - // A bug in gcc 3.2 (and maybe other versions as well) treats - // p1 as a null terminated string, for efficiency reasons - // we work around this elsewhere, but just assert here that - // we adhere to gcc's (buggy) preconditions... - // - BOOST_ASSERT(*p2 == 0); - - string_type result; - // - // swallowing all exceptions here is a bad idea - // however at least one std lib will always throw - // std::bad_alloc for certain arguments... - // -#ifndef BOOST_NO_EXCEPTIONS - try{ -#endif - // - // What we do here depends upon the format of the sort key returned by - // sort key returned by this->transform: - // - switch(m_collate_type) - { - case sort_C: - case sort_unknown: - // the best we can do is translate to lower case, then get a regular sort key: - { - result.assign(p1, p2); - this->m_pctype->tolower(&*result.begin(), &*result.begin() + result.size()); - result = this->m_pcollate->transform(&*result.begin(), &*result.begin() + result.size()); - break; - } - case sort_fixed: - { - // get a regular sort key, and then truncate it: - result.assign(this->m_pcollate->transform(p1, p2)); - result.erase(this->m_collate_delim); - break; - } - case sort_delim: - // get a regular sort key, and then truncate everything after the delim: - result.assign(this->m_pcollate->transform(p1, p2)); - std::size_t i; - for(i = 0; i < result.size(); ++i) - { - if(result[i] == m_collate_delim) - break; - } - result.erase(i); - break; - } -#ifndef BOOST_NO_EXCEPTIONS - }catch(...){} -#endif - while(result.size() && (charT(0) == *result.rbegin())) - result.erase(result.size() - 1); - if(result.empty()) - { - // character is ignorable at the primary level: - result = string_type(1, charT(0)); - } - return result; -} - -template -typename cpp_regex_traits_implementation::string_type - cpp_regex_traits_implementation::transform(const charT* p1, const charT* p2) const -{ - // - // PRECONDITIONS: - // - // A bug in gcc 3.2 (and maybe other versions as well) treats - // p1 as a null terminated string, for efficiency reasons - // we work around this elsewhere, but just assert here that - // we adhere to gcc's (buggy) preconditions... - // - BOOST_ASSERT(*p2 == 0); - // - // swallowing all exceptions here is a bad idea - // however at least one std lib will always throw - // std::bad_alloc for certain arguments... - // - string_type result; -#ifndef BOOST_NO_EXCEPTIONS - try{ -#endif - result = this->m_pcollate->transform(p1, p2); - // - // Borland's STLPort version returns a NULL-terminated - // string that has garbage at the end - each call to - // std::collate::transform returns a different string! - // So as a workaround, we'll truncate the string at the first NULL - // which _seems_ to work.... -#if BOOST_WORKAROUND(__BORLANDC__, < 0x580) - result.erase(result.find(charT(0))); -#else - // - // some implementations (Dinkumware) append unnecessary trailing \0's: - while(result.size() && (charT(0) == *result.rbegin())) - result.erase(result.size() - 1); -#endif - BOOST_ASSERT(std::find(result.begin(), result.end(), charT(0)) == result.end()); -#ifndef BOOST_NO_EXCEPTIONS - } - catch(...) - { - } -#endif - return result; -} - - -template -typename cpp_regex_traits_implementation::string_type - cpp_regex_traits_implementation::lookup_collatename(const charT* p1, const charT* p2) const -{ - typedef typename std::map::const_iterator iter_type; - if(m_custom_collate_names.size()) - { - iter_type pos = m_custom_collate_names.find(string_type(p1, p2)); - if(pos != m_custom_collate_names.end()) - return pos->second; - } -#if !defined(BOOST_NO_TEMPLATED_ITERATOR_CONSTRUCTORS)\ - && !BOOST_WORKAROUND(BOOST_MSVC, < 1300)\ - && !BOOST_WORKAROUND(__BORLANDC__, <= 0x0551) - std::string name(p1, p2); -#else - std::string name; - const charT* p0 = p1; - while(p0 != p2) - name.append(1, char(*p0++)); -#endif - name = lookup_default_collate_name(name); -#if !defined(BOOST_NO_TEMPLATED_ITERATOR_CONSTRUCTORS)\ - && !BOOST_WORKAROUND(BOOST_MSVC, < 1300)\ - && !BOOST_WORKAROUND(__BORLANDC__, <= 0x0551) - if(name.size()) - return string_type(name.begin(), name.end()); -#else - if(name.size()) - { - string_type result; - typedef std::string::const_iterator iter; - iter b = name.begin(); - iter e = name.end(); - while(b != e) - result.append(1, charT(*b++)); - return result; - } -#endif - if(p2 - p1 == 1) - return string_type(1, *p1); - return string_type(); -} - -template -void cpp_regex_traits_implementation::init() -{ -#ifndef BOOST_NO_STD_MESSAGES -#ifndef __IBMCPP__ - typename std::messages::catalog cat = static_cast::catalog>(-1); -#else - typename std::messages::catalog cat = reinterpret_cast::catalog>(-1); -#endif - std::string cat_name(cpp_regex_traits::get_catalog_name()); - if(cat_name.size() && (this->m_pmessages != 0)) - { - cat = this->m_pmessages->open( - cat_name, - this->m_locale); - if((int)cat < 0) - { - std::string m("Unable to open message catalog: "); - std::runtime_error err(m + cat_name); - cutl_details_boost::re_detail::raise_runtime_error(err); - } - } - // - // if we have a valid catalog then load our messages: - // - if((int)cat >= 0) - { - // - // Error messages: - // - for(cutl_details_boost::regex_constants::error_type i = static_cast(0); - i <= cutl_details_boost::regex_constants::error_unknown; - i = static_cast(i + 1)) - { - const char* p = get_default_error_string(i); - string_type default_message; - while(*p) - { - default_message.append(1, this->m_pctype->widen(*p)); - ++p; - } - string_type s = this->m_pmessages->get(cat, 0, i+200, default_message); - std::string result; - for(std::string::size_type j = 0; j < s.size(); ++j) - { - result.append(1, this->m_pctype->narrow(s[j], 0)); - } - m_error_strings[i] = result; - } - // - // Custom class names: - // -#ifndef BOOST_REGEX_BUGGY_CTYPE_FACET - static const char_class_type masks[16] = - { - std::ctype::alnum, - std::ctype::alpha, - std::ctype::cntrl, - std::ctype::digit, - std::ctype::graph, - cpp_regex_traits_implementation::mask_horizontal, - std::ctype::lower, - std::ctype::print, - std::ctype::punct, - std::ctype::space, - std::ctype::upper, - cpp_regex_traits_implementation::mask_vertical, - std::ctype::xdigit, - cpp_regex_traits_implementation::mask_blank, - cpp_regex_traits_implementation::mask_word, - cpp_regex_traits_implementation::mask_unicode, - }; -#else - static const char_class_type masks[16] = - { - ::cutl_details_boost::re_detail::char_class_alnum, - ::cutl_details_boost::re_detail::char_class_alpha, - ::cutl_details_boost::re_detail::char_class_cntrl, - ::cutl_details_boost::re_detail::char_class_digit, - ::cutl_details_boost::re_detail::char_class_graph, - ::cutl_details_boost::re_detail::char_class_horizontal_space, - ::cutl_details_boost::re_detail::char_class_lower, - ::cutl_details_boost::re_detail::char_class_print, - ::cutl_details_boost::re_detail::char_class_punct, - ::cutl_details_boost::re_detail::char_class_space, - ::cutl_details_boost::re_detail::char_class_upper, - ::cutl_details_boost::re_detail::char_class_vertical_space, - ::cutl_details_boost::re_detail::char_class_xdigit, - ::cutl_details_boost::re_detail::char_class_blank, - ::cutl_details_boost::re_detail::char_class_word, - ::cutl_details_boost::re_detail::char_class_unicode, - }; -#endif - static const string_type null_string; - for(unsigned int j = 0; j <= 13; ++j) - { - string_type s(this->m_pmessages->get(cat, 0, j+300, null_string)); - if(s.size()) - this->m_custom_class_names[s] = masks[j]; - } - } -#endif - // - // get the collation format used by m_pcollate: - // - m_collate_type = re_detail::find_sort_syntax(this, &m_collate_delim); -} - -template -typename cpp_regex_traits_implementation::char_class_type - cpp_regex_traits_implementation::lookup_classname_imp(const charT* p1, const charT* p2) const -{ -#ifndef BOOST_REGEX_BUGGY_CTYPE_FACET - static const char_class_type masks[22] = - { - 0, - std::ctype::alnum, - std::ctype::alpha, - cpp_regex_traits_implementation::mask_blank, - std::ctype::cntrl, - std::ctype::digit, - std::ctype::digit, - std::ctype::graph, - cpp_regex_traits_implementation::mask_horizontal, - std::ctype::lower, - std::ctype::lower, - std::ctype::print, - std::ctype::punct, - std::ctype::space, - std::ctype::space, - std::ctype::upper, - cpp_regex_traits_implementation::mask_unicode, - std::ctype::upper, - cpp_regex_traits_implementation::mask_vertical, - std::ctype::alnum | cpp_regex_traits_implementation::mask_word, - std::ctype::alnum | cpp_regex_traits_implementation::mask_word, - std::ctype::xdigit, - }; -#else - static const char_class_type masks[22] = - { - 0, - ::cutl_details_boost::re_detail::char_class_alnum, - ::cutl_details_boost::re_detail::char_class_alpha, - ::cutl_details_boost::re_detail::char_class_blank, - ::cutl_details_boost::re_detail::char_class_cntrl, - ::cutl_details_boost::re_detail::char_class_digit, - ::cutl_details_boost::re_detail::char_class_digit, - ::cutl_details_boost::re_detail::char_class_graph, - ::cutl_details_boost::re_detail::char_class_horizontal_space, - ::cutl_details_boost::re_detail::char_class_lower, - ::cutl_details_boost::re_detail::char_class_lower, - ::cutl_details_boost::re_detail::char_class_print, - ::cutl_details_boost::re_detail::char_class_punct, - ::cutl_details_boost::re_detail::char_class_space, - ::cutl_details_boost::re_detail::char_class_space, - ::cutl_details_boost::re_detail::char_class_upper, - ::cutl_details_boost::re_detail::char_class_unicode, - ::cutl_details_boost::re_detail::char_class_upper, - ::cutl_details_boost::re_detail::char_class_vertical_space, - ::cutl_details_boost::re_detail::char_class_alnum | ::cutl_details_boost::re_detail::char_class_word, - ::cutl_details_boost::re_detail::char_class_alnum | ::cutl_details_boost::re_detail::char_class_word, - ::cutl_details_boost::re_detail::char_class_xdigit, - }; -#endif - if(m_custom_class_names.size()) - { - typedef typename std::map, char_class_type>::const_iterator map_iter; - map_iter pos = m_custom_class_names.find(string_type(p1, p2)); - if(pos != m_custom_class_names.end()) - return pos->second; - } - std::size_t state_id = 1 + re_detail::get_default_class_id(p1, p2); - BOOST_ASSERT(state_id < sizeof(masks) / sizeof(masks[0])); - return masks[state_id]; -} - -#ifdef BOOST_REGEX_BUGGY_CTYPE_FACET -template -bool cpp_regex_traits_implementation::isctype(const charT c, char_class_type mask) const -{ - return - ((mask & ::cutl_details_boost::re_detail::char_class_space) && (this->m_pctype->is(std::ctype::space, c))) - || ((mask & ::cutl_details_boost::re_detail::char_class_print) && (this->m_pctype->is(std::ctype::print, c))) - || ((mask & ::cutl_details_boost::re_detail::char_class_cntrl) && (this->m_pctype->is(std::ctype::cntrl, c))) - || ((mask & ::cutl_details_boost::re_detail::char_class_upper) && (this->m_pctype->is(std::ctype::upper, c))) - || ((mask & ::cutl_details_boost::re_detail::char_class_lower) && (this->m_pctype->is(std::ctype::lower, c))) - || ((mask & ::cutl_details_boost::re_detail::char_class_alpha) && (this->m_pctype->is(std::ctype::alpha, c))) - || ((mask & ::cutl_details_boost::re_detail::char_class_digit) && (this->m_pctype->is(std::ctype::digit, c))) - || ((mask & ::cutl_details_boost::re_detail::char_class_punct) && (this->m_pctype->is(std::ctype::punct, c))) - || ((mask & ::cutl_details_boost::re_detail::char_class_xdigit) && (this->m_pctype->is(std::ctype::xdigit, c))) - || ((mask & ::cutl_details_boost::re_detail::char_class_blank) && (this->m_pctype->is(std::ctype::space, c)) && !::cutl_details_boost::re_detail::is_separator(c)) - || ((mask & ::cutl_details_boost::re_detail::char_class_word) && (c == '_')) - || ((mask & ::cutl_details_boost::re_detail::char_class_unicode) && ::cutl_details_boost::re_detail::is_extended(c)) - || ((mask & ::cutl_details_boost::re_detail::char_class_vertical_space) && (is_separator(c) || (c == '\v'))) - || ((mask & ::cutl_details_boost::re_detail::char_class_horizontal_space) && this->m_pctype->is(std::ctype::space, c) && !(is_separator(c) || (c == '\v'))); -} -#endif - - -template -inline cutl_details_boost::shared_ptr > create_cpp_regex_traits(const std::locale& l BOOST_APPEND_EXPLICIT_TEMPLATE_TYPE(charT)) -{ - cpp_regex_traits_base key(l); - return ::cutl_details_boost::object_cache, cpp_regex_traits_implementation >::get(key, 5); -} - -} // re_detail - -template -class cpp_regex_traits -{ -private: - typedef std::ctype ctype_type; -public: - typedef charT char_type; - typedef std::size_t size_type; - typedef std::basic_string string_type; - typedef std::locale locale_type; - typedef cutl_details_boost::uint_least32_t char_class_type; - - struct boost_extensions_tag{}; - - cpp_regex_traits() - : m_pimpl(re_detail::create_cpp_regex_traits(std::locale())) - { } - static size_type length(const char_type* p) - { - return std::char_traits::length(p); - } - regex_constants::syntax_type syntax_type(charT c)const - { - return m_pimpl->syntax_type(c); - } - regex_constants::escape_syntax_type escape_syntax_type(charT c) const - { - return m_pimpl->escape_syntax_type(c); - } - charT translate(charT c) const - { - return c; - } - charT translate_nocase(charT c) const - { - return m_pimpl->m_pctype->tolower(c); - } - charT translate(charT c, bool icase) const - { - return icase ? m_pimpl->m_pctype->tolower(c) : c; - } - charT tolower(charT c) const - { - return m_pimpl->m_pctype->tolower(c); - } - charT toupper(charT c) const - { - return m_pimpl->m_pctype->toupper(c); - } - string_type transform(const charT* p1, const charT* p2) const - { - return m_pimpl->transform(p1, p2); - } - string_type transform_primary(const charT* p1, const charT* p2) const - { - return m_pimpl->transform_primary(p1, p2); - } - char_class_type lookup_classname(const charT* p1, const charT* p2) const - { - return m_pimpl->lookup_classname(p1, p2); - } - string_type lookup_collatename(const charT* p1, const charT* p2) const - { - return m_pimpl->lookup_collatename(p1, p2); - } - bool isctype(charT c, char_class_type f) const - { -#ifndef BOOST_REGEX_BUGGY_CTYPE_FACET - typedef typename std::ctype::mask ctype_mask; - - static const ctype_mask mask_base = - static_cast( - std::ctype::alnum - | std::ctype::alpha - | std::ctype::cntrl - | std::ctype::digit - | std::ctype::graph - | std::ctype::lower - | std::ctype::print - | std::ctype::punct - | std::ctype::space - | std::ctype::upper - | std::ctype::xdigit); - - if((f & mask_base) - && (m_pimpl->m_pctype->is( - static_cast(f & mask_base), c))) - return true; - else if((f & re_detail::cpp_regex_traits_implementation::mask_unicode) && re_detail::is_extended(c)) - return true; - else if((f & re_detail::cpp_regex_traits_implementation::mask_word) && (c == '_')) - return true; - else if((f & re_detail::cpp_regex_traits_implementation::mask_blank) - && m_pimpl->m_pctype->is(std::ctype::space, c) - && !re_detail::is_separator(c)) - return true; - else if((f & re_detail::cpp_regex_traits_implementation::mask_vertical) - && (::cutl_details_boost::re_detail::is_separator(c) || (c == '\v'))) - return true; - else if((f & re_detail::cpp_regex_traits_implementation::mask_horizontal) - && this->isctype(c, std::ctype::space) && !this->isctype(c, re_detail::cpp_regex_traits_implementation::mask_vertical)) - return true; - return false; -#else - return m_pimpl->isctype(c, f); -#endif - } - int toi(const charT*& p1, const charT* p2, int radix)const; - int value(charT c, int radix)const - { - const charT* pc = &c; - return toi(pc, pc + 1, radix); - } - locale_type imbue(locale_type l) - { - std::locale result(getloc()); - m_pimpl = re_detail::create_cpp_regex_traits(l); - return result; - } - locale_type getloc()const - { - return m_pimpl->m_locale; - } - std::string error_string(regex_constants::error_type n) const - { - return m_pimpl->error_string(n); - } - - // - // extension: - // set the name of the message catalog in use (defaults to "boost_regex"). - // - static std::string catalog_name(const std::string& name); - static std::string get_catalog_name(); - -private: - cutl_details_boost::shared_ptr > m_pimpl; - // - // catalog name handler: - // - static std::string& get_catalog_name_inst(); - -#ifdef BOOST_HAS_THREADS - static static_mutex& get_mutex_inst(); -#endif -}; - - -template -int cpp_regex_traits::toi(const charT*& first, const charT* last, int radix)const -{ - re_detail::parser_buf sbuf; // buffer for parsing numbers. - std::basic_istream is(&sbuf); // stream for parsing numbers. - - // we do NOT want to parse any thousands separators inside the stream: - last = std::find(first, last, BOOST_USE_FACET(std::numpunct, is.getloc()).thousands_sep()); - - sbuf.pubsetbuf(const_cast(static_cast(first)), static_cast(last-first)); - is.clear(); - if(std::abs(radix) == 16) is >> std::hex; - else if(std::abs(radix) == 8) is >> std::oct; - else is >> std::dec; - int val; - if(is >> val) - { - first = first + ((last - first) - sbuf.in_avail()); - return val; - } - else - return -1; -} - -template -std::string cpp_regex_traits::catalog_name(const std::string& name) -{ -#ifdef BOOST_HAS_THREADS - static_mutex::scoped_lock lk(get_mutex_inst()); -#endif - std::string result(get_catalog_name_inst()); - get_catalog_name_inst() = name; - return result; -} - -template -std::string& cpp_regex_traits::get_catalog_name_inst() -{ - static std::string s_name; - return s_name; -} - -template -std::string cpp_regex_traits::get_catalog_name() -{ -#ifdef BOOST_HAS_THREADS - static_mutex::scoped_lock lk(get_mutex_inst()); -#endif - std::string result(get_catalog_name_inst()); - return result; -} - -#ifdef BOOST_HAS_THREADS -template -static_mutex& cpp_regex_traits::get_mutex_inst() -{ - static static_mutex s_mutex = BOOST_STATIC_MUTEX_INIT; - return s_mutex; -} -#endif - - -} // boost - -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif - -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable: 4103) -#endif -#ifdef BOOST_HAS_ABI_HEADERS -# include BOOST_ABI_SUFFIX -#endif -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif - -#endif - -#endif - - diff --git a/cutl/details/boost/regex/v4/cregex.hpp b/cutl/details/boost/regex/v4/cregex.hpp deleted file mode 100644 index 0930a84..0000000 --- a/cutl/details/boost/regex/v4/cregex.hpp +++ /dev/null @@ -1,330 +0,0 @@ -/* - * - * Copyright (c) 1998-2002 - * John Maddock - * - * Use, modification and distribution are subject to the - * Boost Software License, Version 1.0. (See accompanying file - * LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt) - * - */ - - /* - * LOCATION: see http://www.boost.org for most recent version. - * FILE cregex.cpp - * VERSION see - * DESCRIPTION: Declares POSIX API functions - * + cutl_details_boost::RegEx high level wrapper. - */ - -#ifndef BOOST_RE_CREGEX_HPP_INCLUDED -#define BOOST_RE_CREGEX_HPP_INCLUDED - -#ifndef BOOST_REGEX_CONFIG_HPP -#include -#endif -#include -#include - -#ifdef __cplusplus -#include -#else -#include -#endif - -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable: 4103) -#endif -#ifdef BOOST_HAS_ABI_HEADERS -# include BOOST_ABI_PREFIX -#endif -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif - -/* include these defs only for POSIX compatablity */ -#ifdef __cplusplus -namespace cutl_details_boost{ -extern "C" { -#endif - -#if defined(__cplusplus) && !defined(BOOST_NO_STDC_NAMESPACE) -typedef std::ptrdiff_t regoff_t; -typedef std::size_t regsize_t; -#else -typedef ptrdiff_t regoff_t; -typedef size_t regsize_t; -#endif - -typedef struct -{ - unsigned int re_magic; -#ifdef __cplusplus - std::size_t re_nsub; /* number of parenthesized subexpressions */ -#else - size_t re_nsub; -#endif - const char* re_endp; /* end pointer for REG_PEND */ - void* guts; /* none of your business :-) */ - match_flag_type eflags; /* none of your business :-) */ -} regex_tA; - -#ifndef BOOST_NO_WREGEX -typedef struct -{ - unsigned int re_magic; -#ifdef __cplusplus - std::size_t re_nsub; /* number of parenthesized subexpressions */ -#else - size_t re_nsub; -#endif - const wchar_t* re_endp; /* end pointer for REG_PEND */ - void* guts; /* none of your business :-) */ - match_flag_type eflags; /* none of your business :-) */ -} regex_tW; -#endif - -typedef struct -{ - regoff_t rm_so; /* start of match */ - regoff_t rm_eo; /* end of match */ -} regmatch_t; - -/* regcomp() flags */ -typedef enum{ - REG_BASIC = 0000, - REG_EXTENDED = 0001, - REG_ICASE = 0002, - REG_NOSUB = 0004, - REG_NEWLINE = 0010, - REG_NOSPEC = 0020, - REG_PEND = 0040, - REG_DUMP = 0200, - REG_NOCOLLATE = 0400, - REG_ESCAPE_IN_LISTS = 01000, - REG_NEWLINE_ALT = 02000, - REG_PERLEX = 04000, - - REG_PERL = REG_EXTENDED | REG_NOCOLLATE | REG_ESCAPE_IN_LISTS | REG_PERLEX, - REG_AWK = REG_EXTENDED | REG_ESCAPE_IN_LISTS, - REG_GREP = REG_BASIC | REG_NEWLINE_ALT, - REG_EGREP = REG_EXTENDED | REG_NEWLINE_ALT, - - REG_ASSERT = 15, - REG_INVARG = 16, - REG_ATOI = 255, /* convert name to number (!) */ - REG_ITOA = 0400 /* convert number to name (!) */ -} reg_comp_flags; - -/* regexec() flags */ -typedef enum{ - REG_NOTBOL = 00001, - REG_NOTEOL = 00002, - REG_STARTEND = 00004 -} reg_exec_flags; - -/* - * POSIX error codes: - */ -typedef unsigned reg_error_t; -typedef reg_error_t reg_errcode_t; /* backwards compatibility */ - -static const reg_error_t REG_NOERROR = 0; /* Success. */ -static const reg_error_t REG_NOMATCH = 1; /* Didn't find a match (for regexec). */ - - /* POSIX regcomp return error codes. (In the order listed in the - standard.) */ -static const reg_error_t REG_BADPAT = 2; /* Invalid pattern. */ -static const reg_error_t REG_ECOLLATE = 3; /* Undefined collating element. */ -static const reg_error_t REG_ECTYPE = 4; /* Invalid character class name. */ -static const reg_error_t REG_EESCAPE = 5; /* Trailing backslash. */ -static const reg_error_t REG_ESUBREG = 6; /* Invalid back reference. */ -static const reg_error_t REG_EBRACK = 7; /* Unmatched left bracket. */ -static const reg_error_t REG_EPAREN = 8; /* Parenthesis imbalance. */ -static const reg_error_t REG_EBRACE = 9; /* Unmatched \{. */ -static const reg_error_t REG_BADBR = 10; /* Invalid contents of \{\}. */ -static const reg_error_t REG_ERANGE = 11; /* Invalid range end. */ -static const reg_error_t REG_ESPACE = 12; /* Ran out of memory. */ -static const reg_error_t REG_BADRPT = 13; /* No preceding re for repetition op. */ -static const reg_error_t REG_EEND = 14; /* unexpected end of expression */ -static const reg_error_t REG_ESIZE = 15; /* expression too big */ -static const reg_error_t REG_ERPAREN = 8; /* = REG_EPAREN : unmatched right parenthesis */ -static const reg_error_t REG_EMPTY = 17; /* empty expression */ -static const reg_error_t REG_E_MEMORY = 15; /* = REG_ESIZE : out of memory */ -static const reg_error_t REG_ECOMPLEXITY = 18; /* complexity too high */ -static const reg_error_t REG_ESTACK = 19; /* out of stack space */ -static const reg_error_t REG_E_PERL = 20; /* Perl (?...) error */ -static const reg_error_t REG_E_UNKNOWN = 21; /* unknown error */ -static const reg_error_t REG_ENOSYS = 21; /* = REG_E_UNKNOWN : Reserved. */ - -BOOST_REGEX_DECL int BOOST_REGEX_CCALL regcompA(regex_tA*, const char*, int); -BOOST_REGEX_DECL regsize_t BOOST_REGEX_CCALL regerrorA(int, const regex_tA*, char*, regsize_t); -BOOST_REGEX_DECL int BOOST_REGEX_CCALL regexecA(const regex_tA*, const char*, regsize_t, regmatch_t*, int); -BOOST_REGEX_DECL void BOOST_REGEX_CCALL regfreeA(regex_tA*); - -#ifndef BOOST_NO_WREGEX -BOOST_REGEX_DECL int BOOST_REGEX_CCALL regcompW(regex_tW*, const wchar_t*, int); -BOOST_REGEX_DECL regsize_t BOOST_REGEX_CCALL regerrorW(int, const regex_tW*, wchar_t*, regsize_t); -BOOST_REGEX_DECL int BOOST_REGEX_CCALL regexecW(const regex_tW*, const wchar_t*, regsize_t, regmatch_t*, int); -BOOST_REGEX_DECL void BOOST_REGEX_CCALL regfreeW(regex_tW*); -#endif - -#ifdef UNICODE -#define regcomp regcompW -#define regerror regerrorW -#define regexec regexecW -#define regfree regfreeW -#define regex_t regex_tW -#else -#define regcomp regcompA -#define regerror regerrorA -#define regexec regexecA -#define regfree regfreeA -#define regex_t regex_tA -#endif - -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable: 4103) -#endif -#ifdef BOOST_HAS_ABI_HEADERS -# include BOOST_ABI_SUFFIX -#endif -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif - -#ifdef __cplusplus -} /* extern "C" */ -} /* namespace */ -#endif - -#if defined(__cplusplus) -/* - * C++ high level wrapper goes here: - */ -#include -#include -namespace cutl_details_boost{ - -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable: 4103) -#endif -#ifdef BOOST_HAS_ABI_HEADERS -# include BOOST_ABI_PREFIX -#endif -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif - -class RegEx; - -namespace re_detail{ - -class RegExData; -struct pred1; -struct pred2; -struct pred3; -struct pred4; - -} /* namespace re_detail */ - -#if (defined(BOOST_MSVC) || defined(__BORLANDC__)) && !defined(BOOST_DISABLE_WIN32) -typedef bool (__cdecl *GrepCallback)(const RegEx& expression); -typedef bool (__cdecl *GrepFileCallback)(const char* file, const RegEx& expression); -typedef bool (__cdecl *FindFilesCallback)(const char* file); -#else -typedef bool (*GrepCallback)(const RegEx& expression); -typedef bool (*GrepFileCallback)(const char* file, const RegEx& expression); -typedef bool (*FindFilesCallback)(const char* file); -#endif - -class BOOST_REGEX_DECL RegEx -{ -private: - re_detail::RegExData* pdata; -public: - RegEx(); - RegEx(const RegEx& o); - ~RegEx(); - explicit RegEx(const char* c, bool icase = false); - explicit RegEx(const std::string& s, bool icase = false); - RegEx& operator=(const RegEx& o); - RegEx& operator=(const char* p); - RegEx& operator=(const std::string& s){ return this->operator=(s.c_str()); } - unsigned int SetExpression(const char* p, bool icase = false); - unsigned int SetExpression(const std::string& s, bool icase = false){ return SetExpression(s.c_str(), icase); } - std::string Expression()const; - unsigned int error_code()const; - /* - * now matching operators: - */ - bool Match(const char* p, match_flag_type flags = match_default); - bool Match(const std::string& s, match_flag_type flags = match_default) { return Match(s.c_str(), flags); } - bool Search(const char* p, match_flag_type flags = match_default); - bool Search(const std::string& s, match_flag_type flags = match_default) { return Search(s.c_str(), flags); } - unsigned int Grep(GrepCallback cb, const char* p, match_flag_type flags = match_default); - unsigned int Grep(GrepCallback cb, const std::string& s, match_flag_type flags = match_default) { return Grep(cb, s.c_str(), flags); } - unsigned int Grep(std::vector& v, const char* p, match_flag_type flags = match_default); - unsigned int Grep(std::vector& v, const std::string& s, match_flag_type flags = match_default) { return Grep(v, s.c_str(), flags); } - unsigned int Grep(std::vector& v, const char* p, match_flag_type flags = match_default); - unsigned int Grep(std::vector& v, const std::string& s, match_flag_type flags = match_default) { return Grep(v, s.c_str(), flags); } -#ifndef BOOST_REGEX_NO_FILEITER - unsigned int GrepFiles(GrepFileCallback cb, const char* files, bool recurse = false, match_flag_type flags = match_default); - unsigned int GrepFiles(GrepFileCallback cb, const std::string& files, bool recurse = false, match_flag_type flags = match_default) { return GrepFiles(cb, files.c_str(), recurse, flags); } - unsigned int FindFiles(FindFilesCallback cb, const char* files, bool recurse = false, match_flag_type flags = match_default); - unsigned int FindFiles(FindFilesCallback cb, const std::string& files, bool recurse = false, match_flag_type flags = match_default) { return FindFiles(cb, files.c_str(), recurse, flags); } -#endif - - std::string Merge(const std::string& in, const std::string& fmt, - bool copy = true, match_flag_type flags = match_default); - std::string Merge(const char* in, const char* fmt, - bool copy = true, match_flag_type flags = match_default); - - std::size_t Split(std::vector& v, std::string& s, match_flag_type flags = match_default, unsigned max_count = ~0); - /* - * now operators for returning what matched in more detail: - */ - std::size_t Position(int i = 0)const; - std::size_t Length(int i = 0)const; - bool Matched(int i = 0)const; - std::size_t Marks()const; - std::string What(int i = 0)const; - std::string operator[](int i)const { return What(i); } - - static const std::size_t npos; - - friend struct re_detail::pred1; - friend struct re_detail::pred2; - friend struct re_detail::pred3; - friend struct re_detail::pred4; -}; - -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable: 4103) -#endif -#ifdef BOOST_HAS_ABI_HEADERS -# include BOOST_ABI_SUFFIX -#endif -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif - -} /* namespace cutl_details_boost */ - -#endif /* __cplusplus */ - -#endif /* include guard */ - - - - - - - - - - diff --git a/cutl/details/boost/regex/v4/error_type.hpp b/cutl/details/boost/regex/v4/error_type.hpp deleted file mode 100644 index db00e46..0000000 --- a/cutl/details/boost/regex/v4/error_type.hpp +++ /dev/null @@ -1,59 +0,0 @@ -/* - * - * Copyright (c) 2003-2005 - * John Maddock - * - * Use, modification and distribution are subject to the - * Boost Software License, Version 1.0. (See accompanying file - * LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt) - * - */ - - /* - * LOCATION: see http://www.boost.org for most recent version. - * FILE error_type.hpp - * VERSION see - * DESCRIPTION: Declares regular expression error type enumerator. - */ - -#ifndef BOOST_REGEX_ERROR_TYPE_HPP -#define BOOST_REGEX_ERROR_TYPE_HPP - -#ifdef __cplusplus -namespace cutl_details_boost{ -#endif - -#ifdef __cplusplus -namespace regex_constants{ - -enum error_type{ - - error_ok = 0, /* not used */ - error_no_match = 1, /* not used */ - error_bad_pattern = 2, - error_collate = 3, - error_ctype = 4, - error_escape = 5, - error_backref = 6, - error_brack = 7, - error_paren = 8, - error_brace = 9, - error_badbrace = 10, - error_range = 11, - error_space = 12, - error_badrepeat = 13, - error_end = 14, /* not used */ - error_size = 15, - error_right_paren = 16, /* not used */ - error_empty = 17, - error_complexity = 18, - error_stack = 19, - error_perl_extension = 20, - error_unknown = 21 -}; - -} -} -#endif /* __cplusplus */ - -#endif diff --git a/cutl/details/boost/regex/v4/fileiter.hpp b/cutl/details/boost/regex/v4/fileiter.hpp deleted file mode 100644 index 8474f0a..0000000 --- a/cutl/details/boost/regex/v4/fileiter.hpp +++ /dev/null @@ -1,455 +0,0 @@ -/* - * - * Copyright (c) 1998-2002 - * John Maddock - * - * Use, modification and distribution are subject to the - * Boost Software License, Version 1.0. (See accompanying file - * LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt) - * - */ - - /* - * LOCATION: see http://www.boost.org for most recent version. - * FILE fileiter.hpp - * VERSION see - * DESCRIPTION: Declares various platform independent file and - * directory iterators, plus binary file input in - * the form of class map_file. - */ - -#ifndef BOOST_RE_FILEITER_HPP_INCLUDED -#define BOOST_RE_FILEITER_HPP_INCLUDED - -#ifndef BOOST_REGEX_CONFIG_HPP -#include -#endif -#include - -#ifndef BOOST_REGEX_NO_FILEITER - -#if (defined(__CYGWIN__) || defined(__CYGWIN32__)) && !defined(BOOST_REGEX_NO_W32) -#error "Sorry, can't mix with STL code and gcc compiler: if you ran configure, try again with configure --disable-ms-windows" -#define BOOST_REGEX_FI_WIN32_MAP -#define BOOST_REGEX_FI_POSIX_DIR -#elif (defined(__WIN32__) || defined(_WIN32) || defined(WIN32)) && !defined(BOOST_REGEX_NO_W32) -#define BOOST_REGEX_FI_WIN32_MAP -#define BOOST_REGEX_FI_WIN32_DIR -#else -#define BOOST_REGEX_FI_POSIX_MAP -#define BOOST_REGEX_FI_POSIX_DIR -#endif - -#if defined(BOOST_REGEX_FI_WIN32_MAP)||defined(BOOST_REGEX_FI_WIN32_DIR) -#include -#endif - -#if defined(BOOST_REGEX_FI_WIN32_DIR) - -#include - -namespace cutl_details_boost{ - namespace re_detail{ - -#ifndef BOOST_NO_ANSI_APIS -typedef WIN32_FIND_DATAA _fi_find_data; -#else -typedef WIN32_FIND_DATAW _fi_find_data; -#endif -typedef HANDLE _fi_find_handle; - - } // namespace re_detail - -} // namespace cutl_details_boost - -#define _fi_invalid_handle INVALID_HANDLE_VALUE -#define _fi_dir FILE_ATTRIBUTE_DIRECTORY - -#elif defined(BOOST_REGEX_FI_POSIX_DIR) - -#include -#include -#include -#include -#include -#include -#include - -#if defined(__SUNPRO_CC) -using std::list; -#endif - -#ifndef MAX_PATH -#define MAX_PATH 256 -#endif - -namespace cutl_details_boost{ - namespace re_detail{ - -#ifdef BOOST_HAS_ABI_HEADERS -# include BOOST_ABI_PREFIX -#endif - -struct _fi_find_data -{ - unsigned dwFileAttributes; - char cFileName[MAX_PATH]; -}; - -struct _fi_priv_data; - -typedef _fi_priv_data* _fi_find_handle; -#define _fi_invalid_handle 0 -#define _fi_dir 1 - -_fi_find_handle _fi_FindFirstFile(const char* lpFileName, _fi_find_data* lpFindFileData); -bool _fi_FindNextFile(_fi_find_handle hFindFile, _fi_find_data* lpFindFileData); -bool _fi_FindClose(_fi_find_handle hFindFile); - -#ifdef BOOST_HAS_ABI_HEADERS -# include BOOST_ABI_SUFFIX -#endif - - } // namespace re_detail -} // namespace cutl_details_boost - -#ifdef FindFirstFile - #undef FindFirstFile -#endif -#ifdef FindNextFile - #undef FindNextFile -#endif -#ifdef FindClose - #undef FindClose -#endif - -#define FindFirstFileA _fi_FindFirstFile -#define FindNextFileA _fi_FindNextFile -#define FindClose _fi_FindClose - -#endif - -namespace cutl_details_boost{ - namespace re_detail{ - -#ifdef BOOST_HAS_ABI_HEADERS -# include BOOST_ABI_PREFIX -#endif - -#ifdef BOOST_REGEX_FI_WIN32_MAP // win32 mapfile - -class BOOST_REGEX_DECL mapfile -{ - HANDLE hfile; - HANDLE hmap; - const char* _first; - const char* _last; -public: - - typedef const char* iterator; - - mapfile(){ hfile = hmap = 0; _first = _last = 0; } - mapfile(const char* file){ hfile = hmap = 0; _first = _last = 0; open(file); } - ~mapfile(){ close(); } - void open(const char* file); - void close(); - const char* begin(){ return _first; } - const char* end(){ return _last; } - size_t size(){ return _last - _first; } - bool valid(){ return (hfile != 0) && (hfile != INVALID_HANDLE_VALUE); } -}; - - -#else - -class BOOST_REGEX_DECL mapfile_iterator; - -class BOOST_REGEX_DECL mapfile -{ - typedef char* pointer; - std::FILE* hfile; - long int _size; - pointer* _first; - pointer* _last; - mutable std::list condemed; - enum sizes - { - buf_size = 4096 - }; - void lock(pointer* node)const; - void unlock(pointer* node)const; -public: - - typedef mapfile_iterator iterator; - - mapfile(){ hfile = 0; _size = 0; _first = _last = 0; } - mapfile(const char* file){ hfile = 0; _size = 0; _first = _last = 0; open(file); } - ~mapfile(){ close(); } - void open(const char* file); - void close(); - iterator begin()const; - iterator end()const; - unsigned long size()const{ return _size; } - bool valid()const{ return hfile != 0; } - friend class mapfile_iterator; -}; - -class BOOST_REGEX_DECL mapfile_iterator -#if !defined(BOOST_NO_STD_ITERATOR) || defined(BOOST_MSVC_STD_ITERATOR) -: public std::iterator -#endif -{ - typedef mapfile::pointer internal_pointer; - internal_pointer* node; - const mapfile* file; - unsigned long offset; - long position()const - { - return file ? ((node - file->_first) * mapfile::buf_size + offset) : 0; - } - void position(long pos) - { - if(file) - { - node = file->_first + (pos / mapfile::buf_size); - offset = pos % mapfile::buf_size; - } - } -public: - typedef std::ptrdiff_t difference_type; - typedef char value_type; - typedef const char* pointer; - typedef const char& reference; - typedef std::random_access_iterator_tag iterator_category; - - mapfile_iterator() { node = 0; file = 0; offset = 0; } - mapfile_iterator(const mapfile* f, long arg_position) - { - file = f; - node = f->_first + arg_position / mapfile::buf_size; - offset = arg_position % mapfile::buf_size; - if(file) - file->lock(node); - } - mapfile_iterator(const mapfile_iterator& i) - { - file = i.file; - node = i.node; - offset = i.offset; - if(file) - file->lock(node); - } - ~mapfile_iterator() - { - if(file && node) - file->unlock(node); - } - mapfile_iterator& operator = (const mapfile_iterator& i); - char operator* ()const - { - BOOST_ASSERT(node >= file->_first); - BOOST_ASSERT(node < file->_last); - return file ? *(*node + sizeof(int) + offset) : char(0); - } - char operator[] (long off)const - { - mapfile_iterator tmp(*this); - tmp += off; - return *tmp; - } - mapfile_iterator& operator++ (); - mapfile_iterator operator++ (int); - mapfile_iterator& operator-- (); - mapfile_iterator operator-- (int); - - mapfile_iterator& operator += (long off) - { - position(position() + off); - return *this; - } - mapfile_iterator& operator -= (long off) - { - position(position() - off); - return *this; - } - - friend inline bool operator==(const mapfile_iterator& i, const mapfile_iterator& j) - { - return (i.file == j.file) && (i.node == j.node) && (i.offset == j.offset); - } - - friend inline bool operator!=(const mapfile_iterator& i, const mapfile_iterator& j) - { - return !(i == j); - } - - friend inline bool operator<(const mapfile_iterator& i, const mapfile_iterator& j) - { - return i.position() < j.position(); - } - friend inline bool operator>(const mapfile_iterator& i, const mapfile_iterator& j) - { - return i.position() > j.position(); - } - friend inline bool operator<=(const mapfile_iterator& i, const mapfile_iterator& j) - { - return i.position() <= j.position(); - } - friend inline bool operator>=(const mapfile_iterator& i, const mapfile_iterator& j) - { - return i.position() >= j.position(); - } - - friend mapfile_iterator operator + (const mapfile_iterator& i, long off); - friend mapfile_iterator operator + (long off, const mapfile_iterator& i) - { - mapfile_iterator tmp(i); - return tmp += off; - } - friend mapfile_iterator operator - (const mapfile_iterator& i, long off); - friend inline long operator - (const mapfile_iterator& i, const mapfile_iterator& j) - { - return i.position() - j.position(); - } -}; - -#endif - -// _fi_sep determines the directory separator, either '\\' or '/' -BOOST_REGEX_DECL extern const char* _fi_sep; - -struct file_iterator_ref -{ - _fi_find_handle hf; - _fi_find_data _data; - long count; -}; - - -class BOOST_REGEX_DECL file_iterator -{ - char* _root; - char* _path; - char* ptr; - file_iterator_ref* ref; - -public: - typedef std::ptrdiff_t difference_type; - typedef const char* value_type; - typedef const char** pointer; - typedef const char*& reference; - typedef std::input_iterator_tag iterator_category; - - file_iterator(); - file_iterator(const char* wild); - ~file_iterator(); - file_iterator(const file_iterator&); - file_iterator& operator=(const file_iterator&); - const char* root()const { return _root; } - const char* path()const { return _path; } - const char* name()const { return ptr; } - _fi_find_data* data() { return &(ref->_data); } - void next(); - file_iterator& operator++() { next(); return *this; } - file_iterator operator++(int); - const char* operator*() { return path(); } - - friend inline bool operator == (const file_iterator& f1, const file_iterator& f2) - { - return ((f1.ref->hf == _fi_invalid_handle) && (f2.ref->hf == _fi_invalid_handle)); - } - - friend inline bool operator != (const file_iterator& f1, const file_iterator& f2) - { - return !(f1 == f2); - } - -}; - -// dwa 9/13/00 - suppress unused parameter warning -inline bool operator < (const file_iterator&, const file_iterator&) -{ - return false; -} - - -class BOOST_REGEX_DECL directory_iterator -{ - char* _root; - char* _path; - char* ptr; - file_iterator_ref* ref; - -public: - typedef std::ptrdiff_t difference_type; - typedef const char* value_type; - typedef const char** pointer; - typedef const char*& reference; - typedef std::input_iterator_tag iterator_category; - - directory_iterator(); - directory_iterator(const char* wild); - ~directory_iterator(); - directory_iterator(const directory_iterator& other); - directory_iterator& operator=(const directory_iterator& other); - - const char* root()const { return _root; } - const char* path()const { return _path; } - const char* name()const { return ptr; } - _fi_find_data* data() { return &(ref->_data); } - void next(); - directory_iterator& operator++() { next(); return *this; } - directory_iterator operator++(int); - const char* operator*() { return path(); } - - static const char* separator() { return _fi_sep; } - - friend inline bool operator == (const directory_iterator& f1, const directory_iterator& f2) - { - return ((f1.ref->hf == _fi_invalid_handle) && (f2.ref->hf == _fi_invalid_handle)); - } - - - friend inline bool operator != (const directory_iterator& f1, const directory_iterator& f2) - { - return !(f1 == f2); - } - - }; - -inline bool operator < (const directory_iterator&, const directory_iterator&) -{ - return false; -} - -#ifdef BOOST_HAS_ABI_HEADERS -# include BOOST_ABI_SUFFIX -#endif - - -} // namespace re_detail -using cutl_details_boost::re_detail::directory_iterator; -using cutl_details_boost::re_detail::file_iterator; -using cutl_details_boost::re_detail::mapfile; -} // namespace cutl_details_boost - -#endif // BOOST_REGEX_NO_FILEITER -#endif // BOOST_RE_FILEITER_HPP - - - - - - - - - - - - - - - - - - diff --git a/cutl/details/boost/regex/v4/instances.hpp b/cutl/details/boost/regex/v4/instances.hpp deleted file mode 100644 index 831d6ad..0000000 --- a/cutl/details/boost/regex/v4/instances.hpp +++ /dev/null @@ -1,222 +0,0 @@ -/* - * - * Copyright (c) 1998-2002 - * John Maddock - * - * Use, modification and distribution are subject to the - * Boost Software License, Version 1.0. (See accompanying file - * LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt) - * - */ - - /* - * LOCATION: see http://www.boost.org for most recent version. - * FILE instances.cpp - * VERSION see - * DESCRIPTION: Defines those template instances that are placed in the - * library rather than in the users object files. - */ - -// -// note no include guard, we may include this multiple times: -// -#ifndef BOOST_REGEX_NO_EXTERNAL_TEMPLATES - -namespace cutl_details_boost{ - -// -// this header can be included multiple times, each time with -// a different character type, BOOST_REGEX_CHAR_T must be defined -// first: -// -#ifndef BOOST_REGEX_CHAR_T -# error "BOOST_REGEX_CHAR_T not defined" -#endif - -#ifndef BOOST_REGEX_TRAITS_T -# define BOOST_REGEX_TRAITS_T , cutl_details_boost::regex_traits -#endif - -// -// what follows is compiler specific: -// - -#if defined(__BORLANDC__) && (__BORLANDC__ < 0x600) - -#ifdef BOOST_HAS_ABI_HEADERS -# include BOOST_ABI_PREFIX -#endif - -# ifndef BOOST_REGEX_INSTANTIATE -# pragma option push -Jgx -# endif - -template class BOOST_REGEX_DECL basic_regex< BOOST_REGEX_CHAR_T BOOST_REGEX_TRAITS_T >; -template class BOOST_REGEX_DECL match_results< const BOOST_REGEX_CHAR_T* >; -#ifndef BOOST_NO_STD_ALLOCATOR -template class BOOST_REGEX_DECL ::cutl_details_boost::re_detail::perl_matcher::allocator_type BOOST_REGEX_TRAITS_T >; -#endif - -# ifndef BOOST_REGEX_INSTANTIATE -# pragma option pop -# endif - -#ifdef BOOST_HAS_ABI_HEADERS -# include BOOST_ABI_SUFFIX -#endif - -#elif defined(BOOST_MSVC) || defined(__ICL) - -# ifndef BOOST_REGEX_INSTANTIATE -# ifdef __GNUC__ -# define template __extension__ extern template -# else -# if BOOST_MSVC > 1310 -# define BOOST_REGEX_TEMPLATE_DECL -# endif -# define template extern template -# endif -# endif - -#ifndef BOOST_REGEX_TEMPLATE_DECL -# define BOOST_REGEX_TEMPLATE_DECL BOOST_REGEX_DECL -#endif - -# ifdef BOOST_MSVC -# pragma warning(push) -# pragma warning(disable : 4251 4231) -# if BOOST_MSVC < 1600 -# pragma warning(disable : 4660) -# endif -# endif - -template class BOOST_REGEX_TEMPLATE_DECL basic_regex< BOOST_REGEX_CHAR_T BOOST_REGEX_TRAITS_T >; - -#if !BOOST_WORKAROUND(BOOST_MSVC, < 1300) -template class BOOST_REGEX_TEMPLATE_DECL match_results< const BOOST_REGEX_CHAR_T* >; -#endif -#ifndef BOOST_NO_STD_ALLOCATOR -template class BOOST_REGEX_TEMPLATE_DECL ::cutl_details_boost::re_detail::perl_matcher::allocator_type BOOST_REGEX_TRAITS_T >; -#endif -#if !(defined(BOOST_DINKUMWARE_STDLIB) && (BOOST_DINKUMWARE_STDLIB <= 1))\ - && !(defined(BOOST_INTEL_CXX_VERSION) && (BOOST_INTEL_CXX_VERSION <= 800))\ - && !(defined(__SGI_STL_PORT) || defined(_STLPORT_VERSION))\ - && !defined(BOOST_REGEX_ICU_INSTANCES) -#if !BOOST_WORKAROUND(BOOST_MSVC, < 1300) -template class BOOST_REGEX_TEMPLATE_DECL match_results< std::basic_string::const_iterator >; -#endif -#ifndef BOOST_NO_STD_ALLOCATOR -template class BOOST_REGEX_TEMPLATE_DECL ::cutl_details_boost::re_detail::perl_matcher< std::basic_string::const_iterator, match_results< std::basic_string::const_iterator >::allocator_type, cutl_details_boost::regex_traits >; -#endif -#endif - - -# ifdef BOOST_MSVC -# pragma warning(pop) -# endif - -# ifdef template -# undef template -# endif - -#undef BOOST_REGEX_TEMPLATE_DECL - -#elif (defined(__GNUC__) && (__GNUC__ >= 3)) || !defined(BOOST_NO_CXX11_EXTERN_TEMPLATE) - -# ifndef BOOST_REGEX_INSTANTIATE -# ifdef __GNUC__ -# define template __extension__ extern template -# else -# define template extern template -# endif -# endif - -#if !defined(BOOST_NO_STD_LOCALE) && !defined(BOOST_REGEX_ICU_INSTANCES) -namespace re_detail{ -template BOOST_REGEX_DECL -std::locale cpp_regex_traits_base::imbue(const std::locale& l); - -template BOOST_REGEX_DECL -cpp_regex_traits_implementation::string_type - cpp_regex_traits_implementation::transform_primary(const BOOST_REGEX_CHAR_T* p1, const BOOST_REGEX_CHAR_T* p2) const; -template BOOST_REGEX_DECL -cpp_regex_traits_implementation::string_type - cpp_regex_traits_implementation::transform(const BOOST_REGEX_CHAR_T* p1, const BOOST_REGEX_CHAR_T* p2) const; -template BOOST_REGEX_DECL -cpp_regex_traits_implementation::string_type - cpp_regex_traits_implementation::lookup_collatename(const BOOST_REGEX_CHAR_T* p1, const BOOST_REGEX_CHAR_T* p2) const; -template BOOST_REGEX_DECL -void cpp_regex_traits_implementation::init(); -template BOOST_REGEX_DECL -cpp_regex_traits_implementation::char_class_type - cpp_regex_traits_implementation::lookup_classname_imp(const BOOST_REGEX_CHAR_T* p1, const BOOST_REGEX_CHAR_T* p2) const; -#ifdef BOOST_REGEX_BUGGY_CTYPE_FACET -template BOOST_REGEX_DECL -bool cpp_regex_traits_implementation::isctype(const BOOST_REGEX_CHAR_T c, char_class_type mask) const; -#endif -} // namespace -template BOOST_REGEX_DECL -int cpp_regex_traits::toi(const BOOST_REGEX_CHAR_T*& first, const BOOST_REGEX_CHAR_T* last, int radix)const; -template BOOST_REGEX_DECL -std::string cpp_regex_traits::catalog_name(const std::string& name); -template BOOST_REGEX_DECL -std::string& cpp_regex_traits::get_catalog_name_inst(); -template BOOST_REGEX_DECL -std::string cpp_regex_traits::get_catalog_name(); -#ifdef BOOST_HAS_THREADS -template BOOST_REGEX_DECL -static_mutex& cpp_regex_traits::get_mutex_inst(); -#endif -#endif - -template BOOST_REGEX_DECL basic_regex& - basic_regex::do_assign( - const BOOST_REGEX_CHAR_T* p1, - const BOOST_REGEX_CHAR_T* p2, - flag_type f); -template BOOST_REGEX_DECL basic_regex::locale_type BOOST_REGEX_CALL - basic_regex::imbue(locale_type l); - -template BOOST_REGEX_DECL void BOOST_REGEX_CALL - match_results::maybe_assign( - const match_results& m); - -namespace re_detail{ -template BOOST_REGEX_DECL void perl_matcher::allocator_type BOOST_REGEX_TRAITS_T >::construct_init( - const basic_regex& e, match_flag_type f); -template BOOST_REGEX_DECL bool perl_matcher::allocator_type BOOST_REGEX_TRAITS_T >::match(); -template BOOST_REGEX_DECL bool perl_matcher::allocator_type BOOST_REGEX_TRAITS_T >::find(); -} // namespace - -#if (defined(__GLIBCPP__) || defined(__GLIBCXX__)) \ - && !defined(BOOST_REGEX_ICU_INSTANCES)\ - && !defined(__SGI_STL_PORT)\ - && !defined(_STLPORT_VERSION) -// std:basic_string<>::const_iterator instances as well: -template BOOST_REGEX_DECL void BOOST_REGEX_CALL - match_results::const_iterator>::maybe_assign( - const match_results::const_iterator>& m); - -namespace re_detail{ -template BOOST_REGEX_DECL void perl_matcher::const_iterator, match_results< std::basic_string::const_iterator >::allocator_type, cutl_details_boost::regex_traits >::construct_init( - const basic_regex& e, match_flag_type f); -template BOOST_REGEX_DECL bool perl_matcher::const_iterator, match_results< std::basic_string::const_iterator >::allocator_type, cutl_details_boost::regex_traits >::match(); -template BOOST_REGEX_DECL bool perl_matcher::const_iterator, match_results< std::basic_string::const_iterator >::allocator_type, cutl_details_boost::regex_traits >::find(); -} // namespace -#endif - -# ifdef template -# undef template -# endif - - -#endif - -} // namespace cutl_details_boost - -#endif // BOOST_REGEX_NO_EXTERNAL_TEMPLATES - - - - - diff --git a/cutl/details/boost/regex/v4/iterator_category.hpp b/cutl/details/boost/regex/v4/iterator_category.hpp deleted file mode 100644 index 18b1291..0000000 --- a/cutl/details/boost/regex/v4/iterator_category.hpp +++ /dev/null @@ -1,91 +0,0 @@ -/* - * - * Copyright (c) 2002 - * John Maddock - * - * Use, modification and distribution are subject to the - * Boost Software License, Version 1.0. (See accompanying file - * LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt) - * - */ - - /* - * LOCATION: see http://www.boost.org for most recent version. - * FILE regex_match.hpp - * VERSION see - * DESCRIPTION: Iterator traits for selecting an iterator type as - * an integral constant expression. - */ - - -#ifndef BOOST_REGEX_ITERATOR_CATEGORY_HPP -#define BOOST_REGEX_ITERATOR_CATEGORY_HPP - -#include -#include -#include - -namespace cutl_details_boost{ -namespace detail{ - -template -struct is_random_imp -{ -#ifndef BOOST_NO_STD_ITERATOR_TRAITS -private: - typedef typename std::iterator_traits::iterator_category cat; -public: - BOOST_STATIC_CONSTANT(bool, value = (::cutl_details_boost::is_convertible::value)); -#else - BOOST_STATIC_CONSTANT(bool, value = false); -#endif -}; - -template -struct is_random_pointer_imp -{ - BOOST_STATIC_CONSTANT(bool, value = true); -}; - -template -struct is_random_imp_selector -{ - template - struct rebind - { - typedef is_random_imp type; - }; -}; - -template <> -struct is_random_imp_selector -{ - template - struct rebind - { - typedef is_random_pointer_imp type; - }; -}; - -} - -template -struct is_random_access_iterator -{ -private: - typedef detail::is_random_imp_selector< ::cutl_details_boost::is_pointer::value> selector; - typedef typename selector::template rebind bound_type; - typedef typename bound_type::type answer; -public: - BOOST_STATIC_CONSTANT(bool, value = answer::value); -}; - -#ifndef BOOST_NO_INCLASS_MEMBER_INITIALIZATION -template -const bool is_random_access_iterator::value; -#endif - -} - -#endif - diff --git a/cutl/details/boost/regex/v4/iterator_traits.hpp b/cutl/details/boost/regex/v4/iterator_traits.hpp deleted file mode 100644 index 146db8a..0000000 --- a/cutl/details/boost/regex/v4/iterator_traits.hpp +++ /dev/null @@ -1,135 +0,0 @@ -/* - * - * Copyright (c) 1998-2002 - * John Maddock - * - * Use, modification and distribution are subject to the - * Boost Software License, Version 1.0. (See accompanying file - * LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt) - * - */ - - /* - * LOCATION: see http://www.boost.org for most recent version. - * FILE iterator_traits.cpp - * VERSION see - * DESCRIPTION: Declares iterator traits workarounds. - */ - -#ifndef BOOST_REGEX_V4_ITERATOR_TRAITS_HPP -#define BOOST_REGEX_V4_ITERATOR_TRAITS_HPP - -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable: 4103) -#endif -#ifdef BOOST_HAS_ABI_HEADERS -# include BOOST_ABI_PREFIX -#endif -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif - -namespace cutl_details_boost{ -namespace re_detail{ - -#if defined(BOOST_NO_STD_ITERATOR_TRAITS) || defined(BOOST_NO_TEMPLATE_PARTIAL_SPECIALIZATION) - -template -struct regex_iterator_traits -{ - typedef typename T::iterator_category iterator_category; - typedef typename T::value_type value_type; -#if !defined(BOOST_NO_STD_ITERATOR) - typedef typename T::difference_type difference_type; - typedef typename T::pointer pointer; - typedef typename T::reference reference; -#else - typedef std::ptrdiff_t difference_type; - typedef value_type* pointer; - typedef value_type& reference; -#endif -}; - -template -struct pointer_iterator_traits -{ - typedef std::ptrdiff_t difference_type; - typedef T value_type; - typedef T* pointer; - typedef T& reference; - typedef std::random_access_iterator_tag iterator_category; -}; -template -struct const_pointer_iterator_traits -{ - typedef std::ptrdiff_t difference_type; - typedef T value_type; - typedef const T* pointer; - typedef const T& reference; - typedef std::random_access_iterator_tag iterator_category; -}; - -template<> -struct regex_iterator_traits : pointer_iterator_traits{}; -template<> -struct regex_iterator_traits : const_pointer_iterator_traits{}; -template<> -struct regex_iterator_traits : pointer_iterator_traits{}; -template<> -struct regex_iterator_traits : const_pointer_iterator_traits{}; -// -// the follwoing are needed for ICU support: -// -template<> -struct regex_iterator_traits : pointer_iterator_traits{}; -template<> -struct regex_iterator_traits : const_pointer_iterator_traits{}; -template<> -struct regex_iterator_traits : pointer_iterator_traits{}; -template<> -struct regex_iterator_traits : const_pointer_iterator_traits{}; - -#ifdef BOOST_REGEX_HAS_OTHER_WCHAR_T -template<> -struct regex_iterator_traits : pointer_iterator_traits{}; -template<> -struct regex_iterator_traits : const_pointer_iterator_traits{}; -#endif - -#if defined(__SGI_STL_PORT) && defined(__STL_DEBUG) -template<> -struct regex_iterator_traits : pointer_iterator_traits{}; -template<> -struct regex_iterator_traits : const_pointer_iterator_traits{}; -#ifndef BOOST_NO_STD_WSTRING -template<> -struct regex_iterator_traits : pointer_iterator_traits{}; -template<> -struct regex_iterator_traits : const_pointer_iterator_traits{}; -#endif // BOOST_NO_WSTRING -#endif // stport - -#else - -template -struct regex_iterator_traits : public std::iterator_traits {}; - -#endif - -} // namespace re_detail -} // namespace cutl_details_boost - -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable: 4103) -#endif -#ifdef BOOST_HAS_ABI_HEADERS -# include BOOST_ABI_SUFFIX -#endif -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif - -#endif - diff --git a/cutl/details/boost/regex/v4/match_flags.hpp b/cutl/details/boost/regex/v4/match_flags.hpp deleted file mode 100644 index ac6fe21..0000000 --- a/cutl/details/boost/regex/v4/match_flags.hpp +++ /dev/null @@ -1,138 +0,0 @@ -/* - * - * Copyright (c) 1998-2002 - * John Maddock - * - * Use, modification and distribution are subject to the - * Boost Software License, Version 1.0. (See accompanying file - * LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt) - * - */ - - /* - * LOCATION: see http://www.boost.org for most recent version. - * FILE match_flags.hpp - * VERSION see - * DESCRIPTION: Declares match_flags type. - */ - -#ifndef BOOST_REGEX_V4_MATCH_FLAGS -#define BOOST_REGEX_V4_MATCH_FLAGS - -#ifdef __cplusplus -# include -#endif - -#ifdef __cplusplus -namespace cutl_details_boost{ - namespace regex_constants{ -#endif - -typedef enum _match_flags -{ - match_default = 0, - match_not_bol = 1, /* first is not start of line */ - match_not_eol = match_not_bol << 1, /* last is not end of line */ - match_not_bob = match_not_eol << 1, /* first is not start of buffer */ - match_not_eob = match_not_bob << 1, /* last is not end of buffer */ - match_not_bow = match_not_eob << 1, /* first is not start of word */ - match_not_eow = match_not_bow << 1, /* last is not end of word */ - match_not_dot_newline = match_not_eow << 1, /* \n is not matched by '.' */ - match_not_dot_null = match_not_dot_newline << 1, /* '\0' is not matched by '.' */ - match_prev_avail = match_not_dot_null << 1, /* *--first is a valid expression */ - match_init = match_prev_avail << 1, /* internal use */ - match_any = match_init << 1, /* don't care what we match */ - match_not_null = match_any << 1, /* string can't be null */ - match_continuous = match_not_null << 1, /* each grep match must continue from */ - /* uninterupted from the previous one */ - match_partial = match_continuous << 1, /* find partial matches */ - - match_stop = match_partial << 1, /* stop after first match (grep) V3 only */ - match_not_initial_null = match_stop, /* don't match initial null, V4 only */ - match_all = match_stop << 1, /* must find the whole of input even if match_any is set */ - match_perl = match_all << 1, /* Use perl matching rules */ - match_posix = match_perl << 1, /* Use POSIX matching rules */ - match_nosubs = match_posix << 1, /* don't trap marked subs */ - match_extra = match_nosubs << 1, /* include full capture information for repeated captures */ - match_single_line = match_extra << 1, /* treat text as single line and ignor any \n's when matching ^ and $. */ - match_unused1 = match_single_line << 1, /* unused */ - match_unused2 = match_unused1 << 1, /* unused */ - match_unused3 = match_unused2 << 1, /* unused */ - match_max = match_unused3, - - format_perl = 0, /* perl style replacement */ - format_default = 0, /* ditto. */ - format_sed = match_max << 1, /* sed style replacement. */ - format_all = format_sed << 1, /* enable all extentions to sytax. */ - format_no_copy = format_all << 1, /* don't copy non-matching segments. */ - format_first_only = format_no_copy << 1, /* Only replace first occurance. */ - format_is_if = format_first_only << 1, /* internal use only. */ - format_literal = format_is_if << 1 /* treat string as a literal */ - -} match_flags; - -#if (defined(_MSC_VER) && (_MSC_VER < 1300)) || defined(__BORLANDC__) -typedef unsigned long match_flag_type; -#else -typedef match_flags match_flag_type; - - -#ifdef __cplusplus -inline match_flags operator&(match_flags m1, match_flags m2) -{ return static_cast(static_cast(m1) & static_cast(m2)); } -inline match_flags operator|(match_flags m1, match_flags m2) -{ return static_cast(static_cast(m1) | static_cast(m2)); } -inline match_flags operator^(match_flags m1, match_flags m2) -{ return static_cast(static_cast(m1) ^ static_cast(m2)); } -inline match_flags operator~(match_flags m1) -{ return static_cast(~static_cast(m1)); } -inline match_flags& operator&=(match_flags& m1, match_flags m2) -{ m1 = m1&m2; return m1; } -inline match_flags& operator|=(match_flags& m1, match_flags m2) -{ m1 = m1|m2; return m1; } -inline match_flags& operator^=(match_flags& m1, match_flags m2) -{ m1 = m1^m2; return m1; } -#endif -#endif - -#ifdef __cplusplus -} /* namespace regex_constants */ -/* - * import names into boost for backwards compatiblity: - */ -using regex_constants::match_flag_type; -using regex_constants::match_default; -using regex_constants::match_not_bol; -using regex_constants::match_not_eol; -using regex_constants::match_not_bob; -using regex_constants::match_not_eob; -using regex_constants::match_not_bow; -using regex_constants::match_not_eow; -using regex_constants::match_not_dot_newline; -using regex_constants::match_not_dot_null; -using regex_constants::match_prev_avail; -/* using regex_constants::match_init; */ -using regex_constants::match_any; -using regex_constants::match_not_null; -using regex_constants::match_continuous; -using regex_constants::match_partial; -/*using regex_constants::match_stop; */ -using regex_constants::match_all; -using regex_constants::match_perl; -using regex_constants::match_posix; -using regex_constants::match_nosubs; -using regex_constants::match_extra; -using regex_constants::match_single_line; -/*using regex_constants::match_max; */ -using regex_constants::format_all; -using regex_constants::format_sed; -using regex_constants::format_perl; -using regex_constants::format_default; -using regex_constants::format_no_copy; -using regex_constants::format_first_only; -/*using regex_constants::format_is_if;*/ - -} /* namespace cutl_details_boost */ -#endif /* __cplusplus */ -#endif /* include guard */ - diff --git a/cutl/details/boost/regex/v4/match_results.hpp b/cutl/details/boost/regex/v4/match_results.hpp deleted file mode 100644 index 8046cfd..0000000 --- a/cutl/details/boost/regex/v4/match_results.hpp +++ /dev/null @@ -1,702 +0,0 @@ -/* - * - * Copyright (c) 1998-2009 - * John Maddock - * - * Use, modification and distribution are subject to the - * Boost Software License, Version 1.0. (See accompanying file - * LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt) - * - */ - - /* - * LOCATION: see http://www.boost.org for most recent version. - * FILE match_results.cpp - * VERSION see - * DESCRIPTION: Declares template class match_results. - */ - -#ifndef BOOST_REGEX_V4_MATCH_RESULTS_HPP -#define BOOST_REGEX_V4_MATCH_RESULTS_HPP - -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable: 4103) -#endif -#ifdef BOOST_HAS_ABI_HEADERS -# include BOOST_ABI_PREFIX -#endif -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif - -namespace cutl_details_boost{ -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable : 4251 4231) -# if BOOST_MSVC < 1600 -# pragma warning(disable : 4660) -# endif -#endif - -namespace re_detail{ - -class named_subexpressions; - -} - -template -class match_results -{ -private: -#ifndef BOOST_NO_STD_ALLOCATOR - typedef std::vector, Allocator> vector_type; -#else - typedef std::vector > vector_type; -#endif -public: - typedef sub_match value_type; -#if !defined(BOOST_NO_STD_ALLOCATOR) && !(defined(BOOST_MSVC) && defined(_STLPORT_VERSION)) - typedef typename Allocator::const_reference const_reference; -#else - typedef const value_type& const_reference; -#endif - typedef const_reference reference; - typedef typename vector_type::const_iterator const_iterator; - typedef const_iterator iterator; - typedef typename re_detail::regex_iterator_traits< - BidiIterator>::difference_type difference_type; - typedef typename Allocator::size_type size_type; - typedef Allocator allocator_type; - typedef typename re_detail::regex_iterator_traits< - BidiIterator>::value_type char_type; - typedef std::basic_string string_type; - typedef re_detail::named_subexpressions named_sub_type; - - // construct/copy/destroy: - explicit match_results(const Allocator& a = Allocator()) -#ifndef BOOST_NO_STD_ALLOCATOR - : m_subs(a), m_base(), m_last_closed_paren(0), m_is_singular(true) {} -#else - : m_subs(), m_base(), m_last_closed_paren(0), m_is_singular(true) { (void)a; } -#endif - match_results(const match_results& m) - : m_subs(m.m_subs), m_named_subs(m.m_named_subs), m_last_closed_paren(m.m_last_closed_paren), m_is_singular(m.m_is_singular) - { - if(!m_is_singular) - { - m_base = m.m_base; - m_null = m.m_null; - } - } - match_results& operator=(const match_results& m) - { - m_subs = m.m_subs; - m_named_subs = m.m_named_subs; - m_last_closed_paren = m.m_last_closed_paren; - m_is_singular = m.m_is_singular; - if(!m_is_singular) - { - m_base = m.m_base; - m_null = m.m_null; - } - return *this; - } - ~match_results(){} - - // size: - size_type size() const - { return empty() ? 0 : m_subs.size() - 2; } - size_type max_size() const - { return m_subs.max_size(); } - bool empty() const - { return m_subs.size() < 2; } - // element access: - difference_type length(int sub = 0) const - { - if(m_is_singular) - raise_logic_error(); - sub += 2; - if((sub < (int)m_subs.size()) && (sub > 0)) - return m_subs[sub].length(); - return 0; - } - difference_type length(const char_type* sub) const - { - if(m_is_singular) - raise_logic_error(); - const char_type* sub_end = sub; - while(*sub_end) ++sub_end; - return length(named_subexpression_index(sub, sub_end)); - } - template - difference_type length(const charT* sub) const - { - if(m_is_singular) - raise_logic_error(); - const charT* sub_end = sub; - while(*sub_end) ++sub_end; - return length(named_subexpression_index(sub, sub_end)); - } - template - difference_type length(const std::basic_string& sub) const - { - return length(sub.c_str()); - } - difference_type position(size_type sub = 0) const - { - if(m_is_singular) - raise_logic_error(); - sub += 2; - if(sub < m_subs.size()) - { - const sub_match& s = m_subs[sub]; - if(s.matched || (sub == 2)) - { - return ::cutl_details_boost::re_detail::distance((BidiIterator)(m_base), (BidiIterator)(s.first)); - } - } - return ~static_cast(0); - } - difference_type position(const char_type* sub) const - { - const char_type* sub_end = sub; - while(*sub_end) ++sub_end; - return position(named_subexpression_index(sub, sub_end)); - } - template - difference_type position(const charT* sub) const - { - const charT* sub_end = sub; - while(*sub_end) ++sub_end; - return position(named_subexpression_index(sub, sub_end)); - } - template - difference_type position(const std::basic_string& sub) const - { - return position(sub.c_str()); - } - string_type str(int sub = 0) const - { - if(m_is_singular) - raise_logic_error(); - sub += 2; - string_type result; - if(sub < (int)m_subs.size() && (sub > 0)) - { - const sub_match& s = m_subs[sub]; - if(s.matched) - { - result = s.str(); - } - } - return result; - } - string_type str(const char_type* sub) const - { - return (*this)[sub].str(); - } - template - string_type str(const std::basic_string& sub) const - { - return (*this)[sub].str(); - } - template - string_type str(const charT* sub) const - { - return (*this)[sub].str(); - } - template - string_type str(const std::basic_string& sub) const - { - return (*this)[sub].str(); - } - const_reference operator[](int sub) const - { - if(m_is_singular && m_subs.empty()) - raise_logic_error(); - sub += 2; - if(sub < (int)m_subs.size() && (sub >= 0)) - { - return m_subs[sub]; - } - return m_null; - } - // - // Named sub-expressions: - // - const_reference named_subexpression(const char_type* i, const char_type* j) const - { - // - // Scan for the leftmost *matched* subexpression with the specified named: - // - if(m_is_singular) - raise_logic_error(); - re_detail::named_subexpressions::range_type r = m_named_subs->equal_range(i, j); - while((r.first != r.second) && ((*this)[r.first->index].matched == false)) - ++r.first; - return r.first != r.second ? (*this)[r.first->index] : m_null; - } - template - const_reference named_subexpression(const charT* i, const charT* j) const - { - BOOST_STATIC_ASSERT(sizeof(charT) <= sizeof(char_type)); - if(i == j) - return m_null; - std::vector s; - while(i != j) - s.insert(s.end(), *i++); - return named_subexpression(&*s.begin(), &*s.begin() + s.size()); - } - int named_subexpression_index(const char_type* i, const char_type* j) const - { - // - // Scan for the leftmost *matched* subexpression with the specified named. - // If none found then return the leftmost expression with that name, - // otherwise an invalid index: - // - if(m_is_singular) - raise_logic_error(); - re_detail::named_subexpressions::range_type s, r; - s = r = m_named_subs->equal_range(i, j); - while((r.first != r.second) && ((*this)[r.first->index].matched == false)) - ++r.first; - if(r.first == r.second) - r = s; - return r.first != r.second ? r.first->index : -20; - } - template - int named_subexpression_index(const charT* i, const charT* j) const - { - BOOST_STATIC_ASSERT(sizeof(charT) <= sizeof(char_type)); - if(i == j) - return -20; - std::vector s; - while(i != j) - s.insert(s.end(), *i++); - return named_subexpression_index(&*s.begin(), &*s.begin() + s.size()); - } - template - const_reference operator[](const std::basic_string& s) const - { - return named_subexpression(s.c_str(), s.c_str() + s.size()); - } - const_reference operator[](const char_type* p) const - { - const char_type* e = p; - while(*e) ++e; - return named_subexpression(p, e); - } - - template - const_reference operator[](const charT* p) const - { - BOOST_STATIC_ASSERT(sizeof(charT) <= sizeof(char_type)); - if(*p == 0) - return m_null; - std::vector s; - while(*p) - s.insert(s.end(), *p++); - return named_subexpression(&*s.begin(), &*s.begin() + s.size()); - } - template - const_reference operator[](const std::basic_string& ns) const - { - BOOST_STATIC_ASSERT(sizeof(charT) <= sizeof(char_type)); - if(ns.empty()) - return m_null; - std::vector s; - for(unsigned i = 0; i < ns.size(); ++i) - s.insert(s.end(), ns[i]); - return named_subexpression(&*s.begin(), &*s.begin() + s.size()); - } - - const_reference prefix() const - { - if(m_is_singular) - raise_logic_error(); - return (*this)[-1]; - } - - const_reference suffix() const - { - if(m_is_singular) - raise_logic_error(); - return (*this)[-2]; - } - const_iterator begin() const - { - return (m_subs.size() > 2) ? (m_subs.begin() + 2) : m_subs.end(); - } - const_iterator end() const - { - return m_subs.end(); - } - // format: - template - OutputIterator format(OutputIterator out, - Functor fmt, - match_flag_type flags = format_default) const - { - if(m_is_singular) - raise_logic_error(); - typedef typename re_detail::compute_functor_type, OutputIterator>::type F; - F func(fmt); - return func(*this, out, flags); - } - template - string_type format(Functor fmt, match_flag_type flags = format_default) const - { - if(m_is_singular) - raise_logic_error(); - std::basic_string result; - re_detail::string_out_iterator > i(result); - - typedef typename re_detail::compute_functor_type, re_detail::string_out_iterator > >::type F; - F func(fmt); - - func(*this, i, flags); - return result; - } - // format with locale: - template - OutputIterator format(OutputIterator out, - Functor fmt, - match_flag_type flags, - const RegexT& re) const - { - if(m_is_singular) - raise_logic_error(); - typedef ::cutl_details_boost::regex_traits_wrapper traits_type; - typedef typename re_detail::compute_functor_type, OutputIterator, traits_type>::type F; - F func(fmt); - return func(*this, out, flags, re.get_traits()); - } - template - string_type format(Functor fmt, - match_flag_type flags, - const RegexT& re) const - { - if(m_is_singular) - raise_logic_error(); - typedef ::cutl_details_boost::regex_traits_wrapper traits_type; - std::basic_string result; - re_detail::string_out_iterator > i(result); - - typedef typename re_detail::compute_functor_type, re_detail::string_out_iterator >, traits_type >::type F; - F func(fmt); - - func(*this, i, flags, re.get_traits()); - return result; - } - - const_reference get_last_closed_paren()const - { - if(m_is_singular) - raise_logic_error(); - return m_last_closed_paren == 0 ? m_null : (*this)[m_last_closed_paren]; - } - - allocator_type get_allocator() const - { -#ifndef BOOST_NO_STD_ALLOCATOR - return m_subs.get_allocator(); -#else - return allocator_type(); -#endif - } - void swap(match_results& that) - { - std::swap(m_subs, that.m_subs); - std::swap(m_named_subs, that.m_named_subs); - std::swap(m_last_closed_paren, that.m_last_closed_paren); - if(m_is_singular) - { - if(!that.m_is_singular) - { - m_base = that.m_base; - m_null = that.m_null; - } - } - else if(that.m_is_singular) - { - that.m_base = m_base; - that.m_null = m_null; - } - else - { - std::swap(m_base, that.m_base); - std::swap(m_null, that.m_null); - } - std::swap(m_is_singular, that.m_is_singular); - } - bool operator==(const match_results& that)const - { - if(m_is_singular) - { - return that.m_is_singular; - } - else if(that.m_is_singular) - { - return false; - } - return (m_subs == that.m_subs) && (m_base == that.m_base) && (m_last_closed_paren == that.m_last_closed_paren); - } - bool operator!=(const match_results& that)const - { return !(*this == that); } - -#ifdef BOOST_REGEX_MATCH_EXTRA - typedef typename sub_match::capture_sequence_type capture_sequence_type; - - const capture_sequence_type& captures(int i)const - { - if(m_is_singular) - raise_logic_error(); - return (*this)[i].captures(); - } -#endif - - // - // private access functions: - void BOOST_REGEX_CALL set_second(BidiIterator i) - { - BOOST_ASSERT(m_subs.size() > 2); - m_subs[2].second = i; - m_subs[2].matched = true; - m_subs[0].first = i; - m_subs[0].matched = (m_subs[0].first != m_subs[0].second); - m_null.first = i; - m_null.second = i; - m_null.matched = false; - m_is_singular = false; - } - - void BOOST_REGEX_CALL set_second(BidiIterator i, size_type pos, bool m = true, bool escape_k = false) - { - if(pos) - m_last_closed_paren = static_cast(pos); - pos += 2; - BOOST_ASSERT(m_subs.size() > pos); - m_subs[pos].second = i; - m_subs[pos].matched = m; - if((pos == 2) && !escape_k) - { - m_subs[0].first = i; - m_subs[0].matched = (m_subs[0].first != m_subs[0].second); - m_null.first = i; - m_null.second = i; - m_null.matched = false; - m_is_singular = false; - } - } - void BOOST_REGEX_CALL set_size(size_type n, BidiIterator i, BidiIterator j) - { - value_type v(j); - size_type len = m_subs.size(); - if(len > n + 2) - { - m_subs.erase(m_subs.begin()+n+2, m_subs.end()); - std::fill(m_subs.begin(), m_subs.end(), v); - } - else - { - std::fill(m_subs.begin(), m_subs.end(), v); - if(n+2 != len) - m_subs.insert(m_subs.end(), n+2-len, v); - } - m_subs[1].first = i; - m_last_closed_paren = 0; - } - void BOOST_REGEX_CALL set_base(BidiIterator pos) - { - m_base = pos; - } - BidiIterator base()const - { - return m_base; - } - void BOOST_REGEX_CALL set_first(BidiIterator i) - { - BOOST_ASSERT(m_subs.size() > 2); - // set up prefix: - m_subs[1].second = i; - m_subs[1].matched = (m_subs[1].first != i); - // set up $0: - m_subs[2].first = i; - // zero out everything else: - for(size_type n = 3; n < m_subs.size(); ++n) - { - m_subs[n].first = m_subs[n].second = m_subs[0].second; - m_subs[n].matched = false; - } - } - void BOOST_REGEX_CALL set_first(BidiIterator i, size_type pos, bool escape_k = false) - { - BOOST_ASSERT(pos+2 < m_subs.size()); - if(pos || escape_k) - { - m_subs[pos+2].first = i; - if(escape_k) - { - m_subs[1].second = i; - m_subs[1].matched = (m_subs[1].first != m_subs[1].second); - } - } - else - set_first(i); - } - void BOOST_REGEX_CALL maybe_assign(const match_results& m); - - void BOOST_REGEX_CALL set_named_subs(cutl_details_boost::shared_ptr subs) - { - m_named_subs = subs; - } - -private: - // - // Error handler called when an uninitialized match_results is accessed: - // - static void raise_logic_error() - { - std::logic_error e("Attempt to access an uninitialzed cutl_details_boost::match_results<> class."); - cutl_details_boost::throw_exception(e); - } - - - vector_type m_subs; // subexpressions - BidiIterator m_base; // where the search started from - sub_match m_null; // a null match - cutl_details_boost::shared_ptr m_named_subs; // Shared copy of named subs in the regex object - int m_last_closed_paren; // Last ) to be seen - used for formatting - bool m_is_singular; // True if our stored iterators are singular -}; - -template -void BOOST_REGEX_CALL match_results::maybe_assign(const match_results& m) -{ - if(m_is_singular) - { - *this = m; - return; - } - const_iterator p1, p2; - p1 = begin(); - p2 = m.begin(); - // - // Distances are measured from the start of *this* match, unless this isn't - // a valid match in which case we use the start of the whole sequence. Note that - // no subsequent match-candidate can ever be to the left of the first match found. - // This ensures that when we are using bidirectional iterators, that distances - // measured are as short as possible, and therefore as efficient as possible - // to compute. Finally note that we don't use the "matched" data member to test - // whether a sub-expression is a valid match, because partial matches set this - // to false for sub-expression 0. - // - BidiIterator l_end = this->suffix().second; - BidiIterator l_base = (p1->first == l_end) ? this->prefix().first : (*this)[0].first; - difference_type len1 = 0; - difference_type len2 = 0; - difference_type base1 = 0; - difference_type base2 = 0; - std::size_t i; - for(i = 0; i < size(); ++i, ++p1, ++p2) - { - // - // Leftmost takes priority over longest; handle special cases - // where distances need not be computed first (an optimisation - // for bidirectional iterators: ensure that we don't accidently - // compute the length of the whole sequence, as this can be really - // expensive). - // - if(p1->first == l_end) - { - if(p2->first != l_end) - { - // p2 must be better than p1, and no need to calculate - // actual distances: - base1 = 1; - base2 = 0; - break; - } - else - { - // *p1 and *p2 are either unmatched or match end-of sequence, - // either way no need to calculate distances: - if((p1->matched == false) && (p2->matched == true)) - break; - if((p1->matched == true) && (p2->matched == false)) - return; - continue; - } - } - else if(p2->first == l_end) - { - // p1 better than p2, and no need to calculate distances: - return; - } - base1 = ::cutl_details_boost::re_detail::distance(l_base, p1->first); - base2 = ::cutl_details_boost::re_detail::distance(l_base, p2->first); - BOOST_ASSERT(base1 >= 0); - BOOST_ASSERT(base2 >= 0); - if(base1 < base2) return; - if(base2 < base1) break; - - len1 = ::cutl_details_boost::re_detail::distance((BidiIterator)p1->first, (BidiIterator)p1->second); - len2 = ::cutl_details_boost::re_detail::distance((BidiIterator)p2->first, (BidiIterator)p2->second); - BOOST_ASSERT(len1 >= 0); - BOOST_ASSERT(len2 >= 0); - if((len1 != len2) || ((p1->matched == false) && (p2->matched == true))) - break; - if((p1->matched == true) && (p2->matched == false)) - return; - } - if(i == size()) - return; - if(base2 < base1) - *this = m; - else if((len2 > len1) || ((p1->matched == false) && (p2->matched == true)) ) - *this = m; -} - -template -void swap(match_results& a, match_results& b) -{ - a.swap(b); -} - -#ifndef BOOST_NO_STD_LOCALE -template -std::basic_ostream& - operator << (std::basic_ostream& os, - const match_results& s) -{ - return (os << s.str()); -} -#else -template -std::ostream& operator << (std::ostream& os, - const match_results& s) -{ - return (os << s.str()); -} -#endif - -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif -} // namespace cutl_details_boost - -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable: 4103) -#endif -#ifdef BOOST_HAS_ABI_HEADERS -# include BOOST_ABI_SUFFIX -#endif -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif - -#endif - - diff --git a/cutl/details/boost/regex/v4/mem_block_cache.hpp b/cutl/details/boost/regex/v4/mem_block_cache.hpp deleted file mode 100644 index 98e7e8a..0000000 --- a/cutl/details/boost/regex/v4/mem_block_cache.hpp +++ /dev/null @@ -1,99 +0,0 @@ - /* - * Copyright (c) 2002 - * John Maddock - * - * Use, modification and distribution are subject to the - * Boost Software License, Version 1.0. (See accompanying file - * LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt) - * - */ - - /* - * LOCATION: see http://www.boost.org for most recent version. - * FILE mem_block_cache.hpp - * VERSION see - * DESCRIPTION: memory block cache used by the non-recursive matcher. - */ - -#ifndef BOOST_REGEX_V4_MEM_BLOCK_CACHE_HPP -#define BOOST_REGEX_V4_MEM_BLOCK_CACHE_HPP - -#include -#ifdef BOOST_HAS_THREADS -#include -#endif - -#ifdef BOOST_HAS_ABI_HEADERS -# include BOOST_ABI_PREFIX -#endif - -namespace cutl_details_boost{ -namespace re_detail{ - -struct mem_block_node -{ - mem_block_node* next; -}; - -struct mem_block_cache -{ - // this member has to be statically initialsed: - mem_block_node* next; - unsigned cached_blocks; -#ifdef BOOST_HAS_THREADS - cutl_details_boost::static_mutex mut; -#endif - - ~mem_block_cache() - { - while(next) - { - mem_block_node* old = next; - next = next->next; - ::operator delete(old); - } - } - void* get() - { -#ifdef BOOST_HAS_THREADS - cutl_details_boost::static_mutex::scoped_lock g(mut); -#endif - if(next) - { - mem_block_node* result = next; - next = next->next; - --cached_blocks; - return result; - } - return ::operator new(BOOST_REGEX_BLOCKSIZE); - } - void put(void* p) - { -#ifdef BOOST_HAS_THREADS - cutl_details_boost::static_mutex::scoped_lock g(mut); -#endif - if(cached_blocks >= BOOST_REGEX_MAX_CACHE_BLOCKS) - { - ::operator delete(p); - } - else - { - mem_block_node* old = static_cast(p); - old->next = next; - next = old; - ++cached_blocks; - } - } -}; - -extern mem_block_cache block_cache; - -} -} // namespace cutl_details_boost - -#ifdef BOOST_HAS_ABI_HEADERS -# include BOOST_ABI_SUFFIX -#endif - -#endif - diff --git a/cutl/details/boost/regex/v4/perl_matcher.hpp b/cutl/details/boost/regex/v4/perl_matcher.hpp deleted file mode 100644 index 3ef87be..0000000 --- a/cutl/details/boost/regex/v4/perl_matcher.hpp +++ /dev/null @@ -1,587 +0,0 @@ -/* - * - * Copyright (c) 2002 - * John Maddock - * - * Use, modification and distribution are subject to the - * Boost Software License, Version 1.0. (See accompanying file - * LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt) - * - */ - -#ifndef BOOST_REGEX_MATCHER_HPP -#define BOOST_REGEX_MATCHER_HPP - -#include - -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable: 4103) -#endif -#ifdef BOOST_HAS_ABI_HEADERS -# include BOOST_ABI_PREFIX -#endif -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif - -#ifdef BOOST_MSVC -# pragma warning(push) -# pragma warning(disable: 4800) -#endif - -namespace cutl_details_boost{ -namespace re_detail{ - -// -// error checking API: -// -BOOST_REGEX_DECL void BOOST_REGEX_CALL verify_options(cutl_details_boost::regex_constants::syntax_option_type ef, match_flag_type mf); -// -// function can_start: -// -template -inline bool can_start(charT c, const unsigned char* map, unsigned char mask) -{ - return ((c < static_cast(0)) ? true : ((c >= static_cast(1 << CHAR_BIT)) ? true : map[c] & mask)); -} -inline bool can_start(char c, const unsigned char* map, unsigned char mask) -{ - return map[(unsigned char)c] & mask; -} -inline bool can_start(signed char c, const unsigned char* map, unsigned char mask) -{ - return map[(unsigned char)c] & mask; -} -inline bool can_start(unsigned char c, const unsigned char* map, unsigned char mask) -{ - return map[c] & mask; -} -inline bool can_start(unsigned short c, const unsigned char* map, unsigned char mask) -{ - return ((c >= (1 << CHAR_BIT)) ? true : map[c] & mask); -} -#if !defined(__hpux) && !defined(__WINSCW__)// WCHAR_MIN not usable in pp-directives. -#if defined(WCHAR_MIN) && (WCHAR_MIN == 0) && !defined(BOOST_NO_INTRINSIC_WCHAR_T) -inline bool can_start(wchar_t c, const unsigned char* map, unsigned char mask) -{ - return ((c >= static_cast(1u << CHAR_BIT)) ? true : map[c] & mask); -} -#endif -#endif -#if !defined(BOOST_NO_INTRINSIC_WCHAR_T) -inline bool can_start(unsigned int c, const unsigned char* map, unsigned char mask) -{ - return (((c >= static_cast(1u << CHAR_BIT)) ? true : map[c] & mask)); -} -#endif - - -// -// Unfortunately Rogue Waves standard library appears to have a bug -// in std::basic_string::compare that results in eroneous answers -// in some cases (tested with Borland C++ 5.1, Rogue Wave lib version -// 0x020101) the test case was: -// {39135,0} < {0xff,0} -// which succeeds when it should not. -// -#ifndef _RWSTD_VER -#if !BOOST_WORKAROUND(BOOST_MSVC, < 1310) -template -inline int string_compare(const std::basic_string& s, const C* p) -{ - if(0 == *p) - { - if(s.empty() || ((s.size() == 1) && (s[0] == 0))) - return 0; - } - return s.compare(p); -} -#endif -#else -#if !BOOST_WORKAROUND(BOOST_MSVC, < 1310) -template -inline int string_compare(const std::basic_string& s, const C* p) -{ - if(0 == *p) - { - if(s.empty() || ((s.size() == 1) && (s[0] == 0))) - return 0; - } - return s.compare(p); -} -#endif -inline int string_compare(const std::string& s, const char* p) -{ return std::strcmp(s.c_str(), p); } -# ifndef BOOST_NO_WREGEX -inline int string_compare(const std::wstring& s, const wchar_t* p) -{ return std::wcscmp(s.c_str(), p); } -#endif -#endif -template -inline int string_compare(const Seq& s, const C* p) -{ - std::size_t i = 0; - while((i < s.size()) && (p[i] == s[i])) - { - ++i; - } - return (i == s.size()) ? -p[i] : s[i] - p[i]; -} -# define STR_COMP(s,p) string_compare(s,p) - -template -inline const charT* re_skip_past_null(const charT* p) -{ - while (*p != static_cast(0)) ++p; - return ++p; -} - -template -iterator BOOST_REGEX_CALL re_is_set_member(iterator next, - iterator last, - const re_set_long* set_, - const regex_data& e, bool icase) -{ - const charT* p = reinterpret_cast(set_+1); - iterator ptr; - unsigned int i; - //bool icase = e.m_flags & regex_constants::icase; - - if(next == last) return next; - - typedef typename traits_type::string_type traits_string_type; - const ::cutl_details_boost::regex_traits_wrapper& traits_inst = *(e.m_ptraits); - - // dwa 9/13/00 suppress incorrect MSVC warning - it claims this is never - // referenced - (void)traits_inst; - - // try and match a single character, could be a multi-character - // collating element... - for(i = 0; i < set_->csingles; ++i) - { - ptr = next; - if(*p == static_cast(0)) - { - // treat null string as special case: - if(traits_inst.translate(*ptr, icase) != *p) - { - while(*p == static_cast(0))++p; - continue; - } - return set_->isnot ? next : (ptr == next) ? ++next : ptr; - } - else - { - while(*p && (ptr != last)) - { - if(traits_inst.translate(*ptr, icase) != *p) - break; - ++p; - ++ptr; - } - - if(*p == static_cast(0)) // if null we've matched - return set_->isnot ? next : (ptr == next) ? ++next : ptr; - - p = re_skip_past_null(p); // skip null - } - } - - charT col = traits_inst.translate(*next, icase); - - - if(set_->cranges || set_->cequivalents) - { - traits_string_type s1; - // - // try and match a range, NB only a single character can match - if(set_->cranges) - { - if((e.m_flags & regex_constants::collate) == 0) - s1.assign(1, col); - else - { - charT a[2] = { col, charT(0), }; - s1 = traits_inst.transform(a, a + 1); - } - for(i = 0; i < set_->cranges; ++i) - { - if(STR_COMP(s1, p) >= 0) - { - do{ ++p; }while(*p); - ++p; - if(STR_COMP(s1, p) <= 0) - return set_->isnot ? next : ++next; - } - else - { - // skip first string - do{ ++p; }while(*p); - ++p; - } - // skip second string - do{ ++p; }while(*p); - ++p; - } - } - // - // try and match an equivalence class, NB only a single character can match - if(set_->cequivalents) - { - charT a[2] = { col, charT(0), }; - s1 = traits_inst.transform_primary(a, a +1); - for(i = 0; i < set_->cequivalents; ++i) - { - if(STR_COMP(s1, p) == 0) - return set_->isnot ? next : ++next; - // skip string - do{ ++p; }while(*p); - ++p; - } - } - } - if(traits_inst.isctype(col, set_->cclasses) == true) - return set_->isnot ? next : ++next; - if((set_->cnclasses != 0) && (traits_inst.isctype(col, set_->cnclasses) == false)) - return set_->isnot ? next : ++next; - return set_->isnot ? ++next : next; -} - -template -class repeater_count -{ - repeater_count** stack; - repeater_count* next; - int state_id; - std::size_t count; // the number of iterations so far - BidiIterator start_pos; // where the last repeat started -public: - repeater_count(repeater_count** s) - { - stack = s; - next = 0; - state_id = -1; - count = 0; - } - repeater_count(int i, repeater_count** s, BidiIterator start) - : start_pos(start) - { - state_id = i; - stack = s; - next = *stack; - *stack = this; - if(state_id > next->state_id) - count = 0; - else - { - repeater_count* p = next; - while(p && (p->state_id != state_id)) - p = p->next; - if(p) - { - count = p->count; - start_pos = p->start_pos; - } - else - count = 0; - } - } - ~repeater_count() - { - if(next) - *stack = next; - } - std::size_t get_count() { return count; } - int get_id() { return state_id; } - std::size_t operator++() { return ++count; } - bool check_null_repeat(const BidiIterator& pos, std::size_t max) - { - // this is called when we are about to start a new repeat, - // if the last one was NULL move our count to max, - // otherwise save the current position. - bool result = (count == 0) ? false : (pos == start_pos); - if(result) - count = max; - else - start_pos = pos; - return result; - } -}; - -struct saved_state; - -enum saved_state_type -{ - saved_type_end = 0, - saved_type_paren = 1, - saved_type_recurse = 2, - saved_type_assertion = 3, - saved_state_alt = 4, - saved_state_repeater_count = 5, - saved_state_extra_block = 6, - saved_state_greedy_single_repeat = 7, - saved_state_rep_slow_dot = 8, - saved_state_rep_fast_dot = 9, - saved_state_rep_char = 10, - saved_state_rep_short_set = 11, - saved_state_rep_long_set = 12, - saved_state_non_greedy_long_repeat = 13, - saved_state_count = 14 -}; - -template -struct recursion_info -{ - typedef typename Results::value_type value_type; - typedef typename value_type::iterator iterator; - int idx; - const re_syntax_base* preturn_address; - Results results; - repeater_count* repeater_stack; -}; - -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable : 4251 4231) -# if BOOST_MSVC < 1600 -# pragma warning(disable : 4660) -# endif -#endif - -template -class perl_matcher -{ -public: - typedef typename traits::char_type char_type; - typedef perl_matcher self_type; - typedef bool (self_type::*matcher_proc_type)(void); - typedef std::size_t traits_size_type; - typedef typename is_byte::width_type width_type; - typedef typename regex_iterator_traits::difference_type difference_type; - typedef match_results results_type; - - perl_matcher(BidiIterator first, BidiIterator end, - match_results& what, - const basic_regex& e, - match_flag_type f, - BidiIterator l_base) - : m_result(what), base(first), last(end), - position(first), backstop(l_base), re(e), traits_inst(e.get_traits()), - m_independent(false), next_count(&rep_obj), rep_obj(&next_count) - { - construct_init(e, f); - } - - bool match(); - bool find(); - - void setf(match_flag_type f) - { m_match_flags |= f; } - void unsetf(match_flag_type f) - { m_match_flags &= ~f; } - -private: - void construct_init(const basic_regex& e, match_flag_type f); - - bool find_imp(); - bool match_imp(); -#ifdef BOOST_REGEX_HAS_MS_STACK_GUARD - typedef bool (perl_matcher::*protected_proc_type)(); - bool protected_call(protected_proc_type); -#endif - void estimate_max_state_count(std::random_access_iterator_tag*); - void estimate_max_state_count(void*); - bool match_prefix(); - bool match_all_states(); - - // match procs, stored in s_match_vtable: - bool match_startmark(); - bool match_endmark(); - bool match_literal(); - bool match_start_line(); - bool match_end_line(); - bool match_wild(); - bool match_match(); - bool match_word_boundary(); - bool match_within_word(); - bool match_word_start(); - bool match_word_end(); - bool match_buffer_start(); - bool match_buffer_end(); - bool match_backref(); - bool match_long_set(); - bool match_set(); - bool match_jump(); - bool match_alt(); - bool match_rep(); - bool match_combining(); - bool match_soft_buffer_end(); - bool match_restart_continue(); - bool match_long_set_repeat(); - bool match_set_repeat(); - bool match_char_repeat(); - bool match_dot_repeat_fast(); - bool match_dot_repeat_slow(); - bool match_dot_repeat_dispatch() - { - return ::cutl_details_boost::is_random_access_iterator::value ? match_dot_repeat_fast() : match_dot_repeat_slow(); - } - bool match_backstep(); - bool match_assert_backref(); - bool match_toggle_case(); -#ifdef BOOST_REGEX_RECURSIVE - bool backtrack_till_match(std::size_t count); -#endif - bool match_recursion(); - - // find procs stored in s_find_vtable: - bool find_restart_any(); - bool find_restart_word(); - bool find_restart_line(); - bool find_restart_buf(); - bool find_restart_lit(); - -private: - // final result structure to be filled in: - match_results& m_result; - // temporary result for POSIX matches: - scoped_ptr > m_temp_match; - // pointer to actual result structure to fill in: - match_results* m_presult; - // start of sequence being searched: - BidiIterator base; - // end of sequence being searched: - BidiIterator last; - // current character being examined: - BidiIterator position; - // where to restart next search after failed match attempt: - BidiIterator restart; - // where the current search started from, acts as base for $` during grep: - BidiIterator search_base; - // how far we can go back when matching lookbehind: - BidiIterator backstop; - // the expression being examined: - const basic_regex& re; - // the expression's traits class: - const ::cutl_details_boost::regex_traits_wrapper& traits_inst; - // the next state in the machine being matched: - const re_syntax_base* pstate; - // matching flags in use: - match_flag_type m_match_flags; - // how many states we have examined so far: - std::ptrdiff_t state_count; - // max number of states to examine before giving up: - std::ptrdiff_t max_state_count; - // whether we should ignore case or not: - bool icase; - // set to true when (position == last), indicates that we may have a partial match: - bool m_has_partial_match; - // set to true whenever we get a match: - bool m_has_found_match; - // set to true whenever we're inside an independent sub-expression: - bool m_independent; - // the current repeat being examined: - repeater_count* next_count; - // the first repeat being examined (top of linked list): - repeater_count rep_obj; - // the mask to pass when matching word boundaries: - typename traits::char_class_type m_word_mask; - // the bitmask to use when determining whether a match_any matches a newline or not: - unsigned char match_any_mask; - // recursion information: - std::vector > recursion_stack; - -#ifdef BOOST_REGEX_NON_RECURSIVE - // - // additional members for non-recursive version: - // - typedef bool (self_type::*unwind_proc_type)(bool); - - void extend_stack(); - bool unwind(bool); - bool unwind_end(bool); - bool unwind_paren(bool); - bool unwind_recursion_stopper(bool); - bool unwind_assertion(bool); - bool unwind_alt(bool); - bool unwind_repeater_counter(bool); - bool unwind_extra_block(bool); - bool unwind_greedy_single_repeat(bool); - bool unwind_slow_dot_repeat(bool); - bool unwind_fast_dot_repeat(bool); - bool unwind_char_repeat(bool); - bool unwind_short_set_repeat(bool); - bool unwind_long_set_repeat(bool); - bool unwind_non_greedy_repeat(bool); - bool unwind_recursion(bool); - bool unwind_recursion_pop(bool); - void destroy_single_repeat(); - void push_matched_paren(int index, const sub_match& sub); - void push_recursion_stopper(); - void push_assertion(const re_syntax_base* ps, bool positive); - void push_alt(const re_syntax_base* ps); - void push_repeater_count(int i, repeater_count** s); - void push_single_repeat(std::size_t c, const re_repeat* r, BidiIterator last_position, int state_id); - void push_non_greedy_repeat(const re_syntax_base* ps); - void push_recursion(int idx, const re_syntax_base* p, results_type* presults); - void push_recursion_pop(); - - // pointer to base of stack: - saved_state* m_stack_base; - // pointer to current stack position: - saved_state* m_backup_state; - // determines what value to return when unwinding from recursion, - // allows for mixed recursive/non-recursive algorithm: - bool m_recursive_result; - // how many memory blocks have we used up?: - unsigned used_block_count; -#endif - - // these operations aren't allowed, so are declared private, - // bodies are provided to keep explicit-instantiation requests happy: - perl_matcher& operator=(const perl_matcher&) - { - return *this; - } - perl_matcher(const perl_matcher& that) - : m_result(that.m_result), re(that.re), traits_inst(that.traits_inst), rep_obj(0) {} -}; - -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif - -} // namespace re_detail - -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable: 4103) -#endif -#ifdef BOOST_HAS_ABI_HEADERS -# include BOOST_ABI_SUFFIX -#endif -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif - -} // namespace cutl_details_boost - -#ifdef BOOST_MSVC -# pragma warning(pop) -#endif - -// -// include the implementation of perl_matcher: -// -#ifdef BOOST_REGEX_RECURSIVE -#include -#else -#include -#endif -// this one has to be last: -#include - -#endif - diff --git a/cutl/details/boost/regex/v4/perl_matcher_common.hpp b/cutl/details/boost/regex/v4/perl_matcher_common.hpp deleted file mode 100644 index e5dbfd2..0000000 --- a/cutl/details/boost/regex/v4/perl_matcher_common.hpp +++ /dev/null @@ -1,996 +0,0 @@ -/* - * - * Copyright (c) 2002 - * John Maddock - * - * Use, modification and distribution are subject to the - * Boost Software License, Version 1.0. (See accompanying file - * LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt) - * - */ - - /* - * LOCATION: see http://www.boost.org for most recent version. - * FILE perl_matcher_common.cpp - * VERSION see - * DESCRIPTION: Definitions of perl_matcher member functions that are - * common to both the recursive and non-recursive versions. - */ - -#ifndef BOOST_REGEX_V4_PERL_MATCHER_COMMON_HPP -#define BOOST_REGEX_V4_PERL_MATCHER_COMMON_HPP - -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable: 4103) -#endif -#ifdef BOOST_HAS_ABI_HEADERS -# include BOOST_ABI_PREFIX -#endif -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif - -#ifdef __BORLANDC__ -# pragma option push -w-8008 -w-8066 -#endif -#ifdef BOOST_MSVC -# pragma warning(push) -# pragma warning(disable: 4800) -#endif - -namespace cutl_details_boost{ -namespace re_detail{ - -template -void perl_matcher::construct_init(const basic_regex& e, match_flag_type f) -{ - typedef typename regex_iterator_traits::iterator_category category; - typedef typename basic_regex::flag_type expression_flag_type; - - if(e.empty()) - { - // precondition failure: e is not a valid regex. - std::invalid_argument ex("Invalid regular expression object"); - cutl_details_boost::throw_exception(ex); - } - pstate = 0; - m_match_flags = f; - estimate_max_state_count(static_cast(0)); - expression_flag_type re_f = re.flags(); - icase = re_f & regex_constants::icase; - if(!(m_match_flags & (match_perl|match_posix))) - { - if((re_f & (regbase::main_option_type|regbase::no_perl_ex)) == 0) - m_match_flags |= match_perl; - else if((re_f & (regbase::main_option_type|regbase::emacs_ex)) == (regbase::basic_syntax_group|regbase::emacs_ex)) - m_match_flags |= match_perl; - else if((re_f & (regbase::main_option_type|regbase::literal)) == (regbase::literal)) - m_match_flags |= match_perl; - else - m_match_flags |= match_posix; - } - if(m_match_flags & match_posix) - { - m_temp_match.reset(new match_results()); - m_presult = m_temp_match.get(); - } - else - m_presult = &m_result; -#ifdef BOOST_REGEX_NON_RECURSIVE - m_stack_base = 0; - m_backup_state = 0; -#endif - // find the value to use for matching word boundaries: - m_word_mask = re.get_data().m_word_mask; - // find bitmask to use for matching '.': - match_any_mask = static_cast((f & match_not_dot_newline) ? re_detail::test_not_newline : re_detail::test_newline); -} - -template -void perl_matcher::estimate_max_state_count(std::random_access_iterator_tag*) -{ - // - // How many states should we allow our machine to visit before giving up? - // This is a heuristic: it takes the greater of O(N^2) and O(NS^2) - // where N is the length of the string, and S is the number of states - // in the machine. It's tempting to up this to O(N^2S) or even O(N^2S^2) - // but these take unreasonably amounts of time to bale out in pathological - // cases. - // - // Calculate NS^2 first: - // - static const std::ptrdiff_t k = 100000; - std::ptrdiff_t dist = cutl_details_boost::re_detail::distance(base, last); - if(dist == 0) - dist = 1; - std::ptrdiff_t states = re.size(); - if(states == 0) - states = 1; - states *= states; - if((std::numeric_limits::max)() / dist < states) - { - max_state_count = (std::min)((std::ptrdiff_t)BOOST_REGEX_MAX_STATE_COUNT, (std::numeric_limits::max)() - 2); - return; - } - states *= dist; - if((std::numeric_limits::max)() - k < states) - { - max_state_count = (std::min)((std::ptrdiff_t)BOOST_REGEX_MAX_STATE_COUNT, (std::numeric_limits::max)() - 2); - return; - } - states += k; - - max_state_count = states; - - // - // Now calculate N^2: - // - states = dist; - if((std::numeric_limits::max)() / dist < states) - { - max_state_count = (std::min)((std::ptrdiff_t)BOOST_REGEX_MAX_STATE_COUNT, (std::numeric_limits::max)() - 2); - return; - } - states *= dist; - if((std::numeric_limits::max)() - k < states) - { - max_state_count = (std::min)((std::ptrdiff_t)BOOST_REGEX_MAX_STATE_COUNT, (std::numeric_limits::max)() - 2); - return; - } - states += k; - // - // N^2 can be a very large number indeed, to prevent things getting out - // of control, cap the max states: - // - if(states > BOOST_REGEX_MAX_STATE_COUNT) - states = BOOST_REGEX_MAX_STATE_COUNT; - // - // If (the possibly capped) N^2 is larger than our first estimate, - // use this instead: - // - if(states > max_state_count) - max_state_count = states; -} - -template -inline void perl_matcher::estimate_max_state_count(void*) -{ - // we don't know how long the sequence is: - max_state_count = BOOST_REGEX_MAX_STATE_COUNT; -} - -#ifdef BOOST_REGEX_HAS_MS_STACK_GUARD -template -inline bool perl_matcher::protected_call( - protected_proc_type proc) -{ - ::cutl_details_boost::re_detail::concrete_protected_call - > - obj(this, proc); - return obj.execute(); - -} -#endif - -template -inline bool perl_matcher::match() -{ -#ifdef BOOST_REGEX_HAS_MS_STACK_GUARD - return protected_call(&perl_matcher::match_imp); -#else - return match_imp(); -#endif -} - -template -bool perl_matcher::match_imp() -{ - // initialise our stack if we are non-recursive: -#ifdef BOOST_REGEX_NON_RECURSIVE - save_state_init init(&m_stack_base, &m_backup_state); - used_block_count = BOOST_REGEX_MAX_BLOCKS; -#if !defined(BOOST_NO_EXCEPTIONS) - try{ -#endif -#endif - - // reset our state machine: - position = base; - search_base = base; - state_count = 0; - m_match_flags |= regex_constants::match_all; - m_presult->set_size((m_match_flags & match_nosubs) ? 1 : re.mark_count(), search_base, last); - m_presult->set_base(base); - m_presult->set_named_subs(this->re.get_named_subs()); - if(m_match_flags & match_posix) - m_result = *m_presult; - verify_options(re.flags(), m_match_flags); - if(0 == match_prefix()) - return false; - return (m_result[0].second == last) && (m_result[0].first == base); - -#if defined(BOOST_REGEX_NON_RECURSIVE) && !defined(BOOST_NO_EXCEPTIONS) - } - catch(...) - { - // unwind all pushed states, apart from anything else this - // ensures that all the states are correctly destructed - // not just the memory freed. - while(unwind(true)){} - throw; - } -#endif -} - -template -inline bool perl_matcher::find() -{ -#ifdef BOOST_REGEX_HAS_MS_STACK_GUARD - return protected_call(&perl_matcher::find_imp); -#else - return find_imp(); -#endif -} - -template -bool perl_matcher::find_imp() -{ - static matcher_proc_type const s_find_vtable[7] = - { - &perl_matcher::find_restart_any, - &perl_matcher::find_restart_word, - &perl_matcher::find_restart_line, - &perl_matcher::find_restart_buf, - &perl_matcher::match_prefix, - &perl_matcher::find_restart_lit, - &perl_matcher::find_restart_lit, - }; - - // initialise our stack if we are non-recursive: -#ifdef BOOST_REGEX_NON_RECURSIVE - save_state_init init(&m_stack_base, &m_backup_state); - used_block_count = BOOST_REGEX_MAX_BLOCKS; -#if !defined(BOOST_NO_EXCEPTIONS) - try{ -#endif -#endif - - state_count = 0; - if((m_match_flags & regex_constants::match_init) == 0) - { - // reset our state machine: - search_base = position = base; - pstate = re.get_first_state(); - m_presult->set_size((m_match_flags & match_nosubs) ? 1 : re.mark_count(), base, last); - m_presult->set_base(base); - m_presult->set_named_subs(this->re.get_named_subs()); - m_match_flags |= regex_constants::match_init; - } - else - { - // start again: - search_base = position = m_result[0].second; - // If last match was null and match_not_null was not set then increment - // our start position, otherwise we go into an infinite loop: - if(((m_match_flags & match_not_null) == 0) && (m_result.length() == 0)) - { - if(position == last) - return false; - else - ++position; - } - // reset $` start: - m_presult->set_size((m_match_flags & match_nosubs) ? 1 : re.mark_count(), search_base, last); - //if((base != search_base) && (base == backstop)) - // m_match_flags |= match_prev_avail; - } - if(m_match_flags & match_posix) - { - m_result.set_size(re.mark_count(), base, last); - m_result.set_base(base); - } - - verify_options(re.flags(), m_match_flags); - // find out what kind of expression we have: - unsigned type = (m_match_flags & match_continuous) ? - static_cast(regbase::restart_continue) - : static_cast(re.get_restart_type()); - - // call the appropriate search routine: - matcher_proc_type proc = s_find_vtable[type]; - return (this->*proc)(); - -#if defined(BOOST_REGEX_NON_RECURSIVE) && !defined(BOOST_NO_EXCEPTIONS) - } - catch(...) - { - // unwind all pushed states, apart from anything else this - // ensures that all the states are correctly destructed - // not just the memory freed. - while(unwind(true)){} - throw; - } -#endif -} - -template -bool perl_matcher::match_prefix() -{ - m_has_partial_match = false; - m_has_found_match = false; - pstate = re.get_first_state(); - m_presult->set_first(position); - restart = position; - match_all_states(); - if(!m_has_found_match && m_has_partial_match && (m_match_flags & match_partial)) - { - m_has_found_match = true; - m_presult->set_second(last, 0, false); - position = last; - if((m_match_flags & match_posix) == match_posix) - { - m_result.maybe_assign(*m_presult); - } - } -#ifdef BOOST_REGEX_MATCH_EXTRA - if(m_has_found_match && (match_extra & m_match_flags)) - { - // - // we have a match, reverse the capture information: - // - for(unsigned i = 0; i < m_presult->size(); ++i) - { - typename sub_match::capture_sequence_type & seq = ((*m_presult)[i]).get_captures(); - std::reverse(seq.begin(), seq.end()); - } - } -#endif - if(!m_has_found_match) - position = restart; // reset search postion - return m_has_found_match; -} - -template -bool perl_matcher::match_literal() -{ - unsigned int len = static_cast(pstate)->length; - const char_type* what = reinterpret_cast(static_cast(pstate) + 1); - // - // compare string with what we stored in - // our records: - for(unsigned int i = 0; i < len; ++i, ++position) - { - if((position == last) || (traits_inst.translate(*position, icase) != what[i])) - return false; - } - pstate = pstate->next.p; - return true; -} - -template -bool perl_matcher::match_start_line() -{ - if(position == backstop) - { - if((m_match_flags & match_prev_avail) == 0) - { - if((m_match_flags & match_not_bol) == 0) - { - pstate = pstate->next.p; - return true; - } - return false; - } - } - else if(m_match_flags & match_single_line) - return false; - - // check the previous value character: - BidiIterator t(position); - --t; - if(position != last) - { - if(is_separator(*t) && !((*t == static_cast('\r')) && (*position == static_cast('\n'))) ) - { - pstate = pstate->next.p; - return true; - } - } - else if(is_separator(*t)) - { - pstate = pstate->next.p; - return true; - } - return false; -} - -template -bool perl_matcher::match_end_line() -{ - if(position != last) - { - if(m_match_flags & match_single_line) - return false; - // we're not yet at the end so *first is always valid: - if(is_separator(*position)) - { - if((position != backstop) || (m_match_flags & match_prev_avail)) - { - // check that we're not in the middle of \r\n sequence - BidiIterator t(position); - --t; - if((*t == static_cast('\r')) && (*position == static_cast('\n'))) - { - return false; - } - } - pstate = pstate->next.p; - return true; - } - } - else if((m_match_flags & match_not_eol) == 0) - { - pstate = pstate->next.p; - return true; - } - return false; -} - -template -bool perl_matcher::match_wild() -{ - if(position == last) - return false; - if(is_separator(*position) && ((match_any_mask & static_cast(pstate)->mask) == 0)) - return false; - if((*position == char_type(0)) && (m_match_flags & match_not_dot_null)) - return false; - pstate = pstate->next.p; - ++position; - return true; -} - -template -bool perl_matcher::match_word_boundary() -{ - bool b; // indcates whether next character is a word character - if(position != last) - { - // prev and this character must be opposites: - #if defined(BOOST_REGEX_USE_C_LOCALE) && defined(__GNUC__) && (__GNUC__ == 2) && (__GNUC_MINOR__ < 95) - b = traits::isctype(*position, m_word_mask); - #else - b = traits_inst.isctype(*position, m_word_mask); - #endif - } - else - { - b = (m_match_flags & match_not_eow) ? true : false; - } - if((position == backstop) && ((m_match_flags & match_prev_avail) == 0)) - { - if(m_match_flags & match_not_bow) - b ^= true; - else - b ^= false; - } - else - { - --position; - b ^= traits_inst.isctype(*position, m_word_mask); - ++position; - } - if(b) - { - pstate = pstate->next.p; - return true; - } - return false; // no match if we get to here... -} - -template -bool perl_matcher::match_within_word() -{ - if(position == last) - return false; - // both prev and this character must be m_word_mask: - bool prev = traits_inst.isctype(*position, m_word_mask); - { - bool b; - if((position == backstop) && ((m_match_flags & match_prev_avail) == 0)) - return false; - else - { - --position; - b = traits_inst.isctype(*position, m_word_mask); - ++position; - } - if(b == prev) - { - pstate = pstate->next.p; - return true; - } - } - return false; -} - -template -bool perl_matcher::match_word_start() -{ - if(position == last) - return false; // can't be starting a word if we're already at the end of input - if(!traits_inst.isctype(*position, m_word_mask)) - return false; // next character isn't a word character - if((position == backstop) && ((m_match_flags & match_prev_avail) == 0)) - { - if(m_match_flags & match_not_bow) - return false; // no previous input - } - else - { - // otherwise inside buffer: - BidiIterator t(position); - --t; - if(traits_inst.isctype(*t, m_word_mask)) - return false; // previous character not non-word - } - // OK we have a match: - pstate = pstate->next.p; - return true; -} - -template -bool perl_matcher::match_word_end() -{ - if((position == backstop) && ((m_match_flags & match_prev_avail) == 0)) - return false; // start of buffer can't be end of word - BidiIterator t(position); - --t; - if(traits_inst.isctype(*t, m_word_mask) == false) - return false; // previous character wasn't a word character - - if(position == last) - { - if(m_match_flags & match_not_eow) - return false; // end of buffer but not end of word - } - else - { - // otherwise inside buffer: - if(traits_inst.isctype(*position, m_word_mask)) - return false; // next character is a word character - } - pstate = pstate->next.p; - return true; // if we fall through to here then we've succeeded -} - -template -bool perl_matcher::match_buffer_start() -{ - if((position != backstop) || (m_match_flags & match_not_bob)) - return false; - // OK match: - pstate = pstate->next.p; - return true; -} - -template -bool perl_matcher::match_buffer_end() -{ - if((position != last) || (m_match_flags & match_not_eob)) - return false; - // OK match: - pstate = pstate->next.p; - return true; -} - -template -bool perl_matcher::match_backref() -{ - // - // Compare with what we previously matched. - // Note that this succeeds if the backref did not partisipate - // in the match, this is in line with ECMAScript, but not Perl - // or PCRE. - // - int index = static_cast(pstate)->index; - if(index >= 10000) - { - named_subexpressions::range_type r = re.get_data().equal_range(index); - BOOST_ASSERT(r.first != r.second); - do - { - index = r.first->index; - ++r.first; - }while((r.first != r.second) && ((*m_presult)[index].matched != true)); - } - - if((m_match_flags & match_perl) && !(*m_presult)[index].matched) - return false; - - BidiIterator i = (*m_presult)[index].first; - BidiIterator j = (*m_presult)[index].second; - while(i != j) - { - if((position == last) || (traits_inst.translate(*position, icase) != traits_inst.translate(*i, icase))) - return false; - ++i; - ++position; - } - pstate = pstate->next.p; - return true; -} - -template -bool perl_matcher::match_long_set() -{ - typedef typename traits::char_class_type char_class_type; - // let the traits class do the work: - if(position == last) - return false; - BidiIterator t = re_is_set_member(position, last, static_cast*>(pstate), re.get_data(), icase); - if(t != position) - { - pstate = pstate->next.p; - position = t; - return true; - } - return false; -} - -template -bool perl_matcher::match_set() -{ - if(position == last) - return false; - if(static_cast(pstate)->_map[static_cast(traits_inst.translate(*position, icase))]) - { - pstate = pstate->next.p; - ++position; - return true; - } - return false; -} - -template -bool perl_matcher::match_jump() -{ - pstate = static_cast(pstate)->alt.p; - return true; -} - -template -bool perl_matcher::match_combining() -{ - if(position == last) - return false; - if(is_combining(traits_inst.translate(*position, icase))) - return false; - ++position; - while((position != last) && is_combining(traits_inst.translate(*position, icase))) - ++position; - pstate = pstate->next.p; - return true; -} - -template -bool perl_matcher::match_soft_buffer_end() -{ - if(m_match_flags & match_not_eob) - return false; - BidiIterator p(position); - while((p != last) && is_separator(traits_inst.translate(*p, icase)))++p; - if(p != last) - return false; - pstate = pstate->next.p; - return true; -} - -template -bool perl_matcher::match_restart_continue() -{ - if(position == search_base) - { - pstate = pstate->next.p; - return true; - } - return false; -} - -template -bool perl_matcher::match_backstep() -{ -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable:4127) -#endif - if( ::cutl_details_boost::is_random_access_iterator::value) - { - std::ptrdiff_t maxlen = ::cutl_details_boost::re_detail::distance(backstop, position); - if(maxlen < static_cast(pstate)->index) - return false; - std::advance(position, -static_cast(pstate)->index); - } - else - { - int c = static_cast(pstate)->index; - while(c--) - { - if(position == backstop) - return false; - --position; - } - } - pstate = pstate->next.p; - return true; -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif -} - -template -inline bool perl_matcher::match_assert_backref() -{ - // return true if marked sub-expression N has been matched: - int index = static_cast(pstate)->index; - bool result = false; - if(index == 9999) - { - // Magic value for a (DEFINE) block: - return false; - } - else if(index > 0) - { - // Have we matched subexpression "index"? - // Check if index is a hash value: - if(index >= 10000) - { - named_subexpressions::range_type r = re.get_data().equal_range(index); - while(r.first != r.second) - { - if((*m_presult)[r.first->index].matched) - { - result = true; - break; - } - ++r.first; - } - } - else - { - result = (*m_presult)[index].matched; - } - pstate = pstate->next.p; - } - else - { - // Have we recursed into subexpression "index"? - // If index == 0 then check for any recursion at all, otherwise for recursion to -index-1. - int idx = -index-1; - if(idx >= 10000) - { - named_subexpressions::range_type r = re.get_data().equal_range(idx); - int stack_index = recursion_stack.empty() ? -1 : recursion_stack.back().idx; - while(r.first != r.second) - { - result |= (stack_index == r.first->index); - if(result)break; - ++r.first; - } - } - else - { - result = !recursion_stack.empty() && ((recursion_stack.back().idx == idx) || (index == 0)); - } - pstate = pstate->next.p; - } - return result; -} - -template -bool perl_matcher::match_toggle_case() -{ - // change our case sensitivity: - this->icase = static_cast(pstate)->icase; - pstate = pstate->next.p; - return true; -} - - -template -bool perl_matcher::find_restart_any() -{ -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable:4127) -#endif - const unsigned char* _map = re.get_map(); - while(true) - { - // skip everything we can't match: - while((position != last) && !can_start(*position, _map, (unsigned char)mask_any) ) - ++position; - if(position == last) - { - // run out of characters, try a null match if possible: - if(re.can_be_null()) - return match_prefix(); - break; - } - // now try and obtain a match: - if(match_prefix()) - return true; - if(position == last) - return false; - ++position; - } - return false; -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif -} - -template -bool perl_matcher::find_restart_word() -{ -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable:4127) -#endif - // do search optimised for word starts: - const unsigned char* _map = re.get_map(); - if((m_match_flags & match_prev_avail) || (position != base)) - --position; - else if(match_prefix()) - return true; - do - { - while((position != last) && traits_inst.isctype(*position, m_word_mask)) - ++position; - while((position != last) && !traits_inst.isctype(*position, m_word_mask)) - ++position; - if(position == last) - break; - - if(can_start(*position, _map, (unsigned char)mask_any) ) - { - if(match_prefix()) - return true; - } - if(position == last) - break; - } while(true); - return false; -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif -} - -template -bool perl_matcher::find_restart_line() -{ - // do search optimised for line starts: - const unsigned char* _map = re.get_map(); - if(match_prefix()) - return true; - while(position != last) - { - while((position != last) && !is_separator(*position)) - ++position; - if(position == last) - return false; - ++position; - if(position == last) - { - if(re.can_be_null() && match_prefix()) - return true; - return false; - } - - if( can_start(*position, _map, (unsigned char)mask_any) ) - { - if(match_prefix()) - return true; - } - if(position == last) - return false; - //++position; - } - return false; -} - -template -bool perl_matcher::find_restart_buf() -{ - if((position == base) && ((m_match_flags & match_not_bob) == 0)) - return match_prefix(); - return false; -} - -template -bool perl_matcher::find_restart_lit() -{ -#if 0 - if(position == last) - return false; // can't possibly match if we're at the end already - - unsigned type = (m_match_flags & match_continuous) ? - static_cast(regbase::restart_continue) - : static_cast(re.get_restart_type()); - - const kmp_info* info = access::get_kmp(re); - int len = info->len; - const char_type* x = info->pstr; - int j = 0; - while (position != last) - { - while((j > -1) && (x[j] != traits_inst.translate(*position, icase))) - j = info->kmp_next[j]; - ++position; - ++j; - if(j >= len) - { - if(type == regbase::restart_fixed_lit) - { - std::advance(position, -j); - restart = position; - std::advance(restart, len); - m_result.set_first(position); - m_result.set_second(restart); - position = restart; - return true; - } - else - { - restart = position; - std::advance(position, -j); - if(match_prefix()) - return true; - else - { - for(int k = 0; (restart != position) && (k < j); ++k, --restart) - {} // dwa 10/20/2000 - warning suppression for MWCW - if(restart != last) - ++restart; - position = restart; - j = 0; //we could do better than this... - } - } - } - } - if((m_match_flags & match_partial) && (position == last) && j) - { - // we need to check for a partial match: - restart = position; - std::advance(position, -j); - return match_prefix(); - } -#endif - return false; -} - -} // namespace re_detail - -} // namespace cutl_details_boost - -#ifdef BOOST_MSVC -# pragma warning(pop) -#endif - -#ifdef __BORLANDC__ -# pragma option pop -#endif -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable: 4103) -#endif -#ifdef BOOST_HAS_ABI_HEADERS -# include BOOST_ABI_SUFFIX -#endif -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif - -#endif - diff --git a/cutl/details/boost/regex/v4/perl_matcher_non_recursive.hpp b/cutl/details/boost/regex/v4/perl_matcher_non_recursive.hpp deleted file mode 100644 index 9e3d74d..0000000 --- a/cutl/details/boost/regex/v4/perl_matcher_non_recursive.hpp +++ /dev/null @@ -1,1642 +0,0 @@ -/* - * - * Copyright (c) 2002 - * John Maddock - * - * Use, modification and distribution are subject to the - * Boost Software License, Version 1.0. (See accompanying file - * LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt) - * - */ - - /* - * LOCATION: see http://www.boost.org for most recent version. - * FILE perl_matcher_common.cpp - * VERSION see - * DESCRIPTION: Definitions of perl_matcher member functions that are - * specific to the non-recursive implementation. - */ - -#ifndef BOOST_REGEX_V4_PERL_MATCHER_NON_RECURSIVE_HPP -#define BOOST_REGEX_V4_PERL_MATCHER_NON_RECURSIVE_HPP - -#include - -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable: 4103) -#endif -#ifdef BOOST_HAS_ABI_HEADERS -# include BOOST_ABI_PREFIX -#endif -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif -#ifdef BOOST_MSVC -# pragma warning(push) -# pragma warning(disable: 4800) -#endif - -namespace cutl_details_boost{ -namespace re_detail{ - -template -inline void inplace_destroy(T* p) -{ - (void)p; // warning suppression - p->~T(); -} - -struct saved_state -{ - union{ - unsigned int state_id; - // this padding ensures correct alignment on 64-bit platforms: - std::size_t padding1; - std::ptrdiff_t padding2; - void* padding3; - }; - saved_state(unsigned i) : state_id(i) {} -}; - -template -struct saved_matched_paren : public saved_state -{ - int index; - sub_match sub; - saved_matched_paren(int i, const sub_match& s) : saved_state(1), index(i), sub(s){}; -}; - -template -struct saved_position : public saved_state -{ - const re_syntax_base* pstate; - BidiIterator position; - saved_position(const re_syntax_base* ps, BidiIterator pos, int i) : saved_state(i), pstate(ps), position(pos){}; -}; - -template -struct saved_assertion : public saved_position -{ - bool positive; - saved_assertion(bool p, const re_syntax_base* ps, BidiIterator pos) - : saved_position(ps, pos, saved_type_assertion), positive(p){}; -}; - -template -struct saved_repeater : public saved_state -{ - repeater_count count; - saved_repeater(int i, repeater_count** s, BidiIterator start) - : saved_state(saved_state_repeater_count), count(i,s,start){} -}; - -struct saved_extra_block : public saved_state -{ - saved_state *base, *end; - saved_extra_block(saved_state* b, saved_state* e) - : saved_state(saved_state_extra_block), base(b), end(e) {} -}; - -struct save_state_init -{ - saved_state** stack; - save_state_init(saved_state** base, saved_state** end) - : stack(base) - { - *base = static_cast(get_mem_block()); - *end = reinterpret_cast(reinterpret_cast(*base)+BOOST_REGEX_BLOCKSIZE); - --(*end); - (void) new (*end)saved_state(0); - BOOST_ASSERT(*end > *base); - } - ~save_state_init() - { - put_mem_block(*stack); - *stack = 0; - } -}; - -template -struct saved_single_repeat : public saved_state -{ - std::size_t count; - const re_repeat* rep; - BidiIterator last_position; - saved_single_repeat(std::size_t c, const re_repeat* r, BidiIterator lp, int arg_id) - : saved_state(arg_id), count(c), rep(r), last_position(lp){} -}; - -template -struct saved_recursion : public saved_state -{ - saved_recursion(int idx, const re_syntax_base* p, Results* pr) - : saved_state(14), recursion_id(idx), preturn_address(p), results(*pr) - {} - int recursion_id; - const re_syntax_base* preturn_address; - Results results; -}; - -template -bool perl_matcher::match_all_states() -{ - static matcher_proc_type const s_match_vtable[30] = - { - (&perl_matcher::match_startmark), - &perl_matcher::match_endmark, - &perl_matcher::match_literal, - &perl_matcher::match_start_line, - &perl_matcher::match_end_line, - &perl_matcher::match_wild, - &perl_matcher::match_match, - &perl_matcher::match_word_boundary, - &perl_matcher::match_within_word, - &perl_matcher::match_word_start, - &perl_matcher::match_word_end, - &perl_matcher::match_buffer_start, - &perl_matcher::match_buffer_end, - &perl_matcher::match_backref, - &perl_matcher::match_long_set, - &perl_matcher::match_set, - &perl_matcher::match_jump, - &perl_matcher::match_alt, - &perl_matcher::match_rep, - &perl_matcher::match_combining, - &perl_matcher::match_soft_buffer_end, - &perl_matcher::match_restart_continue, - // Although this next line *should* be evaluated at compile time, in practice - // some compilers (VC++) emit run-time initialisation which breaks thread - // safety, so use a dispatch function instead: - //(::cutl_details_boost::is_random_access_iterator::value ? &perl_matcher::match_dot_repeat_fast : &perl_matcher::match_dot_repeat_slow), - &perl_matcher::match_dot_repeat_dispatch, - &perl_matcher::match_char_repeat, - &perl_matcher::match_set_repeat, - &perl_matcher::match_long_set_repeat, - &perl_matcher::match_backstep, - &perl_matcher::match_assert_backref, - &perl_matcher::match_toggle_case, - &perl_matcher::match_recursion, - }; - - push_recursion_stopper(); - do{ - while(pstate) - { - matcher_proc_type proc = s_match_vtable[pstate->type]; - ++state_count; - if(!(this->*proc)()) - { - if(state_count > max_state_count) - raise_error(traits_inst, regex_constants::error_complexity); - if((m_match_flags & match_partial) && (position == last) && (position != search_base)) - m_has_partial_match = true; - bool successful_unwind = unwind(false); - if((m_match_flags & match_partial) && (position == last) && (position != search_base)) - m_has_partial_match = true; - if(false == successful_unwind) - return m_recursive_result; - } - } - }while(unwind(true)); - return m_recursive_result; -} - -template -void perl_matcher::extend_stack() -{ - if(used_block_count) - { - --used_block_count; - saved_state* stack_base; - saved_state* backup_state; - stack_base = static_cast(get_mem_block()); - backup_state = reinterpret_cast(reinterpret_cast(stack_base)+BOOST_REGEX_BLOCKSIZE); - saved_extra_block* block = static_cast(backup_state); - --block; - (void) new (block) saved_extra_block(m_stack_base, m_backup_state); - m_stack_base = stack_base; - m_backup_state = block; - } - else - raise_error(traits_inst, regex_constants::error_stack); -} - -template -inline void perl_matcher::push_matched_paren(int index, const sub_match& sub) -{ - //BOOST_ASSERT(index); - saved_matched_paren* pmp = static_cast*>(m_backup_state); - --pmp; - if(pmp < m_stack_base) - { - extend_stack(); - pmp = static_cast*>(m_backup_state); - --pmp; - } - (void) new (pmp)saved_matched_paren(index, sub); - m_backup_state = pmp; -} - -template -inline void perl_matcher::push_recursion_stopper() -{ - saved_state* pmp = m_backup_state; - --pmp; - if(pmp < m_stack_base) - { - extend_stack(); - pmp = m_backup_state; - --pmp; - } - (void) new (pmp)saved_state(saved_type_recurse); - m_backup_state = pmp; -} - -template -inline void perl_matcher::push_assertion(const re_syntax_base* ps, bool positive) -{ - saved_assertion* pmp = static_cast*>(m_backup_state); - --pmp; - if(pmp < m_stack_base) - { - extend_stack(); - pmp = static_cast*>(m_backup_state); - --pmp; - } - (void) new (pmp)saved_assertion(positive, ps, position); - m_backup_state = pmp; -} - -template -inline void perl_matcher::push_alt(const re_syntax_base* ps) -{ - saved_position* pmp = static_cast*>(m_backup_state); - --pmp; - if(pmp < m_stack_base) - { - extend_stack(); - pmp = static_cast*>(m_backup_state); - --pmp; - } - (void) new (pmp)saved_position(ps, position, saved_state_alt); - m_backup_state = pmp; -} - -template -inline void perl_matcher::push_non_greedy_repeat(const re_syntax_base* ps) -{ - saved_position* pmp = static_cast*>(m_backup_state); - --pmp; - if(pmp < m_stack_base) - { - extend_stack(); - pmp = static_cast*>(m_backup_state); - --pmp; - } - (void) new (pmp)saved_position(ps, position, saved_state_non_greedy_long_repeat); - m_backup_state = pmp; -} - -template -inline void perl_matcher::push_repeater_count(int i, repeater_count** s) -{ - saved_repeater* pmp = static_cast*>(m_backup_state); - --pmp; - if(pmp < m_stack_base) - { - extend_stack(); - pmp = static_cast*>(m_backup_state); - --pmp; - } - (void) new (pmp)saved_repeater(i, s, position); - m_backup_state = pmp; -} - -template -inline void perl_matcher::push_single_repeat(std::size_t c, const re_repeat* r, BidiIterator last_position, int state_id) -{ - saved_single_repeat* pmp = static_cast*>(m_backup_state); - --pmp; - if(pmp < m_stack_base) - { - extend_stack(); - pmp = static_cast*>(m_backup_state); - --pmp; - } - (void) new (pmp)saved_single_repeat(c, r, last_position, state_id); - m_backup_state = pmp; -} - -template -inline void perl_matcher::push_recursion(int idx, const re_syntax_base* p, results_type* presults) -{ - saved_recursion* pmp = static_cast*>(m_backup_state); - --pmp; - if(pmp < m_stack_base) - { - extend_stack(); - pmp = static_cast*>(m_backup_state); - --pmp; - } - (void) new (pmp)saved_recursion(idx, p, presults); - m_backup_state = pmp; -} - -template -bool perl_matcher::match_startmark() -{ - int index = static_cast(pstate)->index; - icase = static_cast(pstate)->icase; - switch(index) - { - case 0: - pstate = pstate->next.p; - break; - case -1: - case -2: - { - // forward lookahead assert: - const re_syntax_base* next_pstate = static_cast(pstate->next.p)->alt.p->next.p; - pstate = pstate->next.p->next.p; - push_assertion(next_pstate, index == -1); - break; - } - case -3: - { - // independent sub-expression, currently this is always recursive: - bool old_independent = m_independent; - m_independent = true; - const re_syntax_base* next_pstate = static_cast(pstate->next.p)->alt.p->next.p; - pstate = pstate->next.p->next.p; - bool r = match_all_states(); - pstate = next_pstate; - m_independent = old_independent; -#ifdef BOOST_REGEX_MATCH_EXTRA - if(r && (m_match_flags & match_extra)) - { - // - // our captures have been stored in *m_presult - // we need to unpack them, and insert them - // back in the right order when we unwind the stack: - // - match_results temp_match(*m_presult); - unsigned i; - for(i = 0; i < temp_match.size(); ++i) - (*m_presult)[i].get_captures().clear(); - // match everything else: - r = match_all_states(); - // now place the stored captures back: - for(i = 0; i < temp_match.size(); ++i) - { - typedef typename sub_match::capture_sequence_type seq; - seq& s1 = (*m_presult)[i].get_captures(); - const seq& s2 = temp_match[i].captures(); - s1.insert( - s1.end(), - s2.begin(), - s2.end()); - } - } -#endif - return r; - } - case -4: - { - // conditional expression: - const re_alt* alt = static_cast(pstate->next.p); - BOOST_ASSERT(alt->type == syntax_element_alt); - pstate = alt->next.p; - if(pstate->type == syntax_element_assert_backref) - { - if(!match_assert_backref()) - pstate = alt->alt.p; - break; - } - else - { - // zero width assertion, have to match this recursively: - BOOST_ASSERT(pstate->type == syntax_element_startmark); - bool negated = static_cast(pstate)->index == -2; - BidiIterator saved_position = position; - const re_syntax_base* next_pstate = static_cast(pstate->next.p)->alt.p->next.p; - pstate = pstate->next.p->next.p; - bool r = match_all_states(); - position = saved_position; - if(negated) - r = !r; - if(r) - pstate = next_pstate; - else - pstate = alt->alt.p; - break; - } - } - case -5: - { - push_matched_paren(0, (*m_presult)[0]); - m_presult->set_first(position, 0, true); - pstate = pstate->next.p; - break; - } - default: - { - BOOST_ASSERT(index > 0); - if((m_match_flags & match_nosubs) == 0) - { - push_matched_paren(index, (*m_presult)[index]); - m_presult->set_first(position, index); - } - pstate = pstate->next.p; - break; - } - } - return true; -} - -template -bool perl_matcher::match_alt() -{ - bool take_first, take_second; - const re_alt* jmp = static_cast(pstate); - - // find out which of these two alternatives we need to take: - if(position == last) - { - take_first = jmp->can_be_null & mask_take; - take_second = jmp->can_be_null & mask_skip; - } - else - { - take_first = can_start(*position, jmp->_map, (unsigned char)mask_take); - take_second = can_start(*position, jmp->_map, (unsigned char)mask_skip); - } - - if(take_first) - { - // we can take the first alternative, - // see if we need to push next alternative: - if(take_second) - { - push_alt(jmp->alt.p); - } - pstate = pstate->next.p; - return true; - } - if(take_second) - { - pstate = jmp->alt.p; - return true; - } - return false; // neither option is possible -} - -template -bool perl_matcher::match_rep() -{ -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable:4127 4244) -#endif -#ifdef __BORLANDC__ -#pragma option push -w-8008 -w-8066 -w-8004 -#endif - const re_repeat* rep = static_cast(pstate); - - // find out which of these two alternatives we need to take: - bool take_first, take_second; - if(position == last) - { - take_first = rep->can_be_null & mask_take; - take_second = rep->can_be_null & mask_skip; - } - else - { - take_first = can_start(*position, rep->_map, (unsigned char)mask_take); - take_second = can_start(*position, rep->_map, (unsigned char)mask_skip); - } - - if((m_backup_state->state_id != saved_state_repeater_count) - || (static_cast*>(m_backup_state)->count.get_id() != rep->state_id) - || (next_count->get_id() != rep->state_id)) - { - // we're moving to a different repeat from the last - // one, so set up a counter object: - push_repeater_count(rep->state_id, &next_count); - } - // - // If we've had at least one repeat already, and the last one - // matched the NULL string then set the repeat count to - // maximum: - // - next_count->check_null_repeat(position, rep->max); - - if(next_count->get_count() < rep->min) - { - // we must take the repeat: - if(take_first) - { - // increase the counter: - ++(*next_count); - pstate = rep->next.p; - return true; - } - return false; - } - - bool greedy = (rep->greedy) && (!(m_match_flags & regex_constants::match_any) || m_independent); - if(greedy) - { - // try and take the repeat if we can: - if((next_count->get_count() < rep->max) && take_first) - { - if(take_second) - { - // store position in case we fail: - push_alt(rep->alt.p); - } - // increase the counter: - ++(*next_count); - pstate = rep->next.p; - return true; - } - else if(take_second) - { - pstate = rep->alt.p; - return true; - } - return false; // can't take anything, fail... - } - else // non-greedy - { - // try and skip the repeat if we can: - if(take_second) - { - if((next_count->get_count() < rep->max) && take_first) - { - // store position in case we fail: - push_non_greedy_repeat(rep->next.p); - } - pstate = rep->alt.p; - return true; - } - if((next_count->get_count() < rep->max) && take_first) - { - // increase the counter: - ++(*next_count); - pstate = rep->next.p; - return true; - } - } - return false; -#ifdef __BORLANDC__ -#pragma option pop -#endif -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif -} - -template -bool perl_matcher::match_dot_repeat_slow() -{ - unsigned count = 0; - const re_repeat* rep = static_cast(pstate); - re_syntax_base* psingle = rep->next.p; - // match compulsary repeats first: - while(count < rep->min) - { - pstate = psingle; - if(!match_wild()) - return false; - ++count; - } - bool greedy = (rep->greedy) && (!(m_match_flags & regex_constants::match_any) || m_independent); - if(greedy) - { - // repeat for as long as we can: - while(count < rep->max) - { - pstate = psingle; - if(!match_wild()) - break; - ++count; - } - // remember where we got to if this is a leading repeat: - if((rep->leading) && (count < rep->max)) - restart = position; - // push backtrack info if available: - if(count - rep->min) - push_single_repeat(count, rep, position, saved_state_greedy_single_repeat); - // jump to next state: - pstate = rep->alt.p; - return true; - } - else - { - // non-greedy, push state and return true if we can skip: - if(count < rep->max) - push_single_repeat(count, rep, position, saved_state_rep_slow_dot); - pstate = rep->alt.p; - return (position == last) ? (rep->can_be_null & mask_skip) : can_start(*position, rep->_map, mask_skip); - } -} - -template -bool perl_matcher::match_dot_repeat_fast() -{ - if(m_match_flags & match_not_dot_null) - return match_dot_repeat_slow(); - if((static_cast(pstate->next.p)->mask & match_any_mask) == 0) - return match_dot_repeat_slow(); - - const re_repeat* rep = static_cast(pstate); - bool greedy = (rep->greedy) && (!(m_match_flags & regex_constants::match_any) || m_independent); - unsigned count = static_cast((std::min)(static_cast(::cutl_details_boost::re_detail::distance(position, last)), static_cast(greedy ? rep->max : rep->min))); - if(rep->min > count) - { - position = last; - return false; // not enough text left to match - } - std::advance(position, count); - - if(greedy) - { - if((rep->leading) && (count < rep->max)) - restart = position; - // push backtrack info if available: - if(count - rep->min) - push_single_repeat(count, rep, position, saved_state_greedy_single_repeat); - // jump to next state: - pstate = rep->alt.p; - return true; - } - else - { - // non-greedy, push state and return true if we can skip: - if(count < rep->max) - push_single_repeat(count, rep, position, saved_state_rep_fast_dot); - pstate = rep->alt.p; - return (position == last) ? (rep->can_be_null & mask_skip) : can_start(*position, rep->_map, mask_skip); - } -} - -template -bool perl_matcher::match_char_repeat() -{ -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable:4127) -#endif -#ifdef __BORLANDC__ -#pragma option push -w-8008 -w-8066 -w-8004 -#endif - const re_repeat* rep = static_cast(pstate); - BOOST_ASSERT(1 == static_cast(rep->next.p)->length); - const char_type what = *reinterpret_cast(static_cast(rep->next.p) + 1); - std::size_t count = 0; - // - // start by working out how much we can skip: - // - bool greedy = (rep->greedy) && (!(m_match_flags & regex_constants::match_any) || m_independent); - std::size_t desired = greedy ? rep->max : rep->min; - if(::cutl_details_boost::is_random_access_iterator::value) - { - BidiIterator end = position; - std::advance(end, (std::min)((std::size_t)::cutl_details_boost::re_detail::distance(position, last), desired)); - BidiIterator origin(position); - while((position != end) && (traits_inst.translate(*position, icase) == what)) - { - ++position; - } - count = (unsigned)::cutl_details_boost::re_detail::distance(origin, position); - } - else - { - while((count < desired) && (position != last) && (traits_inst.translate(*position, icase) == what)) - { - ++position; - ++count; - } - } - - if(count < rep->min) - return false; - - if(greedy) - { - if((rep->leading) && (count < rep->max)) - restart = position; - // push backtrack info if available: - if(count - rep->min) - push_single_repeat(count, rep, position, saved_state_greedy_single_repeat); - // jump to next state: - pstate = rep->alt.p; - return true; - } - else - { - // non-greedy, push state and return true if we can skip: - if(count < rep->max) - push_single_repeat(count, rep, position, saved_state_rep_char); - pstate = rep->alt.p; - return (position == last) ? (rep->can_be_null & mask_skip) : can_start(*position, rep->_map, mask_skip); - } -#ifdef __BORLANDC__ -#pragma option pop -#endif -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif -} - -template -bool perl_matcher::match_set_repeat() -{ -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable:4127) -#endif -#ifdef __BORLANDC__ -#pragma option push -w-8008 -w-8066 -w-8004 -#endif - const re_repeat* rep = static_cast(pstate); - const unsigned char* map = static_cast(rep->next.p)->_map; - std::size_t count = 0; - // - // start by working out how much we can skip: - // - bool greedy = (rep->greedy) && (!(m_match_flags & regex_constants::match_any) || m_independent); - std::size_t desired = greedy ? rep->max : rep->min; - if(::cutl_details_boost::is_random_access_iterator::value) - { - BidiIterator end = position; - std::advance(end, (std::min)((std::size_t)::cutl_details_boost::re_detail::distance(position, last), desired)); - BidiIterator origin(position); - while((position != end) && map[static_cast(traits_inst.translate(*position, icase))]) - { - ++position; - } - count = (unsigned)::cutl_details_boost::re_detail::distance(origin, position); - } - else - { - while((count < desired) && (position != last) && map[static_cast(traits_inst.translate(*position, icase))]) - { - ++position; - ++count; - } - } - - if(count < rep->min) - return false; - - if(greedy) - { - if((rep->leading) && (count < rep->max)) - restart = position; - // push backtrack info if available: - if(count - rep->min) - push_single_repeat(count, rep, position, saved_state_greedy_single_repeat); - // jump to next state: - pstate = rep->alt.p; - return true; - } - else - { - // non-greedy, push state and return true if we can skip: - if(count < rep->max) - push_single_repeat(count, rep, position, saved_state_rep_short_set); - pstate = rep->alt.p; - return (position == last) ? (rep->can_be_null & mask_skip) : can_start(*position, rep->_map, mask_skip); - } -#ifdef __BORLANDC__ -#pragma option pop -#endif -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif -} - -template -bool perl_matcher::match_long_set_repeat() -{ -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable:4127) -#endif -#ifdef __BORLANDC__ -#pragma option push -w-8008 -w-8066 -w-8004 -#endif - typedef typename traits::char_class_type m_type; - const re_repeat* rep = static_cast(pstate); - const re_set_long* set = static_cast*>(pstate->next.p); - std::size_t count = 0; - // - // start by working out how much we can skip: - // - bool greedy = (rep->greedy) && (!(m_match_flags & regex_constants::match_any) || m_independent); - std::size_t desired = greedy ? rep->max : rep->min; - if(::cutl_details_boost::is_random_access_iterator::value) - { - BidiIterator end = position; - std::advance(end, (std::min)((std::size_t)::cutl_details_boost::re_detail::distance(position, last), desired)); - BidiIterator origin(position); - while((position != end) && (position != re_is_set_member(position, last, set, re.get_data(), icase))) - { - ++position; - } - count = (unsigned)::cutl_details_boost::re_detail::distance(origin, position); - } - else - { - while((count < desired) && (position != last) && (position != re_is_set_member(position, last, set, re.get_data(), icase))) - { - ++position; - ++count; - } - } - - if(count < rep->min) - return false; - - if(greedy) - { - if((rep->leading) && (count < rep->max)) - restart = position; - // push backtrack info if available: - if(count - rep->min) - push_single_repeat(count, rep, position, saved_state_greedy_single_repeat); - // jump to next state: - pstate = rep->alt.p; - return true; - } - else - { - // non-greedy, push state and return true if we can skip: - if(count < rep->max) - push_single_repeat(count, rep, position, saved_state_rep_long_set); - pstate = rep->alt.p; - return (position == last) ? (rep->can_be_null & mask_skip) : can_start(*position, rep->_map, mask_skip); - } -#ifdef __BORLANDC__ -#pragma option pop -#endif -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif -} - -template -bool perl_matcher::match_recursion() -{ - BOOST_ASSERT(pstate->type == syntax_element_recurse); - // - // Backup call stack: - // - push_recursion_pop(); - // - // Set new call stack: - // - if(recursion_stack.capacity() == 0) - { - recursion_stack.reserve(50); - } - recursion_stack.push_back(recursion_info()); - recursion_stack.back().preturn_address = pstate->next.p; - recursion_stack.back().results = *m_presult; - if(static_cast(pstate)->state_id > 0) - { - push_repeater_count(static_cast(pstate)->state_id, &next_count); - } - pstate = static_cast(pstate)->alt.p; - recursion_stack.back().idx = static_cast(pstate)->index; - - return true; -} - -template -bool perl_matcher::match_endmark() -{ - int index = static_cast(pstate)->index; - icase = static_cast(pstate)->icase; - if(index > 0) - { - if((m_match_flags & match_nosubs) == 0) - { - m_presult->set_second(position, index); - } - if(!recursion_stack.empty()) - { - if(index == recursion_stack.back().idx) - { - pstate = recursion_stack.back().preturn_address; - *m_presult = recursion_stack.back().results; - push_recursion(recursion_stack.back().idx, recursion_stack.back().preturn_address, &recursion_stack.back().results); - recursion_stack.pop_back(); - } - } - } - else if((index < 0) && (index != -4)) - { - // matched forward lookahead: - pstate = 0; - return true; - } - pstate = pstate->next.p; - return true; -} - -template -bool perl_matcher::match_match() -{ - if(!recursion_stack.empty()) - { - BOOST_ASSERT(0 == recursion_stack.back().idx); - pstate = recursion_stack.back().preturn_address; - *m_presult = recursion_stack.back().results; - push_recursion(recursion_stack.back().idx, recursion_stack.back().preturn_address, &recursion_stack.back().results); - recursion_stack.pop_back(); - return true; - } - if((m_match_flags & match_not_null) && (position == (*m_presult)[0].first)) - return false; - if((m_match_flags & match_all) && (position != last)) - return false; - if((m_match_flags & regex_constants::match_not_initial_null) && (position == search_base)) - return false; - m_presult->set_second(position); - pstate = 0; - m_has_found_match = true; - if((m_match_flags & match_posix) == match_posix) - { - m_result.maybe_assign(*m_presult); - if((m_match_flags & match_any) == 0) - return false; - } -#ifdef BOOST_REGEX_MATCH_EXTRA - if(match_extra & m_match_flags) - { - for(unsigned i = 0; i < m_presult->size(); ++i) - if((*m_presult)[i].matched) - ((*m_presult)[i]).get_captures().push_back((*m_presult)[i]); - } -#endif - return true; -} - -/**************************************************************************** - -Unwind and associated proceedures follow, these perform what normal stack -unwinding does in the recursive implementation. - -****************************************************************************/ - -template -bool perl_matcher::unwind(bool have_match) -{ - static unwind_proc_type const s_unwind_table[18] = - { - &perl_matcher::unwind_end, - &perl_matcher::unwind_paren, - &perl_matcher::unwind_recursion_stopper, - &perl_matcher::unwind_assertion, - &perl_matcher::unwind_alt, - &perl_matcher::unwind_repeater_counter, - &perl_matcher::unwind_extra_block, - &perl_matcher::unwind_greedy_single_repeat, - &perl_matcher::unwind_slow_dot_repeat, - &perl_matcher::unwind_fast_dot_repeat, - &perl_matcher::unwind_char_repeat, - &perl_matcher::unwind_short_set_repeat, - &perl_matcher::unwind_long_set_repeat, - &perl_matcher::unwind_non_greedy_repeat, - &perl_matcher::unwind_recursion, - &perl_matcher::unwind_recursion_pop, - }; - - m_recursive_result = have_match; - unwind_proc_type unwinder; - bool cont; - // - // keep unwinding our stack until we have something to do: - // - do - { - unwinder = s_unwind_table[m_backup_state->state_id]; - cont = (this->*unwinder)(m_recursive_result); - }while(cont); - // - // return true if we have more states to try: - // - return pstate ? true : false; -} - -template -bool perl_matcher::unwind_end(bool) -{ - pstate = 0; // nothing left to search - return false; // end of stack nothing more to search -} - -template -bool perl_matcher::unwind_paren(bool have_match) -{ - saved_matched_paren* pmp = static_cast*>(m_backup_state); - // restore previous values if no match was found: - if(have_match == false) - { - m_presult->set_first(pmp->sub.first, pmp->index, pmp->index == 0); - m_presult->set_second(pmp->sub.second, pmp->index, pmp->sub.matched, pmp->index == 0); - } -#ifdef BOOST_REGEX_MATCH_EXTRA - // - // we have a match, push the capture information onto the stack: - // - else if(pmp->sub.matched && (match_extra & m_match_flags)) - ((*m_presult)[pmp->index]).get_captures().push_back(pmp->sub); -#endif - // unwind stack: - m_backup_state = pmp+1; - cutl_details_boost::re_detail::inplace_destroy(pmp); - return true; // keep looking -} - -template -bool perl_matcher::unwind_recursion_stopper(bool) -{ - cutl_details_boost::re_detail::inplace_destroy(m_backup_state++); - pstate = 0; // nothing left to search - return false; // end of stack nothing more to search -} - -template -bool perl_matcher::unwind_assertion(bool r) -{ - saved_assertion* pmp = static_cast*>(m_backup_state); - pstate = pmp->pstate; - position = pmp->position; - bool result = (r == pmp->positive); - m_recursive_result = pmp->positive ? r : !r; - cutl_details_boost::re_detail::inplace_destroy(pmp++); - m_backup_state = pmp; - return !result; // return false if the assertion was matched to stop search. -} - -template -bool perl_matcher::unwind_alt(bool r) -{ - saved_position* pmp = static_cast*>(m_backup_state); - if(!r) - { - pstate = pmp->pstate; - position = pmp->position; - } - cutl_details_boost::re_detail::inplace_destroy(pmp++); - m_backup_state = pmp; - return r; -} - -template -bool perl_matcher::unwind_repeater_counter(bool) -{ - saved_repeater* pmp = static_cast*>(m_backup_state); - cutl_details_boost::re_detail::inplace_destroy(pmp++); - m_backup_state = pmp; - return true; // keep looking -} - -template -bool perl_matcher::unwind_extra_block(bool) -{ - saved_extra_block* pmp = static_cast(m_backup_state); - void* condemmed = m_stack_base; - m_stack_base = pmp->base; - m_backup_state = pmp->end; - cutl_details_boost::re_detail::inplace_destroy(pmp); - put_mem_block(condemmed); - return true; // keep looking -} - -template -inline void perl_matcher::destroy_single_repeat() -{ - saved_single_repeat* p = static_cast*>(m_backup_state); - cutl_details_boost::re_detail::inplace_destroy(p++); - m_backup_state = p; -} - -template -bool perl_matcher::unwind_greedy_single_repeat(bool r) -{ - saved_single_repeat* pmp = static_cast*>(m_backup_state); - - // if we have a match, just discard this state: - if(r) - { - destroy_single_repeat(); - return true; - } - - const re_repeat* rep = pmp->rep; - std::size_t count = pmp->count; - BOOST_ASSERT(rep->next.p != 0); - BOOST_ASSERT(rep->alt.p != 0); - - count -= rep->min; - - if((m_match_flags & match_partial) && (position == last)) - m_has_partial_match = true; - - BOOST_ASSERT(count); - position = pmp->last_position; - - // backtrack till we can skip out: - do - { - --position; - --count; - ++state_count; - }while(count && !can_start(*position, rep->_map, mask_skip)); - - // if we've hit base, destroy this state: - if(count == 0) - { - destroy_single_repeat(); - if(!can_start(*position, rep->_map, mask_skip)) - return true; - } - else - { - pmp->count = count + rep->min; - pmp->last_position = position; - } - pstate = rep->alt.p; - return false; -} - -template -bool perl_matcher::unwind_slow_dot_repeat(bool r) -{ - saved_single_repeat* pmp = static_cast*>(m_backup_state); - - // if we have a match, just discard this state: - if(r) - { - destroy_single_repeat(); - return true; - } - - const re_repeat* rep = pmp->rep; - std::size_t count = pmp->count; - BOOST_ASSERT(rep->type == syntax_element_dot_rep); - BOOST_ASSERT(rep->next.p != 0); - BOOST_ASSERT(rep->alt.p != 0); - BOOST_ASSERT(rep->next.p->type == syntax_element_wild); - - BOOST_ASSERT(count < rep->max); - pstate = rep->next.p; - position = pmp->last_position; - - if(position != last) - { - // wind forward until we can skip out of the repeat: - do - { - if(!match_wild()) - { - // failed repeat match, discard this state and look for another: - destroy_single_repeat(); - return true; - } - ++count; - ++state_count; - pstate = rep->next.p; - }while((count < rep->max) && (position != last) && !can_start(*position, rep->_map, mask_skip)); - } - if(position == last) - { - // can't repeat any more, remove the pushed state: - destroy_single_repeat(); - if((m_match_flags & match_partial) && (position == last) && (position != search_base)) - m_has_partial_match = true; - if(0 == (rep->can_be_null & mask_skip)) - return true; - } - else if(count == rep->max) - { - // can't repeat any more, remove the pushed state: - destroy_single_repeat(); - if(!can_start(*position, rep->_map, mask_skip)) - return true; - } - else - { - pmp->count = count; - pmp->last_position = position; - } - pstate = rep->alt.p; - return false; -} - -template -bool perl_matcher::unwind_fast_dot_repeat(bool r) -{ - saved_single_repeat* pmp = static_cast*>(m_backup_state); - - // if we have a match, just discard this state: - if(r) - { - destroy_single_repeat(); - return true; - } - - const re_repeat* rep = pmp->rep; - std::size_t count = pmp->count; - - BOOST_ASSERT(count < rep->max); - position = pmp->last_position; - if(position != last) - { - - // wind forward until we can skip out of the repeat: - do - { - ++position; - ++count; - ++state_count; - }while((count < rep->max) && (position != last) && !can_start(*position, rep->_map, mask_skip)); - } - - // remember where we got to if this is a leading repeat: - if((rep->leading) && (count < rep->max)) - restart = position; - if(position == last) - { - // can't repeat any more, remove the pushed state: - destroy_single_repeat(); - if((m_match_flags & match_partial) && (position == last) && (position != search_base)) - m_has_partial_match = true; - if(0 == (rep->can_be_null & mask_skip)) - return true; - } - else if(count == rep->max) - { - // can't repeat any more, remove the pushed state: - destroy_single_repeat(); - if(!can_start(*position, rep->_map, mask_skip)) - return true; - } - else - { - pmp->count = count; - pmp->last_position = position; - } - pstate = rep->alt.p; - return false; -} - -template -bool perl_matcher::unwind_char_repeat(bool r) -{ - saved_single_repeat* pmp = static_cast*>(m_backup_state); - - // if we have a match, just discard this state: - if(r) - { - destroy_single_repeat(); - return true; - } - - const re_repeat* rep = pmp->rep; - std::size_t count = pmp->count; - pstate = rep->next.p; - const char_type what = *reinterpret_cast(static_cast(pstate) + 1); - position = pmp->last_position; - - BOOST_ASSERT(rep->type == syntax_element_char_rep); - BOOST_ASSERT(rep->next.p != 0); - BOOST_ASSERT(rep->alt.p != 0); - BOOST_ASSERT(rep->next.p->type == syntax_element_literal); - BOOST_ASSERT(count < rep->max); - - if(position != last) - { - // wind forward until we can skip out of the repeat: - do - { - if(traits_inst.translate(*position, icase) != what) - { - // failed repeat match, discard this state and look for another: - destroy_single_repeat(); - return true; - } - ++count; - ++ position; - ++state_count; - pstate = rep->next.p; - }while((count < rep->max) && (position != last) && !can_start(*position, rep->_map, mask_skip)); - } - // remember where we got to if this is a leading repeat: - if((rep->leading) && (count < rep->max)) - restart = position; - if(position == last) - { - // can't repeat any more, remove the pushed state: - destroy_single_repeat(); - if((m_match_flags & match_partial) && (position == last) && (position != search_base)) - m_has_partial_match = true; - if(0 == (rep->can_be_null & mask_skip)) - return true; - } - else if(count == rep->max) - { - // can't repeat any more, remove the pushed state: - destroy_single_repeat(); - if(!can_start(*position, rep->_map, mask_skip)) - return true; - } - else - { - pmp->count = count; - pmp->last_position = position; - } - pstate = rep->alt.p; - return false; -} - -template -bool perl_matcher::unwind_short_set_repeat(bool r) -{ - saved_single_repeat* pmp = static_cast*>(m_backup_state); - - // if we have a match, just discard this state: - if(r) - { - destroy_single_repeat(); - return true; - } - - const re_repeat* rep = pmp->rep; - std::size_t count = pmp->count; - pstate = rep->next.p; - const unsigned char* map = static_cast(rep->next.p)->_map; - position = pmp->last_position; - - BOOST_ASSERT(rep->type == syntax_element_short_set_rep); - BOOST_ASSERT(rep->next.p != 0); - BOOST_ASSERT(rep->alt.p != 0); - BOOST_ASSERT(rep->next.p->type == syntax_element_set); - BOOST_ASSERT(count < rep->max); - - if(position != last) - { - // wind forward until we can skip out of the repeat: - do - { - if(!map[static_cast(traits_inst.translate(*position, icase))]) - { - // failed repeat match, discard this state and look for another: - destroy_single_repeat(); - return true; - } - ++count; - ++ position; - ++state_count; - pstate = rep->next.p; - }while((count < rep->max) && (position != last) && !can_start(*position, rep->_map, mask_skip)); - } - // remember where we got to if this is a leading repeat: - if((rep->leading) && (count < rep->max)) - restart = position; - if(position == last) - { - // can't repeat any more, remove the pushed state: - destroy_single_repeat(); - if((m_match_flags & match_partial) && (position == last) && (position != search_base)) - m_has_partial_match = true; - if(0 == (rep->can_be_null & mask_skip)) - return true; - } - else if(count == rep->max) - { - // can't repeat any more, remove the pushed state: - destroy_single_repeat(); - if(!can_start(*position, rep->_map, mask_skip)) - return true; - } - else - { - pmp->count = count; - pmp->last_position = position; - } - pstate = rep->alt.p; - return false; -} - -template -bool perl_matcher::unwind_long_set_repeat(bool r) -{ - typedef typename traits::char_class_type m_type; - saved_single_repeat* pmp = static_cast*>(m_backup_state); - - // if we have a match, just discard this state: - if(r) - { - destroy_single_repeat(); - return true; - } - - const re_repeat* rep = pmp->rep; - std::size_t count = pmp->count; - pstate = rep->next.p; - const re_set_long* set = static_cast*>(pstate); - position = pmp->last_position; - - BOOST_ASSERT(rep->type == syntax_element_long_set_rep); - BOOST_ASSERT(rep->next.p != 0); - BOOST_ASSERT(rep->alt.p != 0); - BOOST_ASSERT(rep->next.p->type == syntax_element_long_set); - BOOST_ASSERT(count < rep->max); - - if(position != last) - { - // wind forward until we can skip out of the repeat: - do - { - if(position == re_is_set_member(position, last, set, re.get_data(), icase)) - { - // failed repeat match, discard this state and look for another: - destroy_single_repeat(); - return true; - } - ++position; - ++count; - ++state_count; - pstate = rep->next.p; - }while((count < rep->max) && (position != last) && !can_start(*position, rep->_map, mask_skip)); - } - // remember where we got to if this is a leading repeat: - if((rep->leading) && (count < rep->max)) - restart = position; - if(position == last) - { - // can't repeat any more, remove the pushed state: - destroy_single_repeat(); - if((m_match_flags & match_partial) && (position == last) && (position != search_base)) - m_has_partial_match = true; - if(0 == (rep->can_be_null & mask_skip)) - return true; - } - else if(count == rep->max) - { - // can't repeat any more, remove the pushed state: - destroy_single_repeat(); - if(!can_start(*position, rep->_map, mask_skip)) - return true; - } - else - { - pmp->count = count; - pmp->last_position = position; - } - pstate = rep->alt.p; - return false; -} - -template -bool perl_matcher::unwind_non_greedy_repeat(bool r) -{ - saved_position* pmp = static_cast*>(m_backup_state); - if(!r) - { - position = pmp->position; - pstate = pmp->pstate; - ++(*next_count); - } - cutl_details_boost::re_detail::inplace_destroy(pmp++); - m_backup_state = pmp; - return r; -} - -template -bool perl_matcher::unwind_recursion(bool r) -{ - saved_recursion* pmp = static_cast*>(m_backup_state); - if(!r) - { - recursion_stack.push_back(recursion_info()); - recursion_stack.back().idx = pmp->recursion_id; - recursion_stack.back().preturn_address = pmp->preturn_address; - recursion_stack.back().results = pmp->results; - } - cutl_details_boost::re_detail::inplace_destroy(pmp++); - m_backup_state = pmp; - return true; -} - -template -bool perl_matcher::unwind_recursion_pop(bool r) -{ - saved_state* pmp = static_cast(m_backup_state); - if(!r) - { - recursion_stack.pop_back(); - } - cutl_details_boost::re_detail::inplace_destroy(pmp++); - m_backup_state = pmp; - return true; -} - -template -void perl_matcher::push_recursion_pop() -{ - saved_state* pmp = static_cast(m_backup_state); - --pmp; - if(pmp < m_stack_base) - { - extend_stack(); - pmp = static_cast(m_backup_state); - --pmp; - } - (void) new (pmp)saved_state(15); - m_backup_state = pmp; -} -/* -template -bool perl_matcher::unwind_parenthesis_pop(bool r) -{ - saved_state* pmp = static_cast(m_backup_state); - if(!r) - { - --parenthesis_stack_position; - } - cutl_details_boost::re_detail::inplace_destroy(pmp++); - m_backup_state = pmp; - return true; -} - -template -void perl_matcher::push_parenthesis_pop() -{ - saved_state* pmp = static_cast(m_backup_state); - --pmp; - if(pmp < m_stack_base) - { - extend_stack(); - pmp = static_cast(m_backup_state); - --pmp; - } - (void) new (pmp)saved_state(16); - m_backup_state = pmp; -} - -template -bool perl_matcher::unwind_parenthesis_push(bool r) -{ - saved_position* pmp = static_cast*>(m_backup_state); - if(!r) - { - parenthesis_stack[parenthesis_stack_position++] = pmp->position; - } - cutl_details_boost::re_detail::inplace_destroy(pmp++); - m_backup_state = pmp; - return true; -} - -template -inline void perl_matcher::push_parenthesis_push(BidiIterator p) -{ - saved_position* pmp = static_cast*>(m_backup_state); - --pmp; - if(pmp < m_stack_base) - { - extend_stack(); - pmp = static_cast*>(m_backup_state); - --pmp; - } - (void) new (pmp)saved_position(0, p, 17); - m_backup_state = pmp; -} -*/ -} // namespace re_detail -} // namespace cutl_details_boost - -#ifdef BOOST_MSVC -# pragma warning(pop) -#endif - -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable: 4103) -#endif -#ifdef BOOST_HAS_ABI_HEADERS -# include BOOST_ABI_SUFFIX -#endif -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif - -#endif - - diff --git a/cutl/details/boost/regex/v4/perl_matcher_recursive.hpp b/cutl/details/boost/regex/v4/perl_matcher_recursive.hpp deleted file mode 100644 index 6b4492c..0000000 --- a/cutl/details/boost/regex/v4/perl_matcher_recursive.hpp +++ /dev/null @@ -1,991 +0,0 @@ -/* - * - * Copyright (c) 2002 - * John Maddock - * - * Use, modification and distribution are subject to the - * Boost Software License, Version 1.0. (See accompanying file - * LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt) - * - */ - - /* - * LOCATION: see http://www.boost.org for most recent version. - * FILE perl_matcher_common.cpp - * VERSION see - * DESCRIPTION: Definitions of perl_matcher member functions that are - * specific to the recursive implementation. - */ - -#ifndef BOOST_REGEX_V4_PERL_MATCHER_RECURSIVE_HPP -#define BOOST_REGEX_V4_PERL_MATCHER_RECURSIVE_HPP - -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable: 4103) -#endif -#ifdef BOOST_HAS_ABI_HEADERS -# include BOOST_ABI_PREFIX -#endif -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif - -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable: 4800) -#endif - -namespace cutl_details_boost{ -namespace re_detail{ - -template -class backup_subex -{ - int index; - sub_match sub; -public: - template - backup_subex(const match_results& w, int i) - : index(i), sub(w[i], false) {} - template - void restore(match_results& w) - { - w.set_first(sub.first, index, index == 0); - w.set_second(sub.second, index, sub.matched, index == 0); - } - const sub_match& get() { return sub; } -}; - -template -bool perl_matcher::match_all_states() -{ - static matcher_proc_type const s_match_vtable[30] = - { - (&perl_matcher::match_startmark), - &perl_matcher::match_endmark, - &perl_matcher::match_literal, - &perl_matcher::match_start_line, - &perl_matcher::match_end_line, - &perl_matcher::match_wild, - &perl_matcher::match_match, - &perl_matcher::match_word_boundary, - &perl_matcher::match_within_word, - &perl_matcher::match_word_start, - &perl_matcher::match_word_end, - &perl_matcher::match_buffer_start, - &perl_matcher::match_buffer_end, - &perl_matcher::match_backref, - &perl_matcher::match_long_set, - &perl_matcher::match_set, - &perl_matcher::match_jump, - &perl_matcher::match_alt, - &perl_matcher::match_rep, - &perl_matcher::match_combining, - &perl_matcher::match_soft_buffer_end, - &perl_matcher::match_restart_continue, - // Although this next line *should* be evaluated at compile time, in practice - // some compilers (VC++) emit run-time initialisation which breaks thread - // safety, so use a dispatch function instead: - //(::cutl_details_boost::is_random_access_iterator::value ? &perl_matcher::match_dot_repeat_fast : &perl_matcher::match_dot_repeat_slow), - &perl_matcher::match_dot_repeat_dispatch, - &perl_matcher::match_char_repeat, - &perl_matcher::match_set_repeat, - &perl_matcher::match_long_set_repeat, - &perl_matcher::match_backstep, - &perl_matcher::match_assert_backref, - &perl_matcher::match_toggle_case, - &perl_matcher::match_recursion, - }; - - if(state_count > max_state_count) - raise_error(traits_inst, regex_constants::error_complexity); - while(pstate) - { - matcher_proc_type proc = s_match_vtable[pstate->type]; - ++state_count; - if(!(this->*proc)()) - { - if((m_match_flags & match_partial) && (position == last) && (position != search_base)) - m_has_partial_match = true; - return 0; - } - } - return true; -} - -template -bool perl_matcher::match_startmark() -{ - int index = static_cast(pstate)->index; - icase = static_cast(pstate)->icase; - bool r = true; - switch(index) - { - case 0: - pstate = pstate->next.p; - break; - case -1: - case -2: - { - // forward lookahead assert: - BidiIterator old_position(position); - const re_syntax_base* next_pstate = static_cast(pstate->next.p)->alt.p->next.p; - pstate = pstate->next.p->next.p; - r = match_all_states(); - pstate = next_pstate; - position = old_position; - if((r && (index != -1)) || (!r && (index != -2))) - r = false; - else - r = true; - break; - } - case -3: - { - // independent sub-expression: - bool old_independent = m_independent; - m_independent = true; - const re_syntax_base* next_pstate = static_cast(pstate->next.p)->alt.p->next.p; - pstate = pstate->next.p->next.p; - r = match_all_states(); - pstate = next_pstate; - m_independent = old_independent; -#ifdef BOOST_REGEX_MATCH_EXTRA - if(r && (m_match_flags & match_extra)) - { - // - // our captures have been stored in *m_presult - // we need to unpack them, and insert them - // back in the right order when we unwind the stack: - // - unsigned i; - match_results tm(*m_presult); - for(i = 0; i < tm.size(); ++i) - (*m_presult)[i].get_captures().clear(); - // match everything else: - r = match_all_states(); - // now place the stored captures back: - for(i = 0; i < tm.size(); ++i) - { - typedef typename sub_match::capture_sequence_type seq; - seq& s1 = (*m_presult)[i].get_captures(); - const seq& s2 = tm[i].captures(); - s1.insert( - s1.end(), - s2.begin(), - s2.end()); - } - } -#endif - break; - } - case -4: - { - // conditional expression: - const re_alt* alt = static_cast(pstate->next.p); - BOOST_ASSERT(alt->type == syntax_element_alt); - pstate = alt->next.p; - if(pstate->type == syntax_element_assert_backref) - { - if(!match_assert_backref()) - pstate = alt->alt.p; - break; - } - else - { - // zero width assertion, have to match this recursively: - BOOST_ASSERT(pstate->type == syntax_element_startmark); - bool negated = static_cast(pstate)->index == -2; - BidiIterator saved_position = position; - const re_syntax_base* next_pstate = static_cast(pstate->next.p)->alt.p->next.p; - pstate = pstate->next.p->next.p; - bool res = match_all_states(); - position = saved_position; - if(negated) - res = !res; - if(res) - pstate = next_pstate; - else - pstate = alt->alt.p; - break; - } - } - case -5: - { - // Reset start of $0, since we have a \K escape - backup_subex sub(*m_presult, 0); - m_presult->set_first(position, 0, true); - pstate = pstate->next.p; - r = match_all_states(); - if(r == false) - sub.restore(*m_presult); - break; - } - default: - { - BOOST_ASSERT(index > 0); - if((m_match_flags & match_nosubs) == 0) - { - backup_subex sub(*m_presult, index); - m_presult->set_first(position, index); - pstate = pstate->next.p; - r = match_all_states(); - if(r == false) - sub.restore(*m_presult); -#ifdef BOOST_REGEX_MATCH_EXTRA - // - // we have a match, push the capture information onto the stack: - // - else if(sub.get().matched && (match_extra & m_match_flags)) - ((*m_presult)[index]).get_captures().push_back(sub.get()); -#endif - } - else - { - pstate = pstate->next.p; - } - break; - } - } - return r; -} - -template -bool perl_matcher::match_alt() -{ - bool take_first, take_second; - const re_alt* jmp = static_cast(pstate); - - // find out which of these two alternatives we need to take: - if(position == last) - { - take_first = jmp->can_be_null & mask_take; - take_second = jmp->can_be_null & mask_skip; - } - else - { - take_first = can_start(*position, jmp->_map, (unsigned char)mask_take); - take_second = can_start(*position, jmp->_map, (unsigned char)mask_skip); - } - - if(take_first) - { - // we can take the first alternative, - // see if we need to push next alternative: - if(take_second) - { - BidiIterator oldposition(position); - const re_syntax_base* old_pstate = jmp->alt.p; - pstate = pstate->next.p; - if(!match_all_states()) - { - pstate = old_pstate; - position = oldposition; - } - return true; - } - pstate = pstate->next.p; - return true; - } - if(take_second) - { - pstate = jmp->alt.p; - return true; - } - return false; // neither option is possible -} - -template -bool perl_matcher::match_rep() -{ -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable:4127 4244) -#endif - const re_repeat* rep = static_cast(pstate); - // - // Always copy the repeat count, so that the state is restored - // when we exit this scope: - // - repeater_count r(rep->state_id, &next_count, position); - // - // If we've had at least one repeat already, and the last one - // matched the NULL string then set the repeat count to - // maximum: - // - next_count->check_null_repeat(position, rep->max); - - // find out which of these two alternatives we need to take: - bool take_first, take_second; - if(position == last) - { - take_first = rep->can_be_null & mask_take; - take_second = rep->can_be_null & mask_skip; - } - else - { - take_first = can_start(*position, rep->_map, (unsigned char)mask_take); - take_second = can_start(*position, rep->_map, (unsigned char)mask_skip); - } - - if(next_count->get_count() < rep->min) - { - // we must take the repeat: - if(take_first) - { - // increase the counter: - ++(*next_count); - pstate = rep->next.p; - return match_all_states(); - } - return false; - } - bool greedy = (rep->greedy) && (!(m_match_flags & regex_constants::match_any) || m_independent); - if(greedy) - { - // try and take the repeat if we can: - if((next_count->get_count() < rep->max) && take_first) - { - // store position in case we fail: - BidiIterator pos = position; - // increase the counter: - ++(*next_count); - pstate = rep->next.p; - if(match_all_states()) - return true; - // failed repeat, reset posistion and fall through for alternative: - position = pos; - } - if(take_second) - { - pstate = rep->alt.p; - return true; - } - return false; // can't take anything, fail... - } - else // non-greedy - { - // try and skip the repeat if we can: - if(take_second) - { - // store position in case we fail: - BidiIterator pos = position; - pstate = rep->alt.p; - if(match_all_states()) - return true; - // failed alternative, reset posistion and fall through for repeat: - position = pos; - } - if((next_count->get_count() < rep->max) && take_first) - { - // increase the counter: - ++(*next_count); - pstate = rep->next.p; - return match_all_states(); - } - } - return false; -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif -} - -template -bool perl_matcher::match_dot_repeat_slow() -{ -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable:4127) -#endif - unsigned count = 0; - const re_repeat* rep = static_cast(pstate); - re_syntax_base* psingle = rep->next.p; - // match compulsary repeats first: - while(count < rep->min) - { - pstate = psingle; - if(!match_wild()) - return false; - ++count; - } - bool greedy = (rep->greedy) && (!(m_match_flags & regex_constants::match_any) || m_independent); - if(greedy) - { - // normal repeat: - while(count < rep->max) - { - pstate = psingle; - if(!match_wild()) - break; - ++count; - } - if((rep->leading) && (count < rep->max)) - restart = position; - pstate = rep; - return backtrack_till_match(count - rep->min); - } - else - { - // non-greedy, keep trying till we get a match: - BidiIterator save_pos; - do - { - if((rep->leading) && (rep->max == UINT_MAX)) - restart = position; - pstate = rep->alt.p; - save_pos = position; - ++state_count; - if(match_all_states()) - return true; - if(count >= rep->max) - return false; - ++count; - pstate = psingle; - position = save_pos; - if(!match_wild()) - return false; - }while(true); - } -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif -} - -template -bool perl_matcher::match_dot_repeat_fast() -{ -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable:4127) -#endif - if(m_match_flags & match_not_dot_null) - return match_dot_repeat_slow(); - if((static_cast(pstate->next.p)->mask & match_any_mask) == 0) - return match_dot_repeat_slow(); - // - // start by working out how much we can skip: - // - const re_repeat* rep = static_cast(pstate); -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable:4267) -#endif - bool greedy = (rep->greedy) && (!(m_match_flags & regex_constants::match_any) || m_independent); - std::size_t count = (std::min)(static_cast(::cutl_details_boost::re_detail::distance(position, last)), static_cast(greedy ? rep->max : rep->min)); - if(rep->min > count) - { - position = last; - return false; // not enough text left to match - } - std::advance(position, count); -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif - if((rep->leading) && (count < rep->max) && greedy) - restart = position; - if(greedy) - return backtrack_till_match(count - rep->min); - - // non-greedy, keep trying till we get a match: - BidiIterator save_pos; - do - { - while((position != last) && (count < rep->max) && !can_start(*position, rep->_map, mask_skip)) - { - ++position; - ++count; - } - if((rep->leading) && (rep->max == UINT_MAX)) - restart = position; - pstate = rep->alt.p; - save_pos = position; - ++state_count; - if(match_all_states()) - return true; - if(count >= rep->max) - return false; - if(save_pos == last) - return false; - position = ++save_pos; - ++count; - }while(true); -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif -} - -template -bool perl_matcher::match_char_repeat() -{ -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable:4127) -#pragma warning(disable:4267) -#endif -#ifdef __BORLANDC__ -#pragma option push -w-8008 -w-8066 -w-8004 -#endif - const re_repeat* rep = static_cast(pstate); - BOOST_ASSERT(1 == static_cast(rep->next.p)->length); - const char_type what = *reinterpret_cast(static_cast(rep->next.p) + 1); - // - // start by working out how much we can skip: - // - bool greedy = (rep->greedy) && (!(m_match_flags & regex_constants::match_any) || m_independent); - std::size_t count, desired; - if(::cutl_details_boost::is_random_access_iterator::value) - { - desired = - (std::min)( - (std::size_t)(greedy ? rep->max : rep->min), - (std::size_t)::cutl_details_boost::re_detail::distance(position, last)); - count = desired; - ++desired; - if(icase) - { - while(--desired && (traits_inst.translate_nocase(*position) == what)) - { - ++position; - } - } - else - { - while(--desired && (traits_inst.translate(*position) == what)) - { - ++position; - } - } - count = count - desired; - } - else - { - count = 0; - desired = greedy ? rep->max : rep->min; - while((count < desired) && (position != last) && (traits_inst.translate(*position, icase) == what)) - { - ++position; - ++count; - } - } - if((rep->leading) && (count < rep->max) && greedy) - restart = position; - if(count < rep->min) - return false; - - if(greedy) - return backtrack_till_match(count - rep->min); - - // non-greedy, keep trying till we get a match: - BidiIterator save_pos; - do - { - while((position != last) && (count < rep->max) && !can_start(*position, rep->_map, mask_skip)) - { - if((traits_inst.translate(*position, icase) == what)) - { - ++position; - ++count; - } - else - return false; // counldn't repeat even though it was the only option - } - if((rep->leading) && (rep->max == UINT_MAX)) - restart = position; - pstate = rep->alt.p; - save_pos = position; - ++state_count; - if(match_all_states()) - return true; - if(count >= rep->max) - return false; - position = save_pos; - if(position == last) - return false; - if(traits_inst.translate(*position, icase) == what) - { - ++position; - ++count; - } - else - { - return false; - } - }while(true); -#ifdef __BORLANDC__ -#pragma option pop -#endif -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif -} - -template -bool perl_matcher::match_set_repeat() -{ -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable:4127) -#endif -#ifdef __BORLANDC__ -#pragma option push -w-8008 -w-8066 -w-8004 -#endif - const re_repeat* rep = static_cast(pstate); - const unsigned char* map = static_cast(rep->next.p)->_map; - unsigned count = 0; - // - // start by working out how much we can skip: - // - bool greedy = (rep->greedy) && (!(m_match_flags & regex_constants::match_any) || m_independent); - std::size_t desired = greedy ? rep->max : rep->min; - if(::cutl_details_boost::is_random_access_iterator::value) - { - BidiIterator end = position; - std::advance(end, (std::min)((std::size_t)::cutl_details_boost::re_detail::distance(position, last), desired)); - BidiIterator origin(position); - while((position != end) && map[static_cast(traits_inst.translate(*position, icase))]) - { - ++position; - } - count = (unsigned)::cutl_details_boost::re_detail::distance(origin, position); - } - else - { - while((count < desired) && (position != last) && map[static_cast(traits_inst.translate(*position, icase))]) - { - ++position; - ++count; - } - } - if((rep->leading) && (count < rep->max) && greedy) - restart = position; - if(count < rep->min) - return false; - - if(greedy) - return backtrack_till_match(count - rep->min); - - // non-greedy, keep trying till we get a match: - BidiIterator save_pos; - do - { - while((position != last) && (count < rep->max) && !can_start(*position, rep->_map, mask_skip)) - { - if(map[static_cast(traits_inst.translate(*position, icase))]) - { - ++position; - ++count; - } - else - return false; // counldn't repeat even though it was the only option - } - if((rep->leading) && (rep->max == UINT_MAX)) - restart = position; - pstate = rep->alt.p; - save_pos = position; - ++state_count; - if(match_all_states()) - return true; - if(count >= rep->max) - return false; - position = save_pos; - if(position == last) - return false; - if(map[static_cast(traits_inst.translate(*position, icase))]) - { - ++position; - ++count; - } - else - { - return false; - } - }while(true); -#ifdef __BORLANDC__ -#pragma option pop -#endif -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif -} - -template -bool perl_matcher::match_long_set_repeat() -{ -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable:4127) -#endif -#ifdef __BORLANDC__ -#pragma option push -w-8008 -w-8066 -w-8004 -#endif - typedef typename traits::char_class_type char_class_type; - const re_repeat* rep = static_cast(pstate); - const re_set_long* set = static_cast*>(pstate->next.p); - unsigned count = 0; - // - // start by working out how much we can skip: - // - bool greedy = (rep->greedy) && (!(m_match_flags & regex_constants::match_any) || m_independent); - std::size_t desired = greedy ? rep->max : rep->min; - if(::cutl_details_boost::is_random_access_iterator::value) - { - BidiIterator end = position; - std::advance(end, (std::min)((std::size_t)::cutl_details_boost::re_detail::distance(position, last), desired)); - BidiIterator origin(position); - while((position != end) && (position != re_is_set_member(position, last, set, re.get_data(), icase))) - { - ++position; - } - count = (unsigned)::cutl_details_boost::re_detail::distance(origin, position); - } - else - { - while((count < desired) && (position != last) && (position != re_is_set_member(position, last, set, re.get_data(), icase))) - { - ++position; - ++count; - } - } - if((rep->leading) && (count < rep->max) && greedy) - restart = position; - if(count < rep->min) - return false; - - if(greedy) - return backtrack_till_match(count - rep->min); - - // non-greedy, keep trying till we get a match: - BidiIterator save_pos; - do - { - while((position != last) && (count < rep->max) && !can_start(*position, rep->_map, mask_skip)) - { - if(position != re_is_set_member(position, last, set, re.get_data(), icase)) - { - ++position; - ++count; - } - else - return false; // counldn't repeat even though it was the only option - } - if((rep->leading) && (rep->max == UINT_MAX)) - restart = position; - pstate = rep->alt.p; - save_pos = position; - ++state_count; - if(match_all_states()) - return true; - if(count >= rep->max) - return false; - position = save_pos; - if(position == last) - return false; - if(position != re_is_set_member(position, last, set, re.get_data(), icase)) - { - ++position; - ++count; - } - else - { - return false; - } - }while(true); -#ifdef __BORLANDC__ -#pragma option pop -#endif -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif -} - -template -bool perl_matcher::backtrack_till_match(std::size_t count) -{ -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable:4127) -#endif - if((m_match_flags & match_partial) && (position == last)) - m_has_partial_match = true; - - const re_repeat* rep = static_cast(pstate); - BidiIterator backtrack = position; - if(position == last) - { - if(rep->can_be_null & mask_skip) - { - pstate = rep->alt.p; - if(match_all_states()) - return true; - } - if(count) - { - position = --backtrack; - --count; - } - else - return false; - } - do - { - while(count && !can_start(*position, rep->_map, mask_skip)) - { - --position; - --count; - ++state_count; - } - pstate = rep->alt.p; - backtrack = position; - if(match_all_states()) - return true; - if(count == 0) - return false; - position = --backtrack; - ++state_count; - --count; - }while(true); -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif -} - -template -bool perl_matcher::match_recursion() -{ - BOOST_ASSERT(pstate->type == syntax_element_recurse); - // - // Set new call stack: - // - if(recursion_stack.capacity() == 0) - { - recursion_stack.reserve(50); - } - recursion_stack.push_back(recursion_info()); - recursion_stack.back().preturn_address = pstate->next.p; - recursion_stack.back().results = *m_presult; - recursion_stack.back().repeater_stack = next_count; - pstate = static_cast(pstate)->alt.p; - recursion_stack.back().idx = static_cast(pstate)->index; - - repeater_count* saved = next_count; - repeater_count r(&next_count); // resets all repeat counts since we're recursing and starting fresh on those - next_count = &r; - bool result = match_all_states(); - next_count = saved; - - if(!result) - { - next_count = recursion_stack.back().repeater_stack; - *m_presult = recursion_stack.back().results; - recursion_stack.pop_back(); - return false; - } - return true; -} - -template -bool perl_matcher::match_endmark() -{ - int index = static_cast(pstate)->index; - icase = static_cast(pstate)->icase; - if(index > 0) - { - if((m_match_flags & match_nosubs) == 0) - { - m_presult->set_second(position, index); - } - if(!recursion_stack.empty()) - { - if(index == recursion_stack.back().idx) - { - recursion_info saved = recursion_stack.back(); - recursion_stack.pop_back(); - pstate = saved.preturn_address; - repeater_count* saved_count = next_count; - next_count = saved.repeater_stack; - *m_presult = saved.results; - if(!match_all_states()) - { - recursion_stack.push_back(saved); - next_count = saved_count; - return false; - } - } - } - } - else if((index < 0) && (index != -4)) - { - // matched forward lookahead: - pstate = 0; - return true; - } - pstate = pstate ? pstate->next.p : 0; - return true; -} - -template -bool perl_matcher::match_match() -{ - if(!recursion_stack.empty()) - { - BOOST_ASSERT(0 == recursion_stack.back().idx); - const re_syntax_base* saved_state = pstate = recursion_stack.back().preturn_address; - *m_presult = recursion_stack.back().results; - recursion_stack.pop_back(); - if(!match_all_states()) - { - recursion_stack.push_back(recursion_info()); - recursion_stack.back().preturn_address = saved_state; - recursion_stack.back().results = *m_presult; - return false; - } - return true; - } - if((m_match_flags & match_not_null) && (position == (*m_presult)[0].first)) - return false; - if((m_match_flags & match_all) && (position != last)) - return false; - if((m_match_flags & regex_constants::match_not_initial_null) && (position == search_base)) - return false; - m_presult->set_second(position); - pstate = 0; - m_has_found_match = true; - if((m_match_flags & match_posix) == match_posix) - { - m_result.maybe_assign(*m_presult); - if((m_match_flags & match_any) == 0) - return false; - } -#ifdef BOOST_REGEX_MATCH_EXTRA - if(match_extra & m_match_flags) - { - for(unsigned i = 0; i < m_presult->size(); ++i) - if((*m_presult)[i].matched) - ((*m_presult)[i]).get_captures().push_back((*m_presult)[i]); - } -#endif - return true; -} - - - -} // namespace re_detail -} // namespace cutl_details_boost -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif - -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable: 4103) -#endif -#ifdef BOOST_HAS_ABI_HEADERS -# include BOOST_ABI_SUFFIX -#endif -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif - -#endif - diff --git a/cutl/details/boost/regex/v4/primary_transform.hpp b/cutl/details/boost/regex/v4/primary_transform.hpp deleted file mode 100644 index ae59e5b..0000000 --- a/cutl/details/boost/regex/v4/primary_transform.hpp +++ /dev/null @@ -1,146 +0,0 @@ -/* - * - * Copyright (c) 1998-2002 - * John Maddock - * - * Use, modification and distribution are subject to the - * Boost Software License, Version 1.0. (See accompanying file - * LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt) - * - */ - - /* - * LOCATION: see http://www.boost.org for most recent version. - * FILE: primary_transform.hpp - * VERSION: see - * DESCRIPTION: Heuristically determines the sort string format in use - * by the current locale. - */ - -#ifndef BOOST_REGEX_PRIMARY_TRANSFORM -#define BOOST_REGEX_PRIMARY_TRANSFORM - -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable: 4103) -#endif -#ifdef BOOST_HAS_ABI_HEADERS -# include BOOST_ABI_PREFIX -#endif -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif - -namespace cutl_details_boost{ - namespace re_detail{ - - -enum{ - sort_C, - sort_fixed, - sort_delim, - sort_unknown -}; - -template -unsigned count_chars(const S& s, charT c) -{ - // - // Count how many occurances of character c occur - // in string s: if c is a delimeter between collation - // fields, then this should be the same value for all - // sort keys: - // - unsigned int count = 0; - for(unsigned pos = 0; pos < s.size(); ++pos) - { - if(s[pos] == c) ++count; - } - return count; -} - - -template -unsigned find_sort_syntax(const traits* pt, charT* delim) -{ - // - // compare 'a' with 'A' to see how similar they are, - // should really use a-accute but we can't portably do that, - // - typedef typename traits::string_type string_type; - typedef typename traits::char_type char_type; - - // Suppress incorrect warning for MSVC - (void)pt; - - char_type a[2] = {'a', '\0', }; - string_type sa(pt->transform(a, a+1)); - if(sa == a) - { - *delim = 0; - return sort_C; - } - char_type A[2] = { 'A', '\0', }; - string_type sA(pt->transform(A, A+1)); - char_type c[2] = { ';', '\0', }; - string_type sc(pt->transform(c, c+1)); - - int pos = 0; - while((pos <= static_cast(sa.size())) && (pos <= static_cast(sA.size())) && (sa[pos] == sA[pos])) ++pos; - --pos; - if(pos < 0) - { - *delim = 0; - return sort_unknown; - } - // - // at this point sa[pos] is either the end of a fixed width field - // or the character that acts as a delimiter: - // - charT maybe_delim = sa[pos]; - if((pos != 0) && (count_chars(sa, maybe_delim) == count_chars(sA, maybe_delim)) && (count_chars(sa, maybe_delim) == count_chars(sc, maybe_delim))) - { - *delim = maybe_delim; - return sort_delim; - } - // - // OK doen't look like a delimiter, try for fixed width field: - // - if((sa.size() == sA.size()) && (sa.size() == sc.size())) - { - // note assumes that the fixed width field is less than - // (numeric_limits::max)(), should be true for all types - // I can't imagine 127 character fields... - *delim = static_cast(++pos); - return sort_fixed; - } - // - // don't know what it is: - // - *delim = 0; - return sort_unknown; -} - - - } // namespace re_detail -} // namespace cutl_details_boost - -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable: 4103) -#endif -#ifdef BOOST_HAS_ABI_HEADERS -# include BOOST_ABI_SUFFIX -#endif -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif - -#endif - - - - - - - diff --git a/cutl/details/boost/regex/v4/protected_call.hpp b/cutl/details/boost/regex/v4/protected_call.hpp deleted file mode 100644 index fbf2a31..0000000 --- a/cutl/details/boost/regex/v4/protected_call.hpp +++ /dev/null @@ -1,81 +0,0 @@ -/* - * - * Copyright (c) 2004 - * John Maddock - * - * Use, modification and distribution are subject to the - * Boost Software License, Version 1.0. (See accompanying file - * LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt) - * - */ - - /* - * LOCATION: see http://www.boost.org for most recent version. - * FILE basic_regex_creator.cpp - * VERSION see - * DESCRIPTION: Declares template class basic_regex_creator which fills in - * the data members of a regex_data object. - */ - -#ifndef BOOST_REGEX_V4_PROTECTED_CALL_HPP -#define BOOST_REGEX_V4_PROTECTED_CALL_HPP - -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable: 4103) -#endif -#ifdef BOOST_HAS_ABI_HEADERS -# include BOOST_ABI_PREFIX -#endif -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif - -namespace cutl_details_boost{ -namespace re_detail{ - -class BOOST_REGEX_DECL abstract_protected_call -{ -public: - bool BOOST_REGEX_CALL execute()const; - // this stops gcc-4 from complaining: - virtual ~abstract_protected_call(){} -private: - virtual bool call()const = 0; -}; - -template -class concrete_protected_call - : public abstract_protected_call -{ -public: - typedef bool (T::*proc_type)(); - concrete_protected_call(T* o, proc_type p) - : obj(o), proc(p) {} -private: - virtual bool call()const; - T* obj; - proc_type proc; -}; - -template -bool concrete_protected_call::call()const -{ - return (obj->*proc)(); -} - -} -} // namespace cutl_details_boost - -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable: 4103) -#endif -#ifdef BOOST_HAS_ABI_HEADERS -# include BOOST_ABI_SUFFIX -#endif -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif - -#endif diff --git a/cutl/details/boost/regex/v4/regbase.hpp b/cutl/details/boost/regex/v4/regbase.hpp deleted file mode 100644 index 0c6f7ae..0000000 --- a/cutl/details/boost/regex/v4/regbase.hpp +++ /dev/null @@ -1,180 +0,0 @@ -/* - * - * Copyright (c) 1998-2002 - * John Maddock - * - * Use, modification and distribution are subject to the - * Boost Software License, Version 1.0. (See accompanying file - * LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt) - * - */ - - /* - * LOCATION: see http://www.boost.org for most recent version. - * FILE regbase.cpp - * VERSION see - * DESCRIPTION: Declares class regbase. - */ - -#ifndef BOOST_REGEX_V4_REGBASE_HPP -#define BOOST_REGEX_V4_REGBASE_HPP - -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable: 4103) -#endif -#ifdef BOOST_HAS_ABI_HEADERS -# include BOOST_ABI_PREFIX -#endif -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif - -namespace cutl_details_boost{ -// -// class regbase -// handles error codes and flags -// -class BOOST_REGEX_DECL regbase -{ -public: - enum flag_type_ - { - // - // Divide the flags up into logical groups: - // bits 0-7 indicate main synatx type. - // bits 8-15 indicate syntax subtype. - // bits 16-31 indicate options that are common to all - // regex syntaxes. - // In all cases the default is 0. - // - // Main synatx group: - // - perl_syntax_group = 0, // default - basic_syntax_group = 1, // POSIX basic - literal = 2, // all characters are literals - main_option_type = literal | basic_syntax_group | perl_syntax_group, // everything! - // - // options specific to perl group: - // - no_bk_refs = 1 << 8, // \d not allowed - no_perl_ex = 1 << 9, // disable perl extensions - no_mod_m = 1 << 10, // disable Perl m modifier - mod_x = 1 << 11, // Perl x modifier - mod_s = 1 << 12, // force s modifier on (overrides match_not_dot_newline) - no_mod_s = 1 << 13, // force s modifier off (overrides match_not_dot_newline) - - // - // options specific to basic group: - // - no_char_classes = 1 << 8, // [[:CLASS:]] not allowed - no_intervals = 1 << 9, // {x,y} not allowed - bk_plus_qm = 1 << 10, // uses \+ and \? - bk_vbar = 1 << 11, // use \| for alternatives - emacs_ex = 1 << 12, // enables emacs extensions - - // - // options common to all groups: - // - no_escape_in_lists = 1 << 16, // '\' not special inside [...] - newline_alt = 1 << 17, // \n is the same as | - no_except = 1 << 18, // no exception on error - failbit = 1 << 19, // error flag - icase = 1 << 20, // characters are matched regardless of case - nocollate = 0, // don't use locale specific collation (deprecated) - collate = 1 << 21, // use locale specific collation - nosubs = 1 << 22, // don't mark sub-expressions - save_subexpression_location = 1 << 23, // save subexpression locations - no_empty_expressions = 1 << 24, // no empty expressions allowed - optimize = 0, // not really supported - - - - basic = basic_syntax_group | collate | no_escape_in_lists, - extended = no_bk_refs | collate | no_perl_ex | no_escape_in_lists, - normal = 0, - emacs = basic_syntax_group | collate | emacs_ex | bk_vbar, - awk = no_bk_refs | collate | no_perl_ex, - grep = basic | newline_alt, - egrep = extended | newline_alt, - sed = basic, - perl = normal, - ECMAScript = normal, - JavaScript = normal, - JScript = normal - }; - typedef unsigned int flag_type; - - enum restart_info - { - restart_any = 0, - restart_word = 1, - restart_line = 2, - restart_buf = 3, - restart_continue = 4, - restart_lit = 5, - restart_fixed_lit = 6, - restart_count = 7 - }; -}; - -// -// provide std lib proposal compatible constants: -// -namespace regex_constants{ - - enum flag_type_ - { - - no_except = ::cutl_details_boost::regbase::no_except, - failbit = ::cutl_details_boost::regbase::failbit, - literal = ::cutl_details_boost::regbase::literal, - icase = ::cutl_details_boost::regbase::icase, - nocollate = ::cutl_details_boost::regbase::nocollate, - collate = ::cutl_details_boost::regbase::collate, - nosubs = ::cutl_details_boost::regbase::nosubs, - optimize = ::cutl_details_boost::regbase::optimize, - bk_plus_qm = ::cutl_details_boost::regbase::bk_plus_qm, - bk_vbar = ::cutl_details_boost::regbase::bk_vbar, - no_intervals = ::cutl_details_boost::regbase::no_intervals, - no_char_classes = ::cutl_details_boost::regbase::no_char_classes, - no_escape_in_lists = ::cutl_details_boost::regbase::no_escape_in_lists, - no_mod_m = ::cutl_details_boost::regbase::no_mod_m, - mod_x = ::cutl_details_boost::regbase::mod_x, - mod_s = ::cutl_details_boost::regbase::mod_s, - no_mod_s = ::cutl_details_boost::regbase::no_mod_s, - save_subexpression_location = ::cutl_details_boost::regbase::save_subexpression_location, - no_empty_expressions = ::cutl_details_boost::regbase::no_empty_expressions, - - basic = ::cutl_details_boost::regbase::basic, - extended = ::cutl_details_boost::regbase::extended, - normal = ::cutl_details_boost::regbase::normal, - emacs = ::cutl_details_boost::regbase::emacs, - awk = ::cutl_details_boost::regbase::awk, - grep = ::cutl_details_boost::regbase::grep, - egrep = ::cutl_details_boost::regbase::egrep, - sed = basic, - perl = normal, - ECMAScript = normal, - JavaScript = normal, - JScript = normal - }; - typedef ::cutl_details_boost::regbase::flag_type syntax_option_type; - -} // namespace regex_constants - -} // namespace cutl_details_boost - -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable: 4103) -#endif -#ifdef BOOST_HAS_ABI_HEADERS -# include BOOST_ABI_SUFFIX -#endif -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif - -#endif - diff --git a/cutl/details/boost/regex/v4/regex.hpp b/cutl/details/boost/regex/v4/regex.hpp deleted file mode 100644 index a442970..0000000 --- a/cutl/details/boost/regex/v4/regex.hpp +++ /dev/null @@ -1,202 +0,0 @@ -/* - * - * Copyright (c) 1998-2002 - * John Maddock - * - * Use, modification and distribution are subject to the - * Boost Software License, Version 1.0. (See accompanying file - * LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt) - * - */ - - /* - * LOCATION: see http://www.boost.org for most recent version. - * FILE regex.cpp - * VERSION see - * DESCRIPTION: Declares cutl_details_boost::basic_regex<> and associated - * functions and classes. This header is the main - * entry point for the template regex code. - */ - -#ifndef BOOST_RE_REGEX_HPP_INCLUDED -#define BOOST_RE_REGEX_HPP_INCLUDED - -#ifdef __cplusplus - -// what follows is all C++ don't include in C builds!! - -#ifndef BOOST_REGEX_CONFIG_HPP -#include -#endif -#ifndef BOOST_REGEX_WORKAROUND_HPP -#include -#endif - -#ifndef BOOST_REGEX_FWD_HPP -#include -#endif -#ifndef BOOST_REGEX_TRAITS_HPP -#include -#endif -#ifndef BOOST_REGEX_RAW_BUFFER_HPP -#include -#endif -#ifndef BOOST_REGEX_V4_MATCH_FLAGS -#include -#endif -#ifndef BOOST_REGEX_RAW_BUFFER_HPP -#include -#endif -#ifndef BOOST_RE_PAT_EXCEPT_HPP -#include -#endif - -#ifndef BOOST_REGEX_V4_CHAR_REGEX_TRAITS_HPP -#include -#endif -#ifndef BOOST_REGEX_V4_STATES_HPP -#include -#endif -#ifndef BOOST_REGEX_V4_REGBASE_HPP -#include -#endif -#ifndef BOOST_REGEX_V4_ITERATOR_TRAITS_HPP -#include -#endif -#ifndef BOOST_REGEX_V4_BASIC_REGEX_HPP -#include -#endif -#ifndef BOOST_REGEX_V4_BASIC_REGEX_CREATOR_HPP -#include -#endif -#ifndef BOOST_REGEX_V4_BASIC_REGEX_PARSER_HPP -#include -#endif -#ifndef BOOST_REGEX_V4_SUB_MATCH_HPP -#include -#endif -#ifndef BOOST_REGEX_FORMAT_HPP -#include -#endif -#ifndef BOOST_REGEX_V4_MATCH_RESULTS_HPP -#include -#endif -#ifndef BOOST_REGEX_V4_PROTECTED_CALL_HPP -#include -#endif -#ifndef BOOST_REGEX_MATCHER_HPP -#include -#endif -// -// template instances: -// -#define BOOST_REGEX_CHAR_T char -#ifdef BOOST_REGEX_NARROW_INSTANTIATE -# define BOOST_REGEX_INSTANTIATE -#endif -#include -#undef BOOST_REGEX_CHAR_T -#ifdef BOOST_REGEX_INSTANTIATE -# undef BOOST_REGEX_INSTANTIATE -#endif - -#ifndef BOOST_NO_WREGEX -#define BOOST_REGEX_CHAR_T wchar_t -#ifdef BOOST_REGEX_WIDE_INSTANTIATE -# define BOOST_REGEX_INSTANTIATE -#endif -#include -#undef BOOST_REGEX_CHAR_T -#ifdef BOOST_REGEX_INSTANTIATE -# undef BOOST_REGEX_INSTANTIATE -#endif -#endif - -#if !defined(BOOST_NO_WREGEX) && defined(BOOST_REGEX_HAS_OTHER_WCHAR_T) -#define BOOST_REGEX_CHAR_T unsigned short -#ifdef BOOST_REGEX_US_INSTANTIATE -# define BOOST_REGEX_INSTANTIATE -#endif -#include -#undef BOOST_REGEX_CHAR_T -#ifdef BOOST_REGEX_INSTANTIATE -# undef BOOST_REGEX_INSTANTIATE -#endif -#endif - - -namespace cutl_details_boost{ -#ifdef BOOST_REGEX_NO_FWD -typedef basic_regex > regex; -#ifndef BOOST_NO_WREGEX -typedef basic_regex > wregex; -#endif -#endif - -typedef match_results cmatch; -typedef match_results smatch; -#ifndef BOOST_NO_WREGEX -typedef match_results wcmatch; -typedef match_results wsmatch; -#endif - -} // namespace cutl_details_boost -#ifndef BOOST_REGEX_MATCH_HPP -#include -#endif -#ifndef BOOST_REGEX_V4_REGEX_SEARCH_HPP -#include -#endif -#ifndef BOOST_REGEX_ITERATOR_HPP -#include -#endif -#ifndef BOOST_REGEX_TOKEN_ITERATOR_HPP -#include -#endif -#ifndef BOOST_REGEX_V4_REGEX_GREP_HPP -#include -#endif -#ifndef BOOST_REGEX_V4_REGEX_REPLACE_HPP -#include -#endif -#ifndef BOOST_REGEX_V4_REGEX_MERGE_HPP -#include -#endif -#ifndef BOOST_REGEX_SPLIT_HPP -#include -#endif - -#endif // __cplusplus - -#endif // include - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - diff --git a/cutl/details/boost/regex/v4/regex_format.hpp b/cutl/details/boost/regex/v4/regex_format.hpp deleted file mode 100644 index c4a60f9..0000000 --- a/cutl/details/boost/regex/v4/regex_format.hpp +++ /dev/null @@ -1,1156 +0,0 @@ -/* - * - * Copyright (c) 1998-2009 John Maddock - * Copyright 2008 Eric Niebler. - * - * Use, modification and distribution are subject to the - * Boost Software License, Version 1.0. (See accompanying file - * LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt) - * - */ - - /* - * LOCATION: see http://www.boost.org for most recent version. - * FILE regex_format.hpp - * VERSION see - * DESCRIPTION: Provides formatting output routines for search and replace - * operations. Note this is an internal header file included - * by regex.hpp, do not include on its own. - */ - -#ifndef BOOST_REGEX_FORMAT_HPP -#define BOOST_REGEX_FORMAT_HPP - -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#ifndef BOOST_NO_SFINAE -#include -#endif -#include - -namespace cutl_details_boost{ - -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable: 4103) -#endif -#ifdef BOOST_HAS_ABI_HEADERS -# include BOOST_ABI_PREFIX -#endif -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif - -// -// Forward declaration: -// - template >::allocator_type > -class match_results; - -namespace re_detail{ - -// -// struct trivial_format_traits: -// defines minimum localisation support for formatting -// in the case that the actual regex traits is unavailable. -// -template -struct trivial_format_traits -{ - typedef charT char_type; - - static std::ptrdiff_t length(const charT* p) - { - return global_length(p); - } - static charT tolower(charT c) - { - return ::cutl_details_boost::re_detail::global_lower(c); - } - static charT toupper(charT c) - { - return ::cutl_details_boost::re_detail::global_upper(c); - } - static int value(const charT c, int radix) - { - int result = global_value(c); - return result >= radix ? -1 : result; - } - int toi(const charT*& p1, const charT* p2, int radix)const - { - return global_toi(p1, p2, radix, *this); - } -}; - -template -class basic_regex_formatter -{ -public: - typedef typename traits::char_type char_type; - basic_regex_formatter(OutputIterator o, const Results& r, const traits& t) - : m_traits(t), m_results(r), m_out(o), m_state(output_copy), m_restore_state(output_copy), m_have_conditional(false) {} - OutputIterator format(ForwardIter p1, ForwardIter p2, match_flag_type f); - OutputIterator format(ForwardIter p1, match_flag_type f) - { - return format(p1, p1 + m_traits.length(p1), f); - } -private: - typedef typename Results::value_type sub_match_type; - enum output_state - { - output_copy, - output_next_lower, - output_next_upper, - output_lower, - output_upper, - output_none - }; - - void put(char_type c); - void put(const sub_match_type& sub); - void format_all(); - void format_perl(); - void format_escape(); - void format_conditional(); - void format_until_scope_end(); - bool handle_perl_verb(bool have_brace); - - inline typename Results::value_type const& get_named_sub(ForwardIter i, ForwardIter j, const mpl::false_&) - { - std::vector v(i, j); - return (i != j) ? this->m_results.named_subexpression(&v[0], &v[0] + v.size()) - : this->m_results.named_subexpression(static_cast(0), static_cast(0)); - } - inline typename Results::value_type const& get_named_sub(ForwardIter i, ForwardIter j, const mpl::true_&) - { - return this->m_results.named_subexpression(i, j); - } - inline typename Results::value_type const& get_named_sub(ForwardIter i, ForwardIter j) - { - typedef typename cutl_details_boost::is_convertible::type tag_type; - return get_named_sub(i, j, tag_type()); - } - inline int get_named_sub_index(ForwardIter i, ForwardIter j, const mpl::false_&) - { - std::vector v(i, j); - return (i != j) ? this->m_results.named_subexpression_index(&v[0], &v[0] + v.size()) - : this->m_results.named_subexpression_index(static_cast(0), static_cast(0)); - } - inline int get_named_sub_index(ForwardIter i, ForwardIter j, const mpl::true_&) - { - return this->m_results.named_subexpression_index(i, j); - } - inline int get_named_sub_index(ForwardIter i, ForwardIter j) - { - typedef typename cutl_details_boost::is_convertible::type tag_type; - return get_named_sub_index(i, j, tag_type()); - } -#ifdef BOOST_MSVC - // msvc-8.0 issues a spurious warning on the call to std::advance here: -#pragma warning(push) -#pragma warning(disable:4244) -#endif - inline int toi(ForwardIter& i, ForwardIter j, int base, const cutl_details_boost::mpl::false_&) - { - if(i != j) - { - std::vector v(i, j); - const char_type* start = &v[0]; - const char_type* pos = start; - int r = m_traits.toi(pos, &v[0] + v.size(), base); - std::advance(i, pos - start); - return r; - } - return -1; - } -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif - inline int toi(ForwardIter& i, ForwardIter j, int base, const cutl_details_boost::mpl::true_&) - { - return m_traits.toi(i, j, base); - } - inline int toi(ForwardIter& i, ForwardIter j, int base) - { -#if defined(_MSC_VER) && defined(__INTEL_COMPILER) && ((__INTEL_COMPILER == 9999) || (__INTEL_COMPILER == 1210)) - // Workaround for Intel support issue #656654. - // See also https://svn.boost.org/trac/boost/ticket/6359 - return toi(i, j, base, mpl::false_()); -#else - typedef typename cutl_details_boost::is_convertible::type tag_type; - return toi(i, j, base, tag_type()); -#endif - } - - const traits& m_traits; // the traits class for localised formatting operations - const Results& m_results; // the match_results being used. - OutputIterator m_out; // where to send output. - ForwardIter m_position; // format string, current position - ForwardIter m_end; // format string end - match_flag_type m_flags; // format flags to use - output_state m_state; // what to do with the next character - output_state m_restore_state; // what state to restore to. - bool m_have_conditional; // we are parsing a conditional -private: - basic_regex_formatter(const basic_regex_formatter&); - basic_regex_formatter& operator=(const basic_regex_formatter&); -}; - -template -OutputIterator basic_regex_formatter::format(ForwardIter p1, ForwardIter p2, match_flag_type f) -{ - m_position = p1; - m_end = p2; - m_flags = f; - format_all(); - return m_out; -} - -template -void basic_regex_formatter::format_all() -{ - // over and over: - while(m_position != m_end) - { - switch(*m_position) - { - case '&': - if(m_flags & ::cutl_details_boost::regex_constants::format_sed) - { - ++m_position; - put(m_results[0]); - break; - } - put(*m_position++); - break; - case '\\': - format_escape(); - break; - case '(': - if(m_flags & cutl_details_boost::regex_constants::format_all) - { - ++m_position; - bool have_conditional = m_have_conditional; - m_have_conditional = false; - format_until_scope_end(); - m_have_conditional = have_conditional; - if(m_position == m_end) - return; - BOOST_ASSERT(*m_position == static_cast(')')); - ++m_position; // skip the closing ')' - break; - } - put(*m_position); - ++m_position; - break; - case ')': - if(m_flags & cutl_details_boost::regex_constants::format_all) - { - return; - } - put(*m_position); - ++m_position; - break; - case ':': - if((m_flags & cutl_details_boost::regex_constants::format_all) && m_have_conditional) - { - return; - } - put(*m_position); - ++m_position; - break; - case '?': - if(m_flags & cutl_details_boost::regex_constants::format_all) - { - ++m_position; - format_conditional(); - break; - } - put(*m_position); - ++m_position; - break; - case '$': - if((m_flags & format_sed) == 0) - { - format_perl(); - break; - } - // not a special character: - BOOST_FALLTHROUGH; - default: - put(*m_position); - ++m_position; - break; - } - } -} - -template -void basic_regex_formatter::format_perl() -{ - // - // On entry *m_position points to a '$' character - // output the information that goes with it: - // - BOOST_ASSERT(*m_position == '$'); - // - // see if this is a trailing '$': - // - if(++m_position == m_end) - { - --m_position; - put(*m_position); - ++m_position; - return; - } - // - // OK find out what kind it is: - // - bool have_brace = false; - ForwardIter save_position = m_position; - switch(*m_position) - { - case '&': - ++m_position; - put(this->m_results[0]); - break; - case '`': - ++m_position; - put(this->m_results.prefix()); - break; - case '\'': - ++m_position; - put(this->m_results.suffix()); - break; - case '$': - put(*m_position++); - break; - case '+': - if((++m_position != m_end) && (*m_position == '{')) - { - ForwardIter base = ++m_position; - while((m_position != m_end) && (*m_position != '}')) ++m_position; - if(m_position != m_end) - { - // Named sub-expression: - put(get_named_sub(base, m_position)); - ++m_position; - break; - } - else - { - m_position = --base; - } - } - put((this->m_results)[this->m_results.size() > 1 ? static_cast(this->m_results.size() - 1) : 1]); - break; - case '{': - have_brace = true; - ++m_position; - BOOST_FALLTHROUGH; - default: - // see if we have a number: - { - std::ptrdiff_t len = ::cutl_details_boost::re_detail::distance(m_position, m_end); - //len = (std::min)(static_cast(2), len); - int v = this->toi(m_position, m_position + len, 10); - if((v < 0) || (have_brace && ((m_position == m_end) || (*m_position != '}')))) - { - // Look for a Perl-5.10 verb: - if(!handle_perl_verb(have_brace)) - { - // leave the $ as is, and carry on: - m_position = --save_position; - put(*m_position); - ++m_position; - } - break; - } - // otherwise output sub v: - put(this->m_results[v]); - if(have_brace) - ++m_position; - } - } -} - -template -bool basic_regex_formatter::handle_perl_verb(bool have_brace) -{ - // - // We may have a capitalised string containing a Perl action: - // - static const char_type MATCH[] = { 'M', 'A', 'T', 'C', 'H' }; - static const char_type PREMATCH[] = { 'P', 'R', 'E', 'M', 'A', 'T', 'C', 'H' }; - static const char_type POSTMATCH[] = { 'P', 'O', 'S', 'T', 'M', 'A', 'T', 'C', 'H' }; - static const char_type LAST_PAREN_MATCH[] = { 'L', 'A', 'S', 'T', '_', 'P', 'A', 'R', 'E', 'N', '_', 'M', 'A', 'T', 'C', 'H' }; - static const char_type LAST_SUBMATCH_RESULT[] = { 'L', 'A', 'S', 'T', '_', 'S', 'U', 'B', 'M', 'A', 'T', 'C', 'H', '_', 'R', 'E', 'S', 'U', 'L', 'T' }; - static const char_type LAST_SUBMATCH_RESULT_ALT[] = { '^', 'N' }; - - if(m_position == m_end) - return false; - if(have_brace && (*m_position == '^')) - ++m_position; - - std::ptrdiff_t max_len = m_end - m_position; - - if((max_len >= 5) && std::equal(m_position, m_position + 5, MATCH)) - { - m_position += 5; - if(have_brace) - { - if((m_position != m_end) && (*m_position == '}')) - ++m_position; - else - { - m_position -= 5; - return false; - } - } - put(this->m_results[0]); - return true; - } - if((max_len >= 8) && std::equal(m_position, m_position + 8, PREMATCH)) - { - m_position += 8; - if(have_brace) - { - if((m_position != m_end) && (*m_position == '}')) - ++m_position; - else - { - m_position -= 8; - return false; - } - } - put(this->m_results.prefix()); - return true; - } - if((max_len >= 9) && std::equal(m_position, m_position + 9, POSTMATCH)) - { - m_position += 9; - if(have_brace) - { - if((m_position != m_end) && (*m_position == '}')) - ++m_position; - else - { - m_position -= 9; - return false; - } - } - put(this->m_results.suffix()); - return true; - } - if((max_len >= 16) && std::equal(m_position, m_position + 16, LAST_PAREN_MATCH)) - { - m_position += 16; - if(have_brace) - { - if((m_position != m_end) && (*m_position == '}')) - ++m_position; - else - { - m_position -= 16; - return false; - } - } - put((this->m_results)[this->m_results.size() > 1 ? static_cast(this->m_results.size() - 1) : 1]); - return true; - } - if((max_len >= 20) && std::equal(m_position, m_position + 20, LAST_SUBMATCH_RESULT)) - { - m_position += 20; - if(have_brace) - { - if((m_position != m_end) && (*m_position == '}')) - ++m_position; - else - { - m_position -= 20; - return false; - } - } - put(this->m_results.get_last_closed_paren()); - return true; - } - if((max_len >= 2) && std::equal(m_position, m_position + 2, LAST_SUBMATCH_RESULT_ALT)) - { - m_position += 2; - if(have_brace) - { - if((m_position != m_end) && (*m_position == '}')) - ++m_position; - else - { - m_position -= 2; - return false; - } - } - put(this->m_results.get_last_closed_paren()); - return true; - } - return false; -} - -template -void basic_regex_formatter::format_escape() -{ - // skip the escape and check for trailing escape: - if(++m_position == m_end) - { - put(static_cast('\\')); - return; - } - // now switch on the escape type: - switch(*m_position) - { - case 'a': - put(static_cast('\a')); - ++m_position; - break; - case 'f': - put(static_cast('\f')); - ++m_position; - break; - case 'n': - put(static_cast('\n')); - ++m_position; - break; - case 'r': - put(static_cast('\r')); - ++m_position; - break; - case 't': - put(static_cast('\t')); - ++m_position; - break; - case 'v': - put(static_cast('\v')); - ++m_position; - break; - case 'x': - if(++m_position == m_end) - { - put(static_cast('x')); - return; - } - // maybe have \x{ddd} - if(*m_position == static_cast('{')) - { - ++m_position; - int val = this->toi(m_position, m_end, 16); - if(val < 0) - { - // invalid value treat everything as literals: - put(static_cast('x')); - put(static_cast('{')); - return; - } - if((m_position == m_end) || (*m_position != static_cast('}'))) - { - --m_position; - while(*m_position != static_cast('\\')) - --m_position; - ++m_position; - put(*m_position++); - return; - } - ++m_position; - put(static_cast(val)); - return; - } - else - { - std::ptrdiff_t len = ::cutl_details_boost::re_detail::distance(m_position, m_end); - len = (std::min)(static_cast(2), len); - int val = this->toi(m_position, m_position + len, 16); - if(val < 0) - { - --m_position; - put(*m_position++); - return; - } - put(static_cast(val)); - } - break; - case 'c': - if(++m_position == m_end) - { - --m_position; - put(*m_position++); - return; - } - put(static_cast(*m_position++ % 32)); - break; - case 'e': - put(static_cast(27)); - ++m_position; - break; - default: - // see if we have a perl specific escape: - if((m_flags & cutl_details_boost::regex_constants::format_sed) == 0) - { - bool breakout = false; - switch(*m_position) - { - case 'l': - ++m_position; - m_restore_state = m_state; - m_state = output_next_lower; - breakout = true; - break; - case 'L': - ++m_position; - m_state = output_lower; - breakout = true; - break; - case 'u': - ++m_position; - m_restore_state = m_state; - m_state = output_next_upper; - breakout = true; - break; - case 'U': - ++m_position; - m_state = output_upper; - breakout = true; - break; - case 'E': - ++m_position; - m_state = output_copy; - breakout = true; - break; - } - if(breakout) - break; - } - // see if we have a \n sed style backreference: - std::ptrdiff_t len = ::cutl_details_boost::re_detail::distance(m_position, m_end); - len = (std::min)(static_cast(1), len); - int v = this->toi(m_position, m_position+len, 10); - if((v > 0) || ((v == 0) && (m_flags & ::cutl_details_boost::regex_constants::format_sed))) - { - put(m_results[v]); - break; - } - else if(v == 0) - { - // octal ecape sequence: - --m_position; - len = ::cutl_details_boost::re_detail::distance(m_position, m_end); - len = (std::min)(static_cast(4), len); - v = this->toi(m_position, m_position + len, 8); - BOOST_ASSERT(v >= 0); - put(static_cast(v)); - break; - } - // Otherwise output the character "as is": - put(*m_position++); - break; - } -} - -template -void basic_regex_formatter::format_conditional() -{ - if(m_position == m_end) - { - // oops trailing '?': - put(static_cast('?')); - return; - } - int v; - if(*m_position == '{') - { - ForwardIter base = m_position; - ++m_position; - v = this->toi(m_position, m_end, 10); - if(v < 0) - { - // Try a named subexpression: - while((m_position != m_end) && (*m_position != '}')) - ++m_position; - v = this->get_named_sub_index(base + 1, m_position); - } - if((v < 0) || (*m_position != '}')) - { - m_position = base; - // oops trailing '?': - put(static_cast('?')); - return; - } - // Skip trailing '}': - ++m_position; - } - else - { - std::ptrdiff_t len = ::cutl_details_boost::re_detail::distance(m_position, m_end); - len = (std::min)(static_cast(2), len); - v = this->toi(m_position, m_position + len, 10); - } - if(v < 0) - { - // oops not a number: - put(static_cast('?')); - return; - } - - // output varies depending upon whether sub-expression v matched or not: - if(m_results[v].matched) - { - m_have_conditional = true; - format_all(); - m_have_conditional = false; - if((m_position != m_end) && (*m_position == static_cast(':'))) - { - // skip the ':': - ++m_position; - // save output state, then turn it off: - output_state saved_state = m_state; - m_state = output_none; - // format the rest of this scope: - format_until_scope_end(); - // restore output state: - m_state = saved_state; - } - } - else - { - // save output state, then turn it off: - output_state saved_state = m_state; - m_state = output_none; - // format until ':' or ')': - m_have_conditional = true; - format_all(); - m_have_conditional = false; - // restore state: - m_state = saved_state; - if((m_position != m_end) && (*m_position == static_cast(':'))) - { - // skip the ':': - ++m_position; - // format the rest of this scope: - format_until_scope_end(); - } - } -} - -template -void basic_regex_formatter::format_until_scope_end() -{ - do - { - format_all(); - if((m_position == m_end) || (*m_position == static_cast(')'))) - return; - put(*m_position++); - }while(m_position != m_end); -} - -template -void basic_regex_formatter::put(char_type c) -{ - // write a single character to output - // according to which case translation mode we are in: - switch(this->m_state) - { - case output_none: - return; - case output_next_lower: - c = m_traits.tolower(c); - this->m_state = m_restore_state; - break; - case output_next_upper: - c = m_traits.toupper(c); - this->m_state = m_restore_state; - break; - case output_lower: - c = m_traits.tolower(c); - break; - case output_upper: - c = m_traits.toupper(c); - break; - default: - break; - } - *m_out = c; - ++m_out; -} - -template -void basic_regex_formatter::put(const sub_match_type& sub) -{ - typedef typename sub_match_type::iterator iterator_type; - iterator_type i = sub.first; - while(i != sub.second) - { - put(*i); - ++i; - } -} - -template -class string_out_iterator -#ifndef BOOST_NO_STD_ITERATOR - : public std::iterator -#endif -{ - S* out; -public: - string_out_iterator(S& s) : out(&s) {} - string_out_iterator& operator++() { return *this; } - string_out_iterator& operator++(int) { return *this; } - string_out_iterator& operator*() { return *this; } - string_out_iterator& operator=(typename S::value_type v) - { - out->append(1, v); - return *this; - } - -#ifdef BOOST_NO_STD_ITERATOR - typedef std::ptrdiff_t difference_type; - typedef typename S::value_type value_type; - typedef value_type* pointer; - typedef value_type& reference; - typedef std::output_iterator_tag iterator_category; -#endif -}; - -template -OutputIterator regex_format_imp(OutputIterator out, - const match_results& m, - ForwardIter p1, ForwardIter p2, - match_flag_type flags, - const traits& t - ) -{ - if(flags & regex_constants::format_literal) - { - return re_detail::copy(p1, p2, out); - } - - re_detail::basic_regex_formatter< - OutputIterator, - match_results, - traits, ForwardIter> f(out, m, t); - return f.format(p1, p2, flags); -} - -#ifndef BOOST_NO_SFINAE - -BOOST_MPL_HAS_XXX_TRAIT_DEF(const_iterator) - -struct any_type -{ - template - any_type(const T&); - template - any_type(const T&, const U&); - template - any_type(const T&, const U&, const V&); -}; -typedef char no_type; -typedef char (&unary_type)[2]; -typedef char (&binary_type)[3]; -typedef char (&ternary_type)[4]; - -no_type check_is_formatter(unary_type, binary_type, ternary_type); -template -unary_type check_is_formatter(T const &, binary_type, ternary_type); -template -binary_type check_is_formatter(unary_type, T const &, ternary_type); -template -binary_type check_is_formatter(T const &, U const &, ternary_type); -template -ternary_type check_is_formatter(unary_type, binary_type, T const &); -template -ternary_type check_is_formatter(T const &, binary_type, U const &); -template -ternary_type check_is_formatter(unary_type, T const &, U const &); -template -ternary_type check_is_formatter(T const &, U const &, V const &); - -struct unary_binary_ternary -{ - typedef unary_type (*unary_fun)(any_type); - typedef binary_type (*binary_fun)(any_type, any_type); - typedef ternary_type (*ternary_fun)(any_type, any_type, any_type); - operator unary_fun(); - operator binary_fun(); - operator ternary_fun(); -}; - -template::value> -struct formatter_wrapper - : Formatter - , unary_binary_ternary -{ - formatter_wrapper(){} -}; - -template -struct formatter_wrapper - : unary_binary_ternary -{ - operator Formatter *(); -}; - -template -struct formatter_wrapper - : unary_binary_ternary -{ - operator Formatter *(); -}; - -template -struct format_traits_imp -{ -private: - // - // F must be a pointer, a function, or a class with a function call operator: - // - BOOST_STATIC_ASSERT((::cutl_details_boost::is_pointer::value || ::cutl_details_boost::is_function::value || ::cutl_details_boost::is_class::value)); - static formatter_wrapper::type> f; - static M m; - static O out; - static cutl_details_boost::regex_constants::match_flag_type flags; -public: - BOOST_STATIC_CONSTANT(int, value = sizeof(check_is_formatter(f(m), f(m, out), f(m, out, flags)))); -}; - -template -struct format_traits -{ -public: - // - // Type is mpl::int_ where N is one of: - // - // 0 : F is a pointer to a presumably null-terminated string. - // 1 : F is a character-container such as a std::string. - // 2 : F is a Unary Functor. - // 3 : F is a Binary Functor. - // 4 : F is a Ternary Functor. - // - typedef typename cutl_details_boost::mpl::if_< - cutl_details_boost::mpl::and_, cutl_details_boost::mpl::not_::type> > >, - cutl_details_boost::mpl::int_<0>, - typename cutl_details_boost::mpl::if_< - has_const_iterator, - cutl_details_boost::mpl::int_<1>, - cutl_details_boost::mpl::int_::value> - >::type - >::type type; - // - // This static assertion will fail if the functor passed does not accept - // the same type of arguments passed. - // - BOOST_STATIC_ASSERT( cutl_details_boost::is_class::value && !has_const_iterator::value ? (type::value > 1) : true); -}; - -#else // BOOST_NO_SFINAE - -template -struct format_traits -{ -public: - // - // Type is mpl::int_ where N is one of: - // - // 0 : F is a pointer to a presumably null-terminated string. - // 1 : F is a character-container such as a std::string. - // - // Other options such as F being a Functor are not supported without - // SFINAE support. - // - typedef typename cutl_details_boost::mpl::if_< - cutl_details_boost::is_pointer, - cutl_details_boost::mpl::int_<0>, - cutl_details_boost::mpl::int_<1> - >::type type; -}; - -#endif // BOOST_NO_SFINAE - -template -struct format_functor3 -{ - format_functor3(Base b) : func(b) {} - template - OutputIter operator()(const Match& m, OutputIter i, cutl_details_boost::regex_constants::match_flag_type f) - { - return cutl_details_boost::unwrap_ref(func)(m, i, f); - } - template - OutputIter operator()(const Match& m, OutputIter i, cutl_details_boost::regex_constants::match_flag_type f, const Traits&) - { - return (*this)(m, i, f); - } -private: - Base func; - format_functor3(const format_functor3&); - format_functor3& operator=(const format_functor3&); -}; - -template -struct format_functor2 -{ - format_functor2(Base b) : func(b) {} - template - OutputIter operator()(const Match& m, OutputIter i, cutl_details_boost::regex_constants::match_flag_type /*f*/) - { - return cutl_details_boost::unwrap_ref(func)(m, i); - } - template - OutputIter operator()(const Match& m, OutputIter i, cutl_details_boost::regex_constants::match_flag_type f, const Traits&) - { - return (*this)(m, i, f); - } -private: - Base func; - format_functor2(const format_functor2&); - format_functor2& operator=(const format_functor2&); -}; - -template -struct format_functor1 -{ - format_functor1(Base b) : func(b) {} - - template - OutputIter do_format_string(const S& s, OutputIter i) - { - return re_detail::copy(s.begin(), s.end(), i); - } - template - inline OutputIter do_format_string(const S* s, OutputIter i) - { - while(s && *s) - { - *i = *s; - ++i; - ++s; - } - return i; - } - template - OutputIter operator()(const Match& m, OutputIter i, cutl_details_boost::regex_constants::match_flag_type /*f*/) - { - return do_format_string(cutl_details_boost::unwrap_ref(func)(m), i); - } - template - OutputIter operator()(const Match& m, OutputIter i, cutl_details_boost::regex_constants::match_flag_type f, const Traits&) - { - return (*this)(m, i, f); - } -private: - Base func; - format_functor1(const format_functor1&); - format_functor1& operator=(const format_functor1&); -}; - -template -struct format_functor_c_string -{ - format_functor_c_string(const charT* ps) : func(ps) {} - - template - OutputIter operator()(const Match& m, OutputIter i, cutl_details_boost::regex_constants::match_flag_type f, const Traits& t = Traits()) - { - //typedef typename Match::char_type char_type; - const charT* end = func; - while(*end) ++end; - return regex_format_imp(i, m, func, end, f, t); - } -private: - const charT* func; - format_functor_c_string(const format_functor_c_string&); - format_functor_c_string& operator=(const format_functor_c_string&); -}; - -template -struct format_functor_container -{ - format_functor_container(const Container& c) : func(c) {} - - template - OutputIter operator()(const Match& m, OutputIter i, cutl_details_boost::regex_constants::match_flag_type f, const Traits& t = Traits()) - { - //typedef typename Match::char_type char_type; - return re_detail::regex_format_imp(i, m, func.begin(), func.end(), f, t); - } -private: - const Container& func; - format_functor_container(const format_functor_container&); - format_functor_container& operator=(const format_functor_container&); -}; - -template > -struct compute_functor_type -{ - typedef typename format_traits::type tag; - typedef typename cutl_details_boost::remove_cv< typename cutl_details_boost::remove_pointer::type>::type maybe_char_type; - - typedef typename mpl::if_< - ::cutl_details_boost::is_same >, format_functor_c_string, - typename mpl::if_< - ::cutl_details_boost::is_same >, format_functor_container, - typename mpl::if_< - ::cutl_details_boost::is_same >, format_functor1, - typename mpl::if_< - ::cutl_details_boost::is_same >, format_functor2, - format_functor3 - >::type - >::type - >::type - >::type type; -}; - -} // namespace re_detail - -template -inline OutputIterator regex_format(OutputIterator out, - const match_results& m, - Functor fmt, - match_flag_type flags = format_all - ) -{ - return m.format(out, fmt, flags); -} - -template -inline std::basic_string::char_type> regex_format(const match_results& m, - Functor fmt, - match_flag_type flags = format_all) -{ - return m.format(fmt, flags); -} - -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable: 4103) -#endif -#ifdef BOOST_HAS_ABI_HEADERS -# include BOOST_ABI_SUFFIX -#endif -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif - -} // namespace cutl_details_boost - -#endif // BOOST_REGEX_FORMAT_HPP - - - - - - diff --git a/cutl/details/boost/regex/v4/regex_fwd.hpp b/cutl/details/boost/regex/v4/regex_fwd.hpp deleted file mode 100644 index da2d66b..0000000 --- a/cutl/details/boost/regex/v4/regex_fwd.hpp +++ /dev/null @@ -1,73 +0,0 @@ -/* - * - * Copyright (c) 1998-2002 - * John Maddock - * - * Use, modification and distribution are subject to the - * Boost Software License, Version 1.0. (See accompanying file - * LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt) - * - */ - - /* - * LOCATION: see http://www.boost.org for most recent version. - * FILE regex_fwd.cpp - * VERSION see - * DESCRIPTION: Forward declares cutl_details_boost::basic_regex<> and - * associated typedefs. - */ - -#ifndef BOOST_REGEX_FWD_HPP_INCLUDED -#define BOOST_REGEX_FWD_HPP_INCLUDED - -#ifndef BOOST_REGEX_CONFIG_HPP -#include -#endif - -// -// define BOOST_REGEX_NO_FWD if this -// header doesn't work! -// -#ifdef BOOST_REGEX_NO_FWD -# ifndef BOOST_RE_REGEX_HPP -# include -# endif -#else - -namespace cutl_details_boost{ - -template -class cpp_regex_traits; -template -struct c_regex_traits; -template -class w32_regex_traits; - -#ifdef BOOST_REGEX_USE_WIN32_LOCALE -template > -struct regex_traits; -#elif defined(BOOST_REGEX_USE_CPP_LOCALE) -template > -struct regex_traits; -#else -template > -struct regex_traits; -#endif - -template > -class basic_regex; - -typedef basic_regex > regex; -#ifndef BOOST_NO_WREGEX -typedef basic_regex > wregex; -#endif - -} // namespace cutl_details_boost - -#endif // BOOST_REGEX_NO_FWD - -#endif - - - - diff --git a/cutl/details/boost/regex/v4/regex_grep.hpp b/cutl/details/boost/regex/v4/regex_grep.hpp deleted file mode 100644 index f3f3eec..0000000 --- a/cutl/details/boost/regex/v4/regex_grep.hpp +++ /dev/null @@ -1,155 +0,0 @@ -/* - * - * Copyright (c) 1998-2002 - * John Maddock - * - * Use, modification and distribution are subject to the - * Boost Software License, Version 1.0. (See accompanying file - * LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt) - * - */ - - /* - * LOCATION: see http://www.boost.org for most recent version. - * FILE regex_grep.hpp - * VERSION see - * DESCRIPTION: Provides regex_grep implementation. - */ - -#ifndef BOOST_REGEX_V4_REGEX_GREP_HPP -#define BOOST_REGEX_V4_REGEX_GREP_HPP - - -namespace cutl_details_boost{ - -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable: 4103) -#endif -#ifdef BOOST_HAS_ABI_HEADERS -# include BOOST_ABI_PREFIX -#endif -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif - -// -// regex_grep: -// find all non-overlapping matches within the sequence first last: -// -template -inline unsigned int regex_grep(Predicate foo, - BidiIterator first, - BidiIterator last, - const basic_regex& e, - match_flag_type flags = match_default) -{ - if(e.flags() & regex_constants::failbit) - return false; - - typedef typename match_results::allocator_type match_allocator_type; - - match_results m; - re_detail::perl_matcher matcher(first, last, m, e, flags, first); - unsigned int count = 0; - while(matcher.find()) - { - ++count; - if(0 == foo(m)) - return count; // caller doesn't want to go on - if(m[0].second == last) - return count; // we've reached the end, don't try and find an extra null match. - if(m.length() == 0) - { - if(m[0].second == last) - return count; - // we found a NULL-match, now try to find - // a non-NULL one at the same position: - match_results m2(m); - matcher.setf(match_not_null | match_continuous); - if(matcher.find()) - { - ++count; - if(0 == foo(m)) - return count; - } - else - { - // reset match back to where it was: - m = m2; - } - matcher.unsetf((match_not_null | match_continuous) & ~flags); - } - } - return count; -} - -// -// regex_grep convenience interfaces: -#ifndef BOOST_NO_FUNCTION_TEMPLATE_ORDERING -// -// this isn't really a partial specialisation, but template function -// overloading - if the compiler doesn't support partial specialisation -// then it really won't support this either: -template -inline unsigned int regex_grep(Predicate foo, const charT* str, - const basic_regex& e, - match_flag_type flags = match_default) -{ - return regex_grep(foo, str, str + traits::length(str), e, flags); -} - -template -inline unsigned int regex_grep(Predicate foo, const std::basic_string& s, - const basic_regex& e, - match_flag_type flags = match_default) -{ - return regex_grep(foo, s.begin(), s.end(), e, flags); -} -#else // partial specialisation -inline unsigned int regex_grep(bool (*foo)(const cmatch&), const char* str, - const regex& e, - match_flag_type flags = match_default) -{ - return regex_grep(foo, str, str + regex::traits_type::length(str), e, flags); -} -#ifndef BOOST_NO_WREGEX -inline unsigned int regex_grep(bool (*foo)(const wcmatch&), const wchar_t* str, - const wregex& e, - match_flag_type flags = match_default) -{ - return regex_grep(foo, str, str + wregex::traits_type::length(str), e, flags); -} -#endif -inline unsigned int regex_grep(bool (*foo)(const match_results&), const std::string& s, - const regex& e, - match_flag_type flags = match_default) -{ - return regex_grep(foo, s.begin(), s.end(), e, flags); -} -#if !defined(BOOST_NO_WREGEX) -inline unsigned int regex_grep(bool (*foo)(const match_results::const_iterator>&), - const std::basic_string& s, - const wregex& e, - match_flag_type flags = match_default) -{ - return regex_grep(foo, s.begin(), s.end(), e, flags); -} -#endif -#endif - -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable: 4103) -#endif -#ifdef BOOST_HAS_ABI_HEADERS -# include BOOST_ABI_SUFFIX -#endif -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif - -} // namespace cutl_details_boost - -#endif // BOOST_REGEX_V4_REGEX_GREP_HPP - diff --git a/cutl/details/boost/regex/v4/regex_iterator.hpp b/cutl/details/boost/regex/v4/regex_iterator.hpp deleted file mode 100644 index 82d0d09..0000000 --- a/cutl/details/boost/regex/v4/regex_iterator.hpp +++ /dev/null @@ -1,201 +0,0 @@ -/* - * - * Copyright (c) 2003 - * John Maddock - * - * Use, modification and distribution are subject to the - * Boost Software License, Version 1.0. (See accompanying file - * LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt) - * - */ - - /* - * LOCATION: see http://www.boost.org for most recent version. - * FILE regex_iterator.hpp - * VERSION see - * DESCRIPTION: Provides regex_iterator implementation. - */ - -#ifndef BOOST_REGEX_V4_REGEX_ITERATOR_HPP -#define BOOST_REGEX_V4_REGEX_ITERATOR_HPP - -#include - -namespace cutl_details_boost{ - -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable: 4103) -#endif -#ifdef BOOST_HAS_ABI_HEADERS -# include BOOST_ABI_PREFIX -#endif -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif - -template -class regex_iterator_implementation -{ - typedef basic_regex regex_type; - - match_results what; // current match - BidirectionalIterator base; // start of sequence - BidirectionalIterator end; // end of sequence - const regex_type re; // the expression - match_flag_type flags; // flags for matching - -public: - regex_iterator_implementation(const regex_type* p, BidirectionalIterator last, match_flag_type f) - : base(), end(last), re(*p), flags(f){} - bool init(BidirectionalIterator first) - { - base = first; - return regex_search(first, end, what, re, flags); - } - bool compare(const regex_iterator_implementation& that) - { - if(this == &that) return true; - return (&re.get_data() == &that.re.get_data()) && (end == that.end) && (flags == that.flags) && (what[0].first == that.what[0].first) && (what[0].second == that.what[0].second); - } - const match_results& get() - { return what; } - bool next() - { - //if(what.prefix().first != what[0].second) - // flags |= match_prev_avail; - BidirectionalIterator next_start = what[0].second; - match_flag_type f(flags); - if(!what.length() || (f & regex_constants::match_posix)) - f |= regex_constants::match_not_initial_null; - //if(base != next_start) - // f |= regex_constants::match_not_bob; - bool result = regex_search(next_start, end, what, re, f, base); - if(result) - what.set_base(base); - return result; - } -private: - regex_iterator_implementation& operator=(const regex_iterator_implementation&); -}; - -template ::value_type, - class traits = regex_traits > -class regex_iterator -#ifndef BOOST_NO_STD_ITERATOR - : public std::iterator< - std::forward_iterator_tag, - match_results, - typename re_detail::regex_iterator_traits::difference_type, - const match_results*, - const match_results& > -#endif -{ -private: - typedef regex_iterator_implementation impl; - typedef shared_ptr pimpl; -public: - typedef basic_regex regex_type; - typedef match_results value_type; - typedef typename re_detail::regex_iterator_traits::difference_type - difference_type; - typedef const value_type* pointer; - typedef const value_type& reference; - typedef std::forward_iterator_tag iterator_category; - - regex_iterator(){} - regex_iterator(BidirectionalIterator a, BidirectionalIterator b, - const regex_type& re, - match_flag_type m = match_default) - : pdata(new impl(&re, b, m)) - { - if(!pdata->init(a)) - { - pdata.reset(); - } - } - regex_iterator(const regex_iterator& that) - : pdata(that.pdata) {} - regex_iterator& operator=(const regex_iterator& that) - { - pdata = that.pdata; - return *this; - } - bool operator==(const regex_iterator& that)const - { - if((pdata.get() == 0) || (that.pdata.get() == 0)) - return pdata.get() == that.pdata.get(); - return pdata->compare(*(that.pdata.get())); - } - bool operator!=(const regex_iterator& that)const - { return !(*this == that); } - const value_type& operator*()const - { return pdata->get(); } - const value_type* operator->()const - { return &(pdata->get()); } - regex_iterator& operator++() - { - cow(); - if(0 == pdata->next()) - { - pdata.reset(); - } - return *this; - } - regex_iterator operator++(int) - { - regex_iterator result(*this); - ++(*this); - return result; - } -private: - - pimpl pdata; - - void cow() - { - // copy-on-write - if(pdata.get() && !pdata.unique()) - { - pdata.reset(new impl(*(pdata.get()))); - } - } -}; - -typedef regex_iterator cregex_iterator; -typedef regex_iterator sregex_iterator; -#ifndef BOOST_NO_WREGEX -typedef regex_iterator wcregex_iterator; -typedef regex_iterator wsregex_iterator; -#endif - -// make_regex_iterator: -template -inline regex_iterator make_regex_iterator(const charT* p, const basic_regex& e, regex_constants::match_flag_type m = regex_constants::match_default) -{ - return regex_iterator(p, p+traits::length(p), e, m); -} -template -inline regex_iterator::const_iterator, charT, traits> make_regex_iterator(const std::basic_string& p, const basic_regex& e, regex_constants::match_flag_type m = regex_constants::match_default) -{ - return regex_iterator::const_iterator, charT, traits>(p.begin(), p.end(), e, m); -} - -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable: 4103) -#endif -#ifdef BOOST_HAS_ABI_HEADERS -# include BOOST_ABI_SUFFIX -#endif -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif - -} // namespace cutl_details_boost - -#endif // BOOST_REGEX_V4_REGEX_ITERATOR_HPP - diff --git a/cutl/details/boost/regex/v4/regex_match.hpp b/cutl/details/boost/regex/v4/regex_match.hpp deleted file mode 100644 index 045c4bc..0000000 --- a/cutl/details/boost/regex/v4/regex_match.hpp +++ /dev/null @@ -1,382 +0,0 @@ -/* - * - * Copyright (c) 1998-2002 - * John Maddock - * - * Use, modification and distribution are subject to the - * Boost Software License, Version 1.0. (See accompanying file - * LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt) - * - */ - - /* - * LOCATION: see http://www.boost.org for most recent version. - * FILE regex_match.hpp - * VERSION see - * DESCRIPTION: Regular expression matching algorithms. - * Note this is an internal header file included - * by regex.hpp, do not include on its own. - */ - - -#ifndef BOOST_REGEX_MATCH_HPP -#define BOOST_REGEX_MATCH_HPP - -namespace cutl_details_boost{ - -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable: 4103) -#endif -#ifdef BOOST_HAS_ABI_HEADERS -# include BOOST_ABI_PREFIX -#endif -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif - -// -// proc regex_match -// returns true if the specified regular expression matches -// the whole of the input. Fills in what matched in m. -// -template -bool regex_match(BidiIterator first, BidiIterator last, - match_results& m, - const basic_regex& e, - match_flag_type flags = match_default) -{ - re_detail::perl_matcher matcher(first, last, m, e, flags, first); - return matcher.match(); -} -template -bool regex_match(iterator first, iterator last, - const basic_regex& e, - match_flag_type flags = match_default) -{ - match_results m; - return regex_match(first, last, m, e, flags | regex_constants::match_any); -} -// -// query_match convenience interfaces: -#ifndef BOOST_NO_FUNCTION_TEMPLATE_ORDERING -// -// this isn't really a partial specialisation, but template function -// overloading - if the compiler doesn't support partial specialisation -// then it really won't support this either: -template -inline bool regex_match(const charT* str, - match_results& m, - const basic_regex& e, - match_flag_type flags = match_default) -{ - return regex_match(str, str + traits::length(str), m, e, flags); -} - -template -inline bool regex_match(const std::basic_string& s, - match_results::const_iterator, Allocator>& m, - const basic_regex& e, - match_flag_type flags = match_default) -{ - return regex_match(s.begin(), s.end(), m, e, flags); -} -template -inline bool regex_match(const charT* str, - const basic_regex& e, - match_flag_type flags = match_default) -{ - match_results m; - return regex_match(str, str + traits::length(str), m, e, flags | regex_constants::match_any); -} - -template -inline bool regex_match(const std::basic_string& s, - const basic_regex& e, - match_flag_type flags = match_default) -{ - typedef typename std::basic_string::const_iterator iterator; - match_results m; - return regex_match(s.begin(), s.end(), m, e, flags | regex_constants::match_any); -} -#else // partial ordering -inline bool regex_match(const char* str, - cmatch& m, - const regex& e, - match_flag_type flags = match_default) -{ - return regex_match(str, str + regex::traits_type::length(str), m, e, flags); -} -inline bool regex_match(const char* str, - const regex& e, - match_flag_type flags = match_default) -{ - match_results m; - return regex_match(str, str + regex::traits_type::length(str), m, e, flags | regex_constants::match_any); -} -#ifndef BOOST_NO_STD_LOCALE -inline bool regex_match(const char* str, - cmatch& m, - const basic_regex >& e, - match_flag_type flags = match_default) -{ - return regex_match(str, str + regex::traits_type::length(str), m, e, flags); -} -inline bool regex_match(const char* str, - const basic_regex >& e, - match_flag_type flags = match_default) -{ - match_results m; - return regex_match(str, str + regex::traits_type::length(str), m, e, flags | regex_constants::match_any); -} -#endif -inline bool regex_match(const char* str, - cmatch& m, - const basic_regex >& e, - match_flag_type flags = match_default) -{ - return regex_match(str, str + regex::traits_type::length(str), m, e, flags); -} -inline bool regex_match(const char* str, - const basic_regex >& e, - match_flag_type flags = match_default) -{ - match_results m; - return regex_match(str, str + regex::traits_type::length(str), m, e, flags | regex_constants::match_any); -} -#if defined(_WIN32) && !defined(BOOST_REGEX_NO_W32) -inline bool regex_match(const char* str, - cmatch& m, - const basic_regex >& e, - match_flag_type flags = match_default) -{ - return regex_match(str, str + regex::traits_type::length(str), m, e, flags); -} -inline bool regex_match(const char* str, - const basic_regex >& e, - match_flag_type flags = match_default) -{ - match_results m; - return regex_match(str, str + regex::traits_type::length(str), m, e, flags | regex_constants::match_any); -} -#endif -#ifndef BOOST_NO_WREGEX -inline bool regex_match(const wchar_t* str, - wcmatch& m, - const wregex& e, - match_flag_type flags = match_default) -{ - return regex_match(str, str + wregex::traits_type::length(str), m, e, flags); -} -inline bool regex_match(const wchar_t* str, - const wregex& e, - match_flag_type flags = match_default) -{ - match_results m; - return regex_match(str, str + wregex::traits_type::length(str), m, e, flags | regex_constants::match_any); -} -#ifndef BOOST_NO_STD_LOCALE -inline bool regex_match(const wchar_t* str, - wcmatch& m, - const basic_regex >& e, - match_flag_type flags = match_default) -{ - return regex_match(str, str + wregex::traits_type::length(str), m, e, flags); -} -inline bool regex_match(const wchar_t* str, - const basic_regex >& e, - match_flag_type flags = match_default) -{ - match_results m; - return regex_match(str, str + wregex::traits_type::length(str), m, e, flags | regex_constants::match_any); -} -#endif -inline bool regex_match(const wchar_t* str, - wcmatch& m, - const basic_regex >& e, - match_flag_type flags = match_default) -{ - return regex_match(str, str + wregex::traits_type::length(str), m, e, flags); -} -inline bool regex_match(const wchar_t* str, - const basic_regex >& e, - match_flag_type flags = match_default) -{ - match_results m; - return regex_match(str, str + wregex::traits_type::length(str), m, e, flags | regex_constants::match_any); -} -#if defined(_WIN32) && !defined(BOOST_REGEX_NO_W32) -inline bool regex_match(const wchar_t* str, - wcmatch& m, - const basic_regex >& e, - match_flag_type flags = match_default) -{ - return regex_match(str, str + wregex::traits_type::length(str), m, e, flags); -} -inline bool regex_match(const wchar_t* str, - const basic_regex >& e, - match_flag_type flags = match_default) -{ - match_results m; - return regex_match(str, str + wregex::traits_type::length(str), m, e, flags | regex_constants::match_any); -} -#endif -#endif -inline bool regex_match(const std::string& s, - smatch& m, - const regex& e, - match_flag_type flags = match_default) -{ - return regex_match(s.begin(), s.end(), m, e, flags); -} -inline bool regex_match(const std::string& s, - const regex& e, - match_flag_type flags = match_default) -{ - match_results m; - return regex_match(s.begin(), s.end(), m, e, flags | regex_constants::match_any); -} -#ifndef BOOST_NO_STD_LOCALE -inline bool regex_match(const std::string& s, - smatch& m, - const basic_regex >& e, - match_flag_type flags = match_default) -{ - return regex_match(s.begin(), s.end(), m, e, flags); -} -inline bool regex_match(const std::string& s, - const basic_regex >& e, - match_flag_type flags = match_default) -{ - match_results m; - return regex_match(s.begin(), s.end(), m, e, flags | regex_constants::match_any); -} -#endif -inline bool regex_match(const std::string& s, - smatch& m, - const basic_regex >& e, - match_flag_type flags = match_default) -{ - return regex_match(s.begin(), s.end(), m, e, flags); -} -inline bool regex_match(const std::string& s, - const basic_regex >& e, - match_flag_type flags = match_default) -{ - match_results m; - return regex_match(s.begin(), s.end(), m, e, flags | regex_constants::match_any); -} -#if defined(_WIN32) && !defined(BOOST_REGEX_NO_W32) -inline bool regex_match(const std::string& s, - smatch& m, - const basic_regex >& e, - match_flag_type flags = match_default) -{ - return regex_match(s.begin(), s.end(), m, e, flags); -} -inline bool regex_match(const std::string& s, - const basic_regex >& e, - match_flag_type flags = match_default) -{ - match_results m; - return regex_match(s.begin(), s.end(), m, e, flags | regex_constants::match_any); -} -#endif -#if !defined(BOOST_NO_WREGEX) -inline bool regex_match(const std::basic_string& s, - match_results::const_iterator>& m, - const wregex& e, - match_flag_type flags = match_default) -{ - return regex_match(s.begin(), s.end(), m, e, flags); -} -inline bool regex_match(const std::basic_string& s, - const wregex& e, - match_flag_type flags = match_default) -{ - match_results::const_iterator> m; - return regex_match(s.begin(), s.end(), m, e, flags | regex_constants::match_any); -} -#ifndef BOOST_NO_STD_LOCALE -inline bool regex_match(const std::basic_string& s, - match_results::const_iterator>& m, - const basic_regex >& e, - match_flag_type flags = match_default) -{ - return regex_match(s.begin(), s.end(), m, e, flags); -} -inline bool regex_match(const std::basic_string& s, - const basic_regex >& e, - match_flag_type flags = match_default) -{ - match_results::const_iterator> m; - return regex_match(s.begin(), s.end(), m, e, flags | regex_constants::match_any); -} -#endif -inline bool regex_match(const std::basic_string& s, - match_results::const_iterator>& m, - const basic_regex >& e, - match_flag_type flags = match_default) -{ - return regex_match(s.begin(), s.end(), m, e, flags); -} -inline bool regex_match(const std::basic_string& s, - const basic_regex >& e, - match_flag_type flags = match_default) -{ - match_results::const_iterator> m; - return regex_match(s.begin(), s.end(), m, e, flags | regex_constants::match_any); -} -#if defined(_WIN32) && !defined(BOOST_REGEX_NO_W32) -inline bool regex_match(const std::basic_string& s, - match_results::const_iterator>& m, - const basic_regex >& e, - match_flag_type flags = match_default) -{ - return regex_match(s.begin(), s.end(), m, e, flags); -} -inline bool regex_match(const std::basic_string& s, - const basic_regex >& e, - match_flag_type flags = match_default) -{ - match_results::const_iterator> m; - return regex_match(s.begin(), s.end(), m, e, flags | regex_constants::match_any); -} -#endif -#endif - -#endif - - -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable: 4103) -#endif -#ifdef BOOST_HAS_ABI_HEADERS -# include BOOST_ABI_SUFFIX -#endif -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif - -} // namespace cutl_details_boost - -#endif // BOOST_REGEX_MATCH_HPP - - - - - - - - - - - - - - - - - - diff --git a/cutl/details/boost/regex/v4/regex_merge.hpp b/cutl/details/boost/regex/v4/regex_merge.hpp deleted file mode 100644 index 80f00ac..0000000 --- a/cutl/details/boost/regex/v4/regex_merge.hpp +++ /dev/null @@ -1,93 +0,0 @@ -/* - * - * Copyright (c) 1998-2002 - * John Maddock - * - * Use, modification and distribution are subject to the - * Boost Software License, Version 1.0. (See accompanying file - * LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt) - * - */ - - /* - * LOCATION: see http://www.boost.org for most recent version. - * FILE regex_format.hpp - * VERSION see - * DESCRIPTION: Provides formatting output routines for search and replace - * operations. Note this is an internal header file included - * by regex.hpp, do not include on its own. - */ - -#ifndef BOOST_REGEX_V4_REGEX_MERGE_HPP -#define BOOST_REGEX_V4_REGEX_MERGE_HPP - - -namespace cutl_details_boost{ - -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable: 4103) -#endif -#ifdef BOOST_HAS_ABI_HEADERS -# include BOOST_ABI_PREFIX -#endif -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif - -template -inline OutputIterator regex_merge(OutputIterator out, - Iterator first, - Iterator last, - const basic_regex& e, - const charT* fmt, - match_flag_type flags = match_default) -{ - return regex_replace(out, first, last, e, fmt, flags); -} - -template -inline OutputIterator regex_merge(OutputIterator out, - Iterator first, - Iterator last, - const basic_regex& e, - const std::basic_string& fmt, - match_flag_type flags = match_default) -{ - return regex_merge(out, first, last, e, fmt.c_str(), flags); -} - -template -inline std::basic_string regex_merge(const std::basic_string& s, - const basic_regex& e, - const charT* fmt, - match_flag_type flags = match_default) -{ - return regex_replace(s, e, fmt, flags); -} - -template -inline std::basic_string regex_merge(const std::basic_string& s, - const basic_regex& e, - const std::basic_string& fmt, - match_flag_type flags = match_default) -{ - return regex_replace(s, e, fmt, flags); -} - -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable: 4103) -#endif -#ifdef BOOST_HAS_ABI_HEADERS -# include BOOST_ABI_SUFFIX -#endif -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif - -} // namespace cutl_details_boost - -#endif // BOOST_REGEX_V4_REGEX_MERGE_HPP - - diff --git a/cutl/details/boost/regex/v4/regex_raw_buffer.hpp b/cutl/details/boost/regex/v4/regex_raw_buffer.hpp deleted file mode 100644 index 5aba83d..0000000 --- a/cutl/details/boost/regex/v4/regex_raw_buffer.hpp +++ /dev/null @@ -1,210 +0,0 @@ -/* - * - * Copyright (c) 1998-2002 - * John Maddock - * - * Use, modification and distribution are subject to the - * Boost Software License, Version 1.0. (See accompanying file - * LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt) - * - */ - - /* - * LOCATION: see http://www.boost.org for most recent version. - * FILE regex_raw_buffer.hpp - * VERSION see - * DESCRIPTION: Raw character buffer for regex code. - * Note this is an internal header file included - * by regex.hpp, do not include on its own. - */ - -#ifndef BOOST_REGEX_RAW_BUFFER_HPP -#define BOOST_REGEX_RAW_BUFFER_HPP - -#ifndef BOOST_REGEX_CONFIG_HPP -#include -#endif - -#include -#include - -namespace cutl_details_boost{ - namespace re_detail{ - -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable: 4103) -#endif -#ifdef BOOST_HAS_ABI_HEADERS -# include BOOST_ABI_PREFIX -#endif -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif - -struct empty_padding{}; - -union padding -{ - void* p; - unsigned int i; -}; - -template -struct padding3 -{ - enum{ - padding_size = 8, - padding_mask = 7 - }; -}; - -template<> -struct padding3<2> -{ - enum{ - padding_size = 2, - padding_mask = 1 - }; -}; - -template<> -struct padding3<4> -{ - enum{ - padding_size = 4, - padding_mask = 3 - }; -}; - -template<> -struct padding3<8> -{ - enum{ - padding_size = 8, - padding_mask = 7 - }; -}; - -template<> -struct padding3<16> -{ - enum{ - padding_size = 16, - padding_mask = 15 - }; -}; - -enum{ - padding_size = padding3::padding_size, - padding_mask = padding3::padding_mask -}; - -// -// class raw_storage -// basically this is a simplified vector -// this is used by basic_regex for expression storage -// - -class BOOST_REGEX_DECL raw_storage -{ -public: - typedef std::size_t size_type; - typedef unsigned char* pointer; -private: - pointer last, start, end; -public: - - raw_storage(); - raw_storage(size_type n); - - ~raw_storage() - { - ::operator delete(start); - } - - void BOOST_REGEX_CALL resize(size_type n); - - void* BOOST_REGEX_CALL extend(size_type n) - { - if(size_type(last - end) < n) - resize(n + (end - start)); - register pointer result = end; - end += n; - return result; - } - - void* BOOST_REGEX_CALL insert(size_type pos, size_type n); - - size_type BOOST_REGEX_CALL size() - { - return end - start; - } - - size_type BOOST_REGEX_CALL capacity() - { - return last - start; - } - - void* BOOST_REGEX_CALL data()const - { - return start; - } - - size_type BOOST_REGEX_CALL index(void* ptr) - { - return static_cast(ptr) - static_cast(data()); - } - - void BOOST_REGEX_CALL clear() - { - end = start; - } - - void BOOST_REGEX_CALL align() - { - // move end up to a boundary: - end = start + (((end - start) + padding_mask) & ~padding_mask); - } - void swap(raw_storage& that) - { - std::swap(start, that.start); - std::swap(end, that.end); - std::swap(last, that.last); - } -}; - -inline raw_storage::raw_storage() -{ - last = start = end = 0; -} - -inline raw_storage::raw_storage(size_type n) -{ - start = end = static_cast(::operator new(n)); - BOOST_REGEX_NOEH_ASSERT(start) - last = start + n; -} - - -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable: 4103) -#endif -#ifdef BOOST_HAS_ABI_HEADERS -# include BOOST_ABI_SUFFIX -#endif -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif - -} // namespace re_detail -} // namespace cutl_details_boost - -#endif - - - - - - diff --git a/cutl/details/boost/regex/v4/regex_replace.hpp b/cutl/details/boost/regex/v4/regex_replace.hpp deleted file mode 100644 index 24867da..0000000 --- a/cutl/details/boost/regex/v4/regex_replace.hpp +++ /dev/null @@ -1,99 +0,0 @@ -/* - * - * Copyright (c) 1998-2009 - * John Maddock - * - * Use, modification and distribution are subject to the - * Boost Software License, Version 1.0. (See accompanying file - * LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt) - * - */ - - /* - * LOCATION: see http://www.boost.org for most recent version. - * FILE regex_format.hpp - * VERSION see - * DESCRIPTION: Provides formatting output routines for search and replace - * operations. Note this is an internal header file included - * by regex.hpp, do not include on its own. - */ - -#ifndef BOOST_REGEX_V4_REGEX_REPLACE_HPP -#define BOOST_REGEX_V4_REGEX_REPLACE_HPP - - -namespace cutl_details_boost{ - -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable: 4103) -#endif -#ifdef BOOST_HAS_ABI_HEADERS -# include BOOST_ABI_PREFIX -#endif -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif - -template -OutputIterator regex_replace(OutputIterator out, - BidirectionalIterator first, - BidirectionalIterator last, - const basic_regex& e, - Formatter fmt, - match_flag_type flags = match_default) -{ - regex_iterator i(first, last, e, flags); - regex_iterator j; - if(i == j) - { - if(!(flags & regex_constants::format_no_copy)) - out = re_detail::copy(first, last, out); - } - else - { - BidirectionalIterator last_m(first); - while(i != j) - { - if(!(flags & regex_constants::format_no_copy)) - out = re_detail::copy(i->prefix().first, i->prefix().second, out); - out = i->format(out, fmt, flags, e); - last_m = (*i)[0].second; - if(flags & regex_constants::format_first_only) - break; - ++i; - } - if(!(flags & regex_constants::format_no_copy)) - out = re_detail::copy(last_m, last, out); - } - return out; -} - -template -std::basic_string regex_replace(const std::basic_string& s, - const basic_regex& e, - Formatter fmt, - match_flag_type flags = match_default) -{ - std::basic_string result; - re_detail::string_out_iterator > i(result); - regex_replace(i, s.begin(), s.end(), e, fmt, flags); - return result; -} - -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable: 4103) -#endif -#ifdef BOOST_HAS_ABI_HEADERS -# include BOOST_ABI_SUFFIX -#endif -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif - -} // namespace cutl_details_boost - -#endif // BOOST_REGEX_V4_REGEX_REPLACE_HPP - - diff --git a/cutl/details/boost/regex/v4/regex_search.hpp b/cutl/details/boost/regex/v4/regex_search.hpp deleted file mode 100644 index e6de6b6..0000000 --- a/cutl/details/boost/regex/v4/regex_search.hpp +++ /dev/null @@ -1,217 +0,0 @@ -/* - * - * Copyright (c) 1998-2002 - * John Maddock - * - * Use, modification and distribution are subject to the - * Boost Software License, Version 1.0. (See accompanying file - * LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt) - * - */ - - /* - * LOCATION: see http://www.boost.org for most recent version. - * FILE regex_search.hpp - * VERSION see - * DESCRIPTION: Provides regex_search implementation. - */ - -#ifndef BOOST_REGEX_V4_REGEX_SEARCH_HPP -#define BOOST_REGEX_V4_REGEX_SEARCH_HPP - - -namespace cutl_details_boost{ - -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable: 4103) -#endif -#ifdef BOOST_HAS_ABI_HEADERS -# include BOOST_ABI_PREFIX -#endif -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif - -template -bool regex_search(BidiIterator first, BidiIterator last, - match_results& m, - const basic_regex& e, - match_flag_type flags = match_default) -{ - return regex_search(first, last, m, e, flags, first); -} - -template -bool regex_search(BidiIterator first, BidiIterator last, - match_results& m, - const basic_regex& e, - match_flag_type flags, - BidiIterator base) -{ - if(e.flags() & regex_constants::failbit) - return false; - - re_detail::perl_matcher matcher(first, last, m, e, flags, base); - return matcher.find(); -} - -// -// regex_search convenience interfaces: -#ifndef BOOST_NO_FUNCTION_TEMPLATE_ORDERING -// -// this isn't really a partial specialisation, but template function -// overloading - if the compiler doesn't support partial specialisation -// then it really won't support this either: -template -inline bool regex_search(const charT* str, - match_results& m, - const basic_regex& e, - match_flag_type flags = match_default) -{ - return regex_search(str, str + traits::length(str), m, e, flags); -} - -template -inline bool regex_search(const std::basic_string& s, - match_results::const_iterator, Allocator>& m, - const basic_regex& e, - match_flag_type flags = match_default) -{ - return regex_search(s.begin(), s.end(), m, e, flags); -} -#else // partial overloads: -inline bool regex_search(const char* str, - cmatch& m, - const regex& e, - match_flag_type flags = match_default) -{ - return regex_search(str, str + regex::traits_type::length(str), m, e, flags); -} -inline bool regex_search(const char* first, const char* last, - const regex& e, - match_flag_type flags = match_default) -{ - cmatch m; - return regex_search(first, last, m, e, flags | regex_constants::match_any); -} - -#ifndef BOOST_NO_WREGEX -inline bool regex_search(const wchar_t* str, - wcmatch& m, - const wregex& e, - match_flag_type flags = match_default) -{ - return regex_search(str, str + wregex::traits_type::length(str), m, e, flags); -} -inline bool regex_search(const wchar_t* first, const wchar_t* last, - const wregex& e, - match_flag_type flags = match_default) -{ - wcmatch m; - return regex_search(first, last, m, e, flags | regex_constants::match_any); -} -#endif -inline bool regex_search(const std::string& s, - smatch& m, - const regex& e, - match_flag_type flags = match_default) -{ - return regex_search(s.begin(), s.end(), m, e, flags); -} -#if !defined(BOOST_NO_WREGEX) -inline bool regex_search(const std::basic_string& s, - wsmatch& m, - const wregex& e, - match_flag_type flags = match_default) -{ - return regex_search(s.begin(), s.end(), m, e, flags); -} -#endif - -#endif - -template -bool regex_search(BidiIterator first, BidiIterator last, - const basic_regex& e, - match_flag_type flags = match_default) -{ - if(e.flags() & regex_constants::failbit) - return false; - - match_results m; - typedef typename match_results::allocator_type match_alloc_type; - re_detail::perl_matcher matcher(first, last, m, e, flags | regex_constants::match_any, first); - return matcher.find(); -} - -#ifndef BOOST_NO_FUNCTION_TEMPLATE_ORDERING - -template -inline bool regex_search(const charT* str, - const basic_regex& e, - match_flag_type flags = match_default) -{ - return regex_search(str, str + traits::length(str), e, flags); -} - -template -inline bool regex_search(const std::basic_string& s, - const basic_regex& e, - match_flag_type flags = match_default) -{ - return regex_search(s.begin(), s.end(), e, flags); -} -#else // non-template function overloads -inline bool regex_search(const char* str, - const regex& e, - match_flag_type flags = match_default) -{ - cmatch m; - return regex_search(str, str + regex::traits_type::length(str), m, e, flags | regex_constants::match_any); -} -#ifndef BOOST_NO_WREGEX -inline bool regex_search(const wchar_t* str, - const wregex& e, - match_flag_type flags = match_default) -{ - wcmatch m; - return regex_search(str, str + wregex::traits_type::length(str), m, e, flags | regex_constants::match_any); -} -#endif -inline bool regex_search(const std::string& s, - const regex& e, - match_flag_type flags = match_default) -{ - smatch m; - return regex_search(s.begin(), s.end(), m, e, flags | regex_constants::match_any); -} -#if !defined(BOOST_NO_WREGEX) -inline bool regex_search(const std::basic_string& s, - const wregex& e, - match_flag_type flags = match_default) -{ - wsmatch m; - return regex_search(s.begin(), s.end(), m, e, flags | regex_constants::match_any); -} - -#endif // BOOST_NO_WREGEX - -#endif // partial overload - -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable: 4103) -#endif -#ifdef BOOST_HAS_ABI_HEADERS -# include BOOST_ABI_SUFFIX -#endif -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif - -} // namespace cutl_details_boost - -#endif // BOOST_REGEX_V4_REGEX_SEARCH_HPP - - diff --git a/cutl/details/boost/regex/v4/regex_split.hpp b/cutl/details/boost/regex/v4/regex_split.hpp deleted file mode 100644 index db1bac0..0000000 --- a/cutl/details/boost/regex/v4/regex_split.hpp +++ /dev/null @@ -1,172 +0,0 @@ -/* - * - * Copyright (c) 1998-2002 - * John Maddock - * - * Use, modification and distribution are subject to the - * Boost Software License, Version 1.0. (See accompanying file - * LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt) - * - */ - - /* - * LOCATION: see http://www.boost.org for most recent version. - * FILE regex_split.hpp - * VERSION see - * DESCRIPTION: Implements regex_split and associated functions. - * Note this is an internal header file included - * by regex.hpp, do not include on its own. - */ - -#ifndef BOOST_REGEX_SPLIT_HPP -#define BOOST_REGEX_SPLIT_HPP - -namespace cutl_details_boost{ - -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable: 4103) -#endif -#ifdef BOOST_HAS_ABI_HEADERS -# include BOOST_ABI_PREFIX -#endif -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif - -#ifdef BOOST_MSVC -# pragma warning(push) -# pragma warning(disable: 4800) -#endif - -namespace re_detail{ - -template -const basic_regex& get_default_expression(charT) -{ - static const charT expression_text[4] = { '\\', 's', '+', '\00', }; - static const basic_regex e(expression_text); - return e; -} - -template -class split_pred -{ - typedef std::basic_string string_type; - typedef typename string_type::const_iterator iterator_type; - iterator_type* p_last; - OutputIterator* p_out; - std::size_t* p_max; - std::size_t initial_max; -public: - split_pred(iterator_type* a, OutputIterator* b, std::size_t* c) - : p_last(a), p_out(b), p_max(c), initial_max(*c) {} - - bool operator()(const match_results& what); -}; - -template -bool split_pred::operator() - (const match_results& what) -{ - *p_last = what[0].second; - if(what.size() > 1) - { - // output sub-expressions only: - for(unsigned i = 1; i < what.size(); ++i) - { - *(*p_out) = what.str(i); - ++(*p_out); - if(0 == --*p_max) return false; - } - return *p_max != 0; - } - else - { - // output $` only if it's not-null or not at the start of the input: - const sub_match& sub = what[-1]; - if((sub.first != sub.second) || (*p_max != initial_max)) - { - *(*p_out) = sub.str(); - ++(*p_out); - return --*p_max; - } - } - // - // initial null, do nothing: - return true; -} - -} // namespace re_detail - -template -std::size_t regex_split(OutputIterator out, - std::basic_string& s, - const basic_regex& e, - match_flag_type flags, - std::size_t max_split) -{ - typedef typename std::basic_string::const_iterator ci_t; - //typedef typename match_results::allocator_type match_allocator; - ci_t last = s.begin(); - std::size_t init_size = max_split; - re_detail::split_pred pred(&last, &out, &max_split); - ci_t i, j; - i = s.begin(); - j = s.end(); - regex_grep(pred, i, j, e, flags); - // - // if there is still input left, do a final push as long as max_split - // is not exhausted, and we're not splitting sub-expressions rather - // than whitespace: - if(max_split && (last != s.end()) && (e.mark_count() == 1)) - { - *out = std::basic_string((ci_t)last, (ci_t)s.end()); - ++out; - last = s.end(); - --max_split; - } - // - // delete from the string everything that has been processed so far: - s.erase(0, last - s.begin()); - // - // return the number of new records pushed: - return init_size - max_split; -} - -template -inline std::size_t regex_split(OutputIterator out, - std::basic_string& s, - const basic_regex& e, - match_flag_type flags = match_default) -{ - return regex_split(out, s, e, flags, UINT_MAX); -} - -template -inline std::size_t regex_split(OutputIterator out, - std::basic_string& s) -{ - return regex_split(out, s, re_detail::get_default_expression(charT(0)), match_default, UINT_MAX); -} - -#ifdef BOOST_MSVC -# pragma warning(pop) -#endif - -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable: 4103) -#endif -#ifdef BOOST_HAS_ABI_HEADERS -# include BOOST_ABI_SUFFIX -#endif -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif - -} // namespace cutl_details_boost - -#endif - - diff --git a/cutl/details/boost/regex/v4/regex_token_iterator.hpp b/cutl/details/boost/regex/v4/regex_token_iterator.hpp deleted file mode 100644 index 608862e..0000000 --- a/cutl/details/boost/regex/v4/regex_token_iterator.hpp +++ /dev/null @@ -1,342 +0,0 @@ -/* - * - * Copyright (c) 2003 - * John Maddock - * - * Use, modification and distribution are subject to the - * Boost Software License, Version 1.0. (See accompanying file - * LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt) - * - */ - - /* - * LOCATION: see http://www.boost.org for most recent version. - * FILE regex_token_iterator.hpp - * VERSION see - * DESCRIPTION: Provides regex_token_iterator implementation. - */ - -#ifndef BOOST_REGEX_V4_REGEX_TOKEN_ITERATOR_HPP -#define BOOST_REGEX_V4_REGEX_TOKEN_ITERATOR_HPP - -#include -#include -#if (BOOST_WORKAROUND(__BORLANDC__, >= 0x560) && BOOST_WORKAROUND(__BORLANDC__, BOOST_TESTED_AT(0x570)))\ - || BOOST_WORKAROUND(BOOST_MSVC, < 1300) \ - || BOOST_WORKAROUND(__MWERKS__, BOOST_TESTED_AT(0x3003)) -// -// Borland C++ Builder 6, and Visual C++ 6, -// can't cope with the array template constructor -// so we have a template member that will accept any type as -// argument, and then assert that is really is an array: -// -#include -#include -#endif - -namespace cutl_details_boost{ - -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable: 4103) -#endif -#ifdef BOOST_HAS_ABI_HEADERS -# include BOOST_ABI_PREFIX -#endif -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif -#if BOOST_WORKAROUND(BOOST_MSVC, > 1300) -# pragma warning(push) -# pragma warning(disable:4700) -#endif - -template -class regex_token_iterator_implementation -{ - typedef basic_regex regex_type; - typedef sub_match value_type; - - match_results what; // current match - BidirectionalIterator base; // start of search area - BidirectionalIterator end; // end of search area - const regex_type re; // the expression - match_flag_type flags; // match flags - value_type result; // the current string result - int N; // the current sub-expression being enumerated - std::vector subs; // the sub-expressions to enumerate - -public: - regex_token_iterator_implementation(const regex_type* p, BidirectionalIterator last, int sub, match_flag_type f) - : end(last), re(*p), flags(f){ subs.push_back(sub); } - regex_token_iterator_implementation(const regex_type* p, BidirectionalIterator last, const std::vector& v, match_flag_type f) - : end(last), re(*p), flags(f), subs(v){} -#if !BOOST_WORKAROUND(__HP_aCC, < 60700) -#if (BOOST_WORKAROUND(__BORLANDC__, >= 0x560) && BOOST_WORKAROUND(__BORLANDC__, BOOST_TESTED_AT(0x570)))\ - || BOOST_WORKAROUND(BOOST_MSVC, < 1300) \ - || BOOST_WORKAROUND(__MWERKS__, BOOST_TESTED_AT(0x3003)) \ - || BOOST_WORKAROUND(__HP_aCC, < 60700) - template - regex_token_iterator_implementation(const regex_type* p, BidirectionalIterator last, const T& submatches, match_flag_type f) - : end(last), re(*p), flags(f) - { - // assert that T really is an array: - BOOST_STATIC_ASSERT(::cutl_details_boost::is_array::value); - const std::size_t array_size = sizeof(T) / sizeof(submatches[0]); - for(std::size_t i = 0; i < array_size; ++i) - { - subs.push_back(submatches[i]); - } - } -#else - template - regex_token_iterator_implementation(const regex_type* p, BidirectionalIterator last, const int (&submatches)[CN], match_flag_type f) - : end(last), re(*p), flags(f) - { - for(std::size_t i = 0; i < CN; ++i) - { - subs.push_back(submatches[i]); - } - } -#endif -#endif - bool init(BidirectionalIterator first) - { - N = 0; - base = first; - if(regex_search(first, end, what, re, flags, base) == true) - { - N = 0; - result = ((subs[N] == -1) ? what.prefix() : what[(int)subs[N]]); - return true; - } - else if((subs[N] == -1) && (first != end)) - { - result.first = first; - result.second = end; - result.matched = (first != end); - N = -1; - return true; - } - return false; - } - bool compare(const regex_token_iterator_implementation& that) - { - if(this == &that) return true; - return (&re.get_data() == &that.re.get_data()) - && (end == that.end) - && (flags == that.flags) - && (N == that.N) - && (what[0].first == that.what[0].first) - && (what[0].second == that.what[0].second); - } - const value_type& get() - { return result; } - bool next() - { - if(N == -1) - return false; - if(N+1 < (int)subs.size()) - { - ++N; - result =((subs[N] == -1) ? what.prefix() : what[subs[N]]); - return true; - } - //if(what.prefix().first != what[0].second) - // flags |= /*match_prev_avail |*/ regex_constants::match_not_bob; - BidirectionalIterator last_end(what[0].second); - if(regex_search(last_end, end, what, re, ((what[0].first == what[0].second) ? flags | regex_constants::match_not_initial_null : flags), base)) - { - N =0; - result =((subs[N] == -1) ? what.prefix() : what[subs[N]]); - return true; - } - else if((last_end != end) && (subs[0] == -1)) - { - N =-1; - result.first = last_end; - result.second = end; - result.matched = (last_end != end); - return true; - } - return false; - } -private: - regex_token_iterator_implementation& operator=(const regex_token_iterator_implementation&); -}; - -template ::value_type, - class traits = regex_traits > -class regex_token_iterator -#ifndef BOOST_NO_STD_ITERATOR - : public std::iterator< - std::forward_iterator_tag, - sub_match, - typename re_detail::regex_iterator_traits::difference_type, - const sub_match*, - const sub_match& > -#endif -{ -private: - typedef regex_token_iterator_implementation impl; - typedef shared_ptr pimpl; -public: - typedef basic_regex regex_type; - typedef sub_match value_type; - typedef typename re_detail::regex_iterator_traits::difference_type - difference_type; - typedef const value_type* pointer; - typedef const value_type& reference; - typedef std::forward_iterator_tag iterator_category; - - regex_token_iterator(){} - regex_token_iterator(BidirectionalIterator a, BidirectionalIterator b, const regex_type& re, - int submatch = 0, match_flag_type m = match_default) - : pdata(new impl(&re, b, submatch, m)) - { - if(!pdata->init(a)) - pdata.reset(); - } - regex_token_iterator(BidirectionalIterator a, BidirectionalIterator b, const regex_type& re, - const std::vector& submatches, match_flag_type m = match_default) - : pdata(new impl(&re, b, submatches, m)) - { - if(!pdata->init(a)) - pdata.reset(); - } -#if !BOOST_WORKAROUND(__HP_aCC, < 60700) -#if (BOOST_WORKAROUND(__BORLANDC__, >= 0x560) && BOOST_WORKAROUND(__BORLANDC__, BOOST_TESTED_AT(0x570)))\ - || BOOST_WORKAROUND(BOOST_MSVC, < 1300) \ - || BOOST_WORKAROUND(__MWERKS__, BOOST_TESTED_AT(0x3003)) \ - || BOOST_WORKAROUND(__HP_aCC, < 60700) - template - regex_token_iterator(BidirectionalIterator a, BidirectionalIterator b, const regex_type& re, - const T& submatches, match_flag_type m = match_default) - : pdata(new impl(&re, b, submatches, m)) - { - if(!pdata->init(a)) - pdata.reset(); - } -#else - template - regex_token_iterator(BidirectionalIterator a, BidirectionalIterator b, const regex_type& re, - const int (&submatches)[N], match_flag_type m = match_default) - : pdata(new impl(&re, b, submatches, m)) - { - if(!pdata->init(a)) - pdata.reset(); - } -#endif -#endif - regex_token_iterator(const regex_token_iterator& that) - : pdata(that.pdata) {} - regex_token_iterator& operator=(const regex_token_iterator& that) - { - pdata = that.pdata; - return *this; - } - bool operator==(const regex_token_iterator& that)const - { - if((pdata.get() == 0) || (that.pdata.get() == 0)) - return pdata.get() == that.pdata.get(); - return pdata->compare(*(that.pdata.get())); - } - bool operator!=(const regex_token_iterator& that)const - { return !(*this == that); } - const value_type& operator*()const - { return pdata->get(); } - const value_type* operator->()const - { return &(pdata->get()); } - regex_token_iterator& operator++() - { - cow(); - if(0 == pdata->next()) - { - pdata.reset(); - } - return *this; - } - regex_token_iterator operator++(int) - { - regex_token_iterator result(*this); - ++(*this); - return result; - } -private: - - pimpl pdata; - - void cow() - { - // copy-on-write - if(pdata.get() && !pdata.unique()) - { - pdata.reset(new impl(*(pdata.get()))); - } - } -}; - -typedef regex_token_iterator cregex_token_iterator; -typedef regex_token_iterator sregex_token_iterator; -#ifndef BOOST_NO_WREGEX -typedef regex_token_iterator wcregex_token_iterator; -typedef regex_token_iterator wsregex_token_iterator; -#endif - -template -inline regex_token_iterator make_regex_token_iterator(const charT* p, const basic_regex& e, int submatch = 0, regex_constants::match_flag_type m = regex_constants::match_default) -{ - return regex_token_iterator(p, p+traits::length(p), e, submatch, m); -} -template -inline regex_token_iterator::const_iterator, charT, traits> make_regex_token_iterator(const std::basic_string& p, const basic_regex& e, int submatch = 0, regex_constants::match_flag_type m = regex_constants::match_default) -{ - return regex_token_iterator::const_iterator, charT, traits>(p.begin(), p.end(), e, submatch, m); -} -#if !BOOST_WORKAROUND(BOOST_MSVC, < 1300) -template -inline regex_token_iterator make_regex_token_iterator(const charT* p, const basic_regex& e, const int (&submatch)[N], regex_constants::match_flag_type m = regex_constants::match_default) -{ - return regex_token_iterator(p, p+traits::length(p), e, submatch, m); -} -template -inline regex_token_iterator::const_iterator, charT, traits> make_regex_token_iterator(const std::basic_string& p, const basic_regex& e, const int (&submatch)[N], regex_constants::match_flag_type m = regex_constants::match_default) -{ - return regex_token_iterator::const_iterator, charT, traits>(p.begin(), p.end(), e, submatch, m); -} -#endif -template -inline regex_token_iterator make_regex_token_iterator(const charT* p, const basic_regex& e, const std::vector& submatch, regex_constants::match_flag_type m = regex_constants::match_default) -{ - return regex_token_iterator(p, p+traits::length(p), e, submatch, m); -} -template -inline regex_token_iterator::const_iterator, charT, traits> make_regex_token_iterator(const std::basic_string& p, const basic_regex& e, const std::vector& submatch, regex_constants::match_flag_type m = regex_constants::match_default) -{ - return regex_token_iterator::const_iterator, charT, traits>(p.begin(), p.end(), e, submatch, m); -} - -#if BOOST_WORKAROUND(BOOST_MSVC, > 1300) -# pragma warning(pop) -#endif -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable: 4103) -#endif -#ifdef BOOST_HAS_ABI_HEADERS -# include BOOST_ABI_SUFFIX -#endif -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif - -} // namespace cutl_details_boost - -#endif // BOOST_REGEX_V4_REGEX_TOKEN_ITERATOR_HPP - - - - diff --git a/cutl/details/boost/regex/v4/regex_traits.hpp b/cutl/details/boost/regex/v4/regex_traits.hpp deleted file mode 100644 index b632864..0000000 --- a/cutl/details/boost/regex/v4/regex_traits.hpp +++ /dev/null @@ -1,189 +0,0 @@ -/* - * - * Copyright (c) 2003 - * John Maddock - * - * Use, modification and distribution are subject to the - * Boost Software License, Version 1.0. (See accompanying file - * LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt) - * - */ - - /* - * LOCATION: see http://www.boost.org for most recent version. - * FILE regex_traits.hpp - * VERSION see - * DESCRIPTION: Declares regular expression traits classes. - */ - -#ifndef BOOST_REGEX_TRAITS_HPP_INCLUDED -#define BOOST_REGEX_TRAITS_HPP_INCLUDED - -#ifndef BOOST_REGEX_CONFIG_HPP -#include -#endif -#ifndef BOOST_REGEX_WORKAROUND_HPP -#include -#endif -#ifndef BOOST_REGEX_SYNTAX_TYPE_HPP -#include -#endif -#ifndef BOOST_REGEX_ERROR_TYPE_HPP -#include -#endif -#ifndef BOOST_REGEX_TRAITS_DEFAULTS_HPP_INCLUDED -#include -#endif -#ifndef BOOST_NO_STD_LOCALE -# ifndef BOOST_CPP_REGEX_TRAITS_HPP_INCLUDED -# include -# endif -#endif -#if !BOOST_WORKAROUND(__BORLANDC__, < 0x560) -# ifndef BOOST_C_REGEX_TRAITS_HPP_INCLUDED -# include -# endif -#endif -#if defined(_WIN32) && !defined(BOOST_REGEX_NO_W32) -# ifndef BOOST_W32_REGEX_TRAITS_HPP_INCLUDED -# include -# endif -#endif -#ifndef BOOST_REGEX_FWD_HPP_INCLUDED -#include -#endif - -#include "cutl/details/boost/mpl/has_xxx.hpp" -#include - -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable: 4103) -#endif -#ifdef BOOST_HAS_ABI_HEADERS -# include BOOST_ABI_PREFIX -#endif -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif - -namespace cutl_details_boost{ - -template -struct regex_traits : public implementationT -{ - regex_traits() : implementationT() {} -}; - -// -// class regex_traits_wrapper. -// this is what our implementation will actually store; -// it provides default implementations of the "optional" -// interfaces that we support, in addition to the -// required "standard" ones: -// -namespace re_detail{ -#if !defined(BOOST_NO_TEMPLATE_PARTIAL_SPECIALIZATION) && !BOOST_WORKAROUND(__HP_aCC, < 60000) -BOOST_MPL_HAS_XXX_TRAIT_DEF(boost_extensions_tag) -#else -template -struct has_boost_extensions_tag -{ - BOOST_STATIC_CONSTANT(bool, value = false); -}; -#endif - -template -struct default_wrapper : public BaseT -{ - typedef typename BaseT::char_type char_type; - std::string error_string(::cutl_details_boost::regex_constants::error_type e)const - { - return ::cutl_details_boost::re_detail::get_default_error_string(e); - } - ::cutl_details_boost::regex_constants::syntax_type syntax_type(char_type c)const - { - return ((c & 0x7f) == c) ? get_default_syntax_type(static_cast(c)) : ::cutl_details_boost::regex_constants::syntax_char; - } - ::cutl_details_boost::regex_constants::escape_syntax_type escape_syntax_type(char_type c)const - { - return ((c & 0x7f) == c) ? get_default_escape_syntax_type(static_cast(c)) : ::cutl_details_boost::regex_constants::escape_type_identity; - } - int toi(const char_type*& p1, const char_type* p2, int radix)const - { - return ::cutl_details_boost::re_detail::global_toi(p1, p2, radix, *this); - } - char_type translate(char_type c, bool icase)const - { - return (icase ? this->translate_nocase(c) : this->translate(c)); - } - char_type translate(char_type c)const - { - return BaseT::translate(c); - } - char_type tolower(char_type c)const - { - return ::cutl_details_boost::re_detail::global_lower(c); - } - char_type toupper(char_type c)const - { - return ::cutl_details_boost::re_detail::global_upper(c); - } -}; - -template -struct compute_wrapper_base -{ - typedef BaseT type; -}; -#if !defined(BOOST_NO_TEMPLATE_PARTIAL_SPECIALIZATION) && !BOOST_WORKAROUND(__HP_aCC, < 60000) -template -struct compute_wrapper_base -{ - typedef default_wrapper type; -}; -#else -template <> -struct compute_wrapper_base, false> -{ - typedef default_wrapper > type; -}; -#ifndef BOOST_NO_WREGEX -template <> -struct compute_wrapper_base, false> -{ - typedef default_wrapper > type; -}; -#endif -#endif - -} // namespace re_detail - -template -struct regex_traits_wrapper - : public ::cutl_details_boost::re_detail::compute_wrapper_base< - BaseT, - ::cutl_details_boost::re_detail::has_boost_extensions_tag::value - >::type -{ - regex_traits_wrapper(){} -private: - regex_traits_wrapper(const regex_traits_wrapper&); - regex_traits_wrapper& operator=(const regex_traits_wrapper&); -}; - -} // namespace cutl_details_boost - -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable: 4103) -#endif -#ifdef BOOST_HAS_ABI_HEADERS -# include BOOST_ABI_SUFFIX -#endif -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif - -#endif // include - diff --git a/cutl/details/boost/regex/v4/regex_traits_defaults.hpp b/cutl/details/boost/regex/v4/regex_traits_defaults.hpp deleted file mode 100644 index eeb0f2e..0000000 --- a/cutl/details/boost/regex/v4/regex_traits_defaults.hpp +++ /dev/null @@ -1,371 +0,0 @@ -/* - * - * Copyright (c) 2004 - * John Maddock - * - * Use, modification and distribution are subject to the - * Boost Software License, Version 1.0. (See accompanying file - * LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt) - * - */ - - /* - * LOCATION: see http://www.boost.org for most recent version. - * FILE regex_traits_defaults.hpp - * VERSION see - * DESCRIPTION: Declares API's for access to regex_traits default properties. - */ - -#ifndef BOOST_REGEX_TRAITS_DEFAULTS_HPP_INCLUDED -#define BOOST_REGEX_TRAITS_DEFAULTS_HPP_INCLUDED - -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable: 4103) -#endif -#ifdef BOOST_HAS_ABI_HEADERS -# include BOOST_ABI_PREFIX -#endif -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif - -#ifndef BOOST_REGEX_SYNTAX_TYPE_HPP -#include -#endif -#ifndef BOOST_REGEX_ERROR_TYPE_HPP -#include -#endif - -#ifdef BOOST_NO_STDC_NAMESPACE -namespace std{ - using ::strlen; -} -#endif - -namespace cutl_details_boost{ namespace re_detail{ - - -// -// helpers to suppress warnings: -// -template -inline bool is_extended(charT c) -{ return c > 256; } -inline bool is_extended(char) -{ return false; } - - -BOOST_REGEX_DECL const char* BOOST_REGEX_CALL get_default_syntax(regex_constants::syntax_type n); -BOOST_REGEX_DECL const char* BOOST_REGEX_CALL get_default_error_string(regex_constants::error_type n); -BOOST_REGEX_DECL regex_constants::syntax_type BOOST_REGEX_CALL get_default_syntax_type(char c); -BOOST_REGEX_DECL regex_constants::escape_syntax_type BOOST_REGEX_CALL get_default_escape_syntax_type(char c); - -// is charT c a combining character? -BOOST_REGEX_DECL bool BOOST_REGEX_CALL is_combining_implementation(uint_least16_t s); - -template -inline bool is_combining(charT c) -{ - return (c <= static_cast(0)) ? false : ((c >= static_cast((std::numeric_limits::max)())) ? false : is_combining_implementation(static_cast(c))); -} -template <> -inline bool is_combining(char) -{ - return false; -} -template <> -inline bool is_combining(signed char) -{ - return false; -} -template <> -inline bool is_combining(unsigned char) -{ - return false; -} -#if !defined(__hpux) && !defined(__WINSCW__) // can't use WCHAR_MAX/MIN in pp-directives -#ifdef _MSC_VER -template<> -inline bool is_combining(wchar_t c) -{ - return is_combining_implementation(static_cast(c)); -} -#elif !defined(__DECCXX) && !defined(__osf__) && !defined(__OSF__) && defined(WCHAR_MIN) && (WCHAR_MIN == 0) && !defined(BOOST_NO_INTRINSIC_WCHAR_T) -#if defined(WCHAR_MAX) && (WCHAR_MAX <= USHRT_MAX) -template<> -inline bool is_combining(wchar_t c) -{ - return is_combining_implementation(static_cast(c)); -} -#else -template<> -inline bool is_combining(wchar_t c) -{ - return (c >= (std::numeric_limits::max)()) ? false : is_combining_implementation(static_cast(c)); -} -#endif -#endif -#endif - -// -// is a charT c a line separator? -// -template -inline bool is_separator(charT c) -{ - return BOOST_REGEX_MAKE_BOOL( - (c == static_cast('\n')) - || (c == static_cast('\r')) - || (c == static_cast('\f')) - || (static_cast(c) == 0x2028u) - || (static_cast(c) == 0x2029u) - || (static_cast(c) == 0x85u)); -} -template <> -inline bool is_separator(char c) -{ - return BOOST_REGEX_MAKE_BOOL((c == '\n') || (c == '\r') || (c == '\f')); -} - -// -// get a default collating element: -// -BOOST_REGEX_DECL std::string BOOST_REGEX_CALL lookup_default_collate_name(const std::string& name); - -// -// get the state_id of a character clasification, the individual -// traits classes then transform that state_id into a bitmask: -// -template -struct character_pointer_range -{ - const charT* p1; - const charT* p2; - - bool operator < (const character_pointer_range& r)const - { - return std::lexicographical_compare(p1, p2, r.p1, r.p2); - } - bool operator == (const character_pointer_range& r)const - { - // Not only do we check that the ranges are of equal size before - // calling std::equal, but there is no other algorithm available: - // not even a non-standard MS one. So forward to unchecked_equal - // in the MS case. - return ((p2 - p1) == (r.p2 - r.p1)) && re_detail::equal(p1, p2, r.p1); - } -}; -template -int get_default_class_id(const charT* p1, const charT* p2) -{ - static const charT data[73] = { - 'a', 'l', 'n', 'u', 'm', - 'a', 'l', 'p', 'h', 'a', - 'b', 'l', 'a', 'n', 'k', - 'c', 'n', 't', 'r', 'l', - 'd', 'i', 'g', 'i', 't', - 'g', 'r', 'a', 'p', 'h', - 'l', 'o', 'w', 'e', 'r', - 'p', 'r', 'i', 'n', 't', - 'p', 'u', 'n', 'c', 't', - 's', 'p', 'a', 'c', 'e', - 'u', 'n', 'i', 'c', 'o', 'd', 'e', - 'u', 'p', 'p', 'e', 'r', - 'v', - 'w', 'o', 'r', 'd', - 'x', 'd', 'i', 'g', 'i', 't', - }; - - static const character_pointer_range ranges[21] = - { - {data+0, data+5,}, // alnum - {data+5, data+10,}, // alpha - {data+10, data+15,}, // blank - {data+15, data+20,}, // cntrl - {data+20, data+21,}, // d - {data+20, data+25,}, // digit - {data+25, data+30,}, // graph - {data+29, data+30,}, // h - {data+30, data+31,}, // l - {data+30, data+35,}, // lower - {data+35, data+40,}, // print - {data+40, data+45,}, // punct - {data+45, data+46,}, // s - {data+45, data+50,}, // space - {data+57, data+58,}, // u - {data+50, data+57,}, // unicode - {data+57, data+62,}, // upper - {data+62, data+63,}, // v - {data+63, data+64,}, // w - {data+63, data+67,}, // word - {data+67, data+73,}, // xdigit - }; - static const character_pointer_range* ranges_begin = ranges; - static const character_pointer_range* ranges_end = ranges + (sizeof(ranges)/sizeof(ranges[0])); - - character_pointer_range t = { p1, p2, }; - const character_pointer_range* p = std::lower_bound(ranges_begin, ranges_end, t); - if((p != ranges_end) && (t == *p)) - return static_cast(p - ranges); - return -1; -} - -// -// helper functions: -// -template -std::ptrdiff_t global_length(const charT* p) -{ - std::ptrdiff_t n = 0; - while(*p) - { - ++p; - ++n; - } - return n; -} -template<> -inline std::ptrdiff_t global_length(const char* p) -{ - return (std::strlen)(p); -} -#ifndef BOOST_NO_WREGEX -template<> -inline std::ptrdiff_t global_length(const wchar_t* p) -{ - return (std::wcslen)(p); -} -#endif -template -inline charT BOOST_REGEX_CALL global_lower(charT c) -{ - return c; -} -template -inline charT BOOST_REGEX_CALL global_upper(charT c) -{ - return c; -} - -BOOST_REGEX_DECL char BOOST_REGEX_CALL do_global_lower(char c); -BOOST_REGEX_DECL char BOOST_REGEX_CALL do_global_upper(char c); -#ifndef BOOST_NO_WREGEX -BOOST_REGEX_DECL wchar_t BOOST_REGEX_CALL do_global_lower(wchar_t c); -BOOST_REGEX_DECL wchar_t BOOST_REGEX_CALL do_global_upper(wchar_t c); -#endif -#ifdef BOOST_REGEX_HAS_OTHER_WCHAR_T -BOOST_REGEX_DECL unsigned short BOOST_REGEX_CALL do_global_lower(unsigned short c); -BOOST_REGEX_DECL unsigned short BOOST_REGEX_CALL do_global_upper(unsigned short c); -#endif -// -// This sucks: declare template specialisations of global_lower/global_upper -// that just forward to the non-template implementation functions. We do -// this because there is one compiler (Compaq Tru64 C++) that doesn't seem -// to differentiate between templates and non-template overloads.... -// what's more, the primary template, plus all overloads have to be -// defined in the same translation unit (if one is inline they all must be) -// otherwise the "local template instantiation" compiler option can pick -// the wrong instantiation when linking: -// -template<> inline char BOOST_REGEX_CALL global_lower(char c){ return do_global_lower(c); } -template<> inline char BOOST_REGEX_CALL global_upper(char c){ return do_global_upper(c); } -#ifndef BOOST_NO_WREGEX -template<> inline wchar_t BOOST_REGEX_CALL global_lower(wchar_t c){ return do_global_lower(c); } -template<> inline wchar_t BOOST_REGEX_CALL global_upper(wchar_t c){ return do_global_upper(c); } -#endif -#ifdef BOOST_REGEX_HAS_OTHER_WCHAR_T -template<> inline unsigned short BOOST_REGEX_CALL global_lower(unsigned short c){ return do_global_lower(c); } -template<> inline unsigned short BOOST_REGEX_CALL global_upper(unsigned short c){ return do_global_upper(c); } -#endif - -template -int global_value(charT c) -{ - static const charT zero = '0'; - static const charT nine = '9'; - static const charT a = 'a'; - static const charT f = 'f'; - static const charT A = 'A'; - static const charT F = 'F'; - - if(c > f) return -1; - if(c >= a) return 10 + (c - a); - if(c > F) return -1; - if(c >= A) return 10 + (c - A); - if(c > nine) return -1; - if(c >= zero) return c - zero; - return -1; -} -template -int global_toi(const charT*& p1, const charT* p2, int radix, const traits& t) -{ - (void)t; // warning suppression - int next_value = t.value(*p1, radix); - if((p1 == p2) || (next_value < 0) || (next_value >= radix)) - return -1; - int result = 0; - while(p1 != p2) - { - next_value = t.value(*p1, radix); - if((next_value < 0) || (next_value >= radix)) - break; - result *= radix; - result += next_value; - ++p1; - } - return result; -} - -template -inline const charT* get_escape_R_string() -{ -#ifdef BOOST_MSVC -# pragma warning(push) -# pragma warning(disable:4309 4245) -#endif - static const charT e1[] = { '(', '?', '>', '\x0D', '\x0A', '?', - '|', '[', '\x0A', '\x0B', '\x0C', static_cast('\x85'), '\\', 'x', '{', '2', '0', '2', '8', '}', - '\\', 'x', '{', '2', '0', '2', '9', '}', ']', ')', '\0' }; - static const charT e2[] = { '(', '?', '>', '\x0D', '\x0A', '?', - '|', '[', '\x0A', '\x0B', '\x0C', static_cast('\x85'), ']', ')', '\0' }; - - charT c = static_cast(0x2029u); - bool b = (static_cast(c) == 0x2029u); - - return (b ? e1 : e2); -#ifdef BOOST_MSVC -# pragma warning(pop) -#endif -} - -template <> -inline const char* get_escape_R_string() -{ -#ifdef BOOST_MSVC -# pragma warning(push) -# pragma warning(disable:4309) -#endif - static const char e2[] = { '(', '?', '>', '\x0D', '\x0A', '?', - '|', '[', '\x0A', '\x0B', '\x0C', '\x85', ']', ')', '\0' }; - return e2; -#ifdef BOOST_MSVC -# pragma warning(pop) -#endif -} - -} // re_detail -} // boost - -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable: 4103) -#endif -#ifdef BOOST_HAS_ABI_HEADERS -# include BOOST_ABI_SUFFIX -#endif -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif - -#endif diff --git a/cutl/details/boost/regex/v4/regex_workaround.hpp b/cutl/details/boost/regex/v4/regex_workaround.hpp deleted file mode 100644 index 496ab8c..0000000 --- a/cutl/details/boost/regex/v4/regex_workaround.hpp +++ /dev/null @@ -1,232 +0,0 @@ -/* - * - * Copyright (c) 1998-2005 - * John Maddock - * - * Use, modification and distribution are subject to the - * Boost Software License, Version 1.0. (See accompanying file - * LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt) - * - */ - - /* - * LOCATION: see http://www.boost.org for most recent version. - * FILE regex_workarounds.cpp - * VERSION see - * DESCRIPTION: Declares Misc workarounds. - */ - -#ifndef BOOST_REGEX_WORKAROUND_HPP -#define BOOST_REGEX_WORKAROUND_HPP - - -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#ifndef BOOST_NO_STD_LOCALE -# include -#endif - -#if defined(BOOST_NO_STDC_NAMESPACE) -namespace std{ - using ::sprintf; using ::strcpy; using ::strcat; using ::strlen; -} -#endif - -namespace cutl_details_boost{ namespace re_detail{ -#ifdef BOOST_NO_STD_DISTANCE -template -std::ptrdiff_t distance(const T& x, const T& y) -{ return y - x; } -#else -using std::distance; -#endif -}} - - -#ifdef BOOST_REGEX_NO_BOOL -# define BOOST_REGEX_MAKE_BOOL(x) static_cast((x) ? true : false) -#else -# define BOOST_REGEX_MAKE_BOOL(x) static_cast(x) -#endif - -/***************************************************************************** - * - * Fix broken broken namespace support: - * - ****************************************************************************/ - -#if defined(BOOST_NO_STDC_NAMESPACE) && defined(__cplusplus) - -namespace std{ - using ::ptrdiff_t; - using ::size_t; - using ::abs; - using ::memset; - using ::memcpy; -} - -#endif - -/***************************************************************************** - * - * helper functions pointer_construct/pointer_destroy: - * - ****************************************************************************/ - -#ifdef __cplusplus -namespace cutl_details_boost{ namespace re_detail{ - -#ifdef BOOST_MSVC -#pragma warning (push) -#pragma warning (disable : 4100) -#endif - -template -inline void pointer_destroy(T* p) -{ p->~T(); (void)p; } - -#ifdef BOOST_MSVC -#pragma warning (pop) -#endif - -template -inline void pointer_construct(T* p, const T& t) -{ new (p) T(t); } - -}} // namespaces -#endif - -/***************************************************************************** - * - * helper function copy: - * - ****************************************************************************/ - -#ifdef __cplusplus -namespace cutl_details_boost{ namespace re_detail{ -#if BOOST_WORKAROUND(BOOST_MSVC,>=1400) && BOOST_WORKAROUND(BOOST_MSVC, <1600) && defined(_CPPLIB_VER) && defined(BOOST_DINKUMWARE_STDLIB) && !(defined(__SGI_STL_PORT) || defined(_STLPORT_VERSION)) - // - // MSVC 8 will either emit warnings or else refuse to compile - // code that makes perfectly legitimate use of std::copy, when - // the OutputIterator type is a user-defined class (apparently all user - // defined iterators are "unsafe"). This code works around that: - // - template - inline OutputIterator copy( - InputIterator first, - InputIterator last, - OutputIterator dest - ) - { - return stdext::unchecked_copy(first, last, dest); - } - template - inline bool equal( - InputIterator1 first, - InputIterator1 last, - InputIterator2 with - ) - { - return stdext::unchecked_equal(first, last, with); - } -#elif BOOST_WORKAROUND(BOOST_MSVC, > 1500) - // - // MSVC 10 will either emit warnings or else refuse to compile - // code that makes perfectly legitimate use of std::copy, when - // the OutputIterator type is a user-defined class (apparently all user - // defined iterators are "unsafe"). What's more Microsoft have removed their - // non-standard "unchecked" versions, even though their still in the MS - // documentation!! Work around this as best we can: - // - template - inline OutputIterator copy( - InputIterator first, - InputIterator last, - OutputIterator dest - ) - { - while(first != last) - *dest++ = *first++; - return dest; - } - template - inline bool equal( - InputIterator1 first, - InputIterator1 last, - InputIterator2 with - ) - { - while(first != last) - if(*first++ != *with++) return false; - return true; - } -#else - using std::copy; - using std::equal; -#endif -#if BOOST_WORKAROUND(BOOST_MSVC,>=1400) && defined(__STDC_WANT_SECURE_LIB__) && __STDC_WANT_SECURE_LIB__ - - // use safe versions of strcpy etc: - using ::strcpy_s; - using ::strcat_s; -#else - inline std::size_t strcpy_s( - char *strDestination, - std::size_t sizeInBytes, - const char *strSource - ) - { - if(std::strlen(strSource)+1 > sizeInBytes) - return 1; - std::strcpy(strDestination, strSource); - return 0; - } - inline std::size_t strcat_s( - char *strDestination, - std::size_t sizeInBytes, - const char *strSource - ) - { - if(std::strlen(strSource) + std::strlen(strDestination) + 1 > sizeInBytes) - return 1; - std::strcat(strDestination, strSource); - return 0; - } - -#endif - - inline void overflow_error_if_not_zero(std::size_t i) - { - if(i) - { - std::overflow_error e("String buffer too small"); - cutl_details_boost::throw_exception(e); - } - } - -}} // namespaces - -#endif // __cplusplus - -#endif // include guard - diff --git a/cutl/details/boost/regex/v4/states.hpp b/cutl/details/boost/regex/v4/states.hpp deleted file mode 100644 index 8f1152a..0000000 --- a/cutl/details/boost/regex/v4/states.hpp +++ /dev/null @@ -1,301 +0,0 @@ -/* - * - * Copyright (c) 1998-2002 - * John Maddock - * - * Use, modification and distribution are subject to the - * Boost Software License, Version 1.0. (See accompanying file - * LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt) - * - */ - - /* - * LOCATION: see http://www.boost.org for most recent version. - * FILE states.cpp - * VERSION see - * DESCRIPTION: Declares internal state machine structures. - */ - -#ifndef BOOST_REGEX_V4_STATES_HPP -#define BOOST_REGEX_V4_STATES_HPP - -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable: 4103) -#endif -#ifdef BOOST_HAS_ABI_HEADERS -# include BOOST_ABI_PREFIX -#endif -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif - -namespace cutl_details_boost{ -namespace re_detail{ - -/*** mask_type ******************************************************* -Whenever we have a choice of two alternatives, we use an array of bytes -to indicate which of the two alternatives it is possible to take for any -given input character. If mask_take is set, then we can take the next -state, and if mask_skip is set then we can take the alternative. -***********************************************************************/ -enum mask_type -{ - mask_take = 1, - mask_skip = 2, - mask_init = 4, - mask_any = mask_skip | mask_take, - mask_all = mask_any -}; - -/*** helpers ********************************************************** -These helpers let us use function overload resolution to detect whether -we have narrow or wide character strings: -***********************************************************************/ -struct _narrow_type{}; -struct _wide_type{}; -template struct is_byte; -template<> struct is_byte { typedef _narrow_type width_type; }; -template<> struct is_byte{ typedef _narrow_type width_type; }; -template<> struct is_byte { typedef _narrow_type width_type; }; -template struct is_byte { typedef _wide_type width_type; }; - -/*** enum syntax_element_type ****************************************** -Every record in the state machine falls into one of the following types: -***********************************************************************/ -enum syntax_element_type -{ - // start of a marked sub-expression, or perl-style (?...) extension - syntax_element_startmark = 0, - // end of a marked sub-expression, or perl-style (?...) extension - syntax_element_endmark = syntax_element_startmark + 1, - // any sequence of literal characters - syntax_element_literal = syntax_element_endmark + 1, - // start of line assertion: ^ - syntax_element_start_line = syntax_element_literal + 1, - // end of line assertion $ - syntax_element_end_line = syntax_element_start_line + 1, - // match any character: . - syntax_element_wild = syntax_element_end_line + 1, - // end of expression: we have a match when we get here - syntax_element_match = syntax_element_wild + 1, - // perl style word boundary: \b - syntax_element_word_boundary = syntax_element_match + 1, - // perl style within word boundary: \B - syntax_element_within_word = syntax_element_word_boundary + 1, - // start of word assertion: \< - syntax_element_word_start = syntax_element_within_word + 1, - // end of word assertion: \> - syntax_element_word_end = syntax_element_word_start + 1, - // start of buffer assertion: \` - syntax_element_buffer_start = syntax_element_word_end + 1, - // end of buffer assertion: \' - syntax_element_buffer_end = syntax_element_buffer_start + 1, - // backreference to previously matched sub-expression - syntax_element_backref = syntax_element_buffer_end + 1, - // either a wide character set [..] or one with multicharacter collating elements: - syntax_element_long_set = syntax_element_backref + 1, - // narrow character set: [...] - syntax_element_set = syntax_element_long_set + 1, - // jump to a new state in the machine: - syntax_element_jump = syntax_element_set + 1, - // choose between two production states: - syntax_element_alt = syntax_element_jump + 1, - // a repeat - syntax_element_rep = syntax_element_alt + 1, - // match a combining character sequence - syntax_element_combining = syntax_element_rep + 1, - // perl style soft buffer end: \z - syntax_element_soft_buffer_end = syntax_element_combining + 1, - // perl style continuation: \G - syntax_element_restart_continue = syntax_element_soft_buffer_end + 1, - // single character repeats: - syntax_element_dot_rep = syntax_element_restart_continue + 1, - syntax_element_char_rep = syntax_element_dot_rep + 1, - syntax_element_short_set_rep = syntax_element_char_rep + 1, - syntax_element_long_set_rep = syntax_element_short_set_rep + 1, - // a backstep for lookbehind repeats: - syntax_element_backstep = syntax_element_long_set_rep + 1, - // an assertion that a mark was matched: - syntax_element_assert_backref = syntax_element_backstep + 1, - syntax_element_toggle_case = syntax_element_assert_backref + 1, - // a recursive expression: - syntax_element_recurse = syntax_element_toggle_case + 1 -}; - -#ifdef BOOST_REGEX_DEBUG -// dwa 09/26/00 - This is needed to suppress warnings about an ambiguous conversion -std::ostream& operator<<(std::ostream&, syntax_element_type); -#endif - -struct re_syntax_base; - -/*** union offset_type ************************************************ -Points to another state in the machine. During machine construction -we use integral offsets, but these are converted to pointers before -execution of the machine. -***********************************************************************/ -union offset_type -{ - re_syntax_base* p; - std::ptrdiff_t i; -}; - -/*** struct re_syntax_base ******************************************** -Base class for all states in the machine. -***********************************************************************/ -struct re_syntax_base -{ - syntax_element_type type; // what kind of state this is - offset_type next; // next state in the machine -}; - -/*** struct re_brace ************************************************** -A marked parenthesis. -***********************************************************************/ -struct re_brace : public re_syntax_base -{ - // The index to match, can be zero (don't mark the sub-expression) - // or negative (for perl style (?...) extentions): - int index; - bool icase; -}; - -/*** struct re_dot ************************************************** -Match anything. -***********************************************************************/ -enum -{ - dont_care = 1, - force_not_newline = 0, - force_newline = 2, - - test_not_newline = 2, - test_newline = 3 -}; -struct re_dot : public re_syntax_base -{ - unsigned char mask; -}; - -/*** struct re_literal ************************************************ -A string of literals, following this structure will be an -array of characters: charT[length] -***********************************************************************/ -struct re_literal : public re_syntax_base -{ - unsigned int length; -}; - -/*** struct re_case ************************************************ -Indicates whether we are moving to a case insensive block or not -***********************************************************************/ -struct re_case : public re_syntax_base -{ - bool icase; -}; - -/*** struct re_set_long *********************************************** -A wide character set of characters, following this structure will be -an array of type charT: -First csingles null-terminated strings -Then 2 * cranges NULL terminated strings -Then cequivalents NULL terminated strings -***********************************************************************/ -template -struct re_set_long : public re_syntax_base -{ - unsigned int csingles, cranges, cequivalents; - mask_type cclasses; - mask_type cnclasses; - bool isnot; - bool singleton; -}; - -/*** struct re_set **************************************************** -A set of narrow-characters, matches any of _map which is none-zero -***********************************************************************/ -struct re_set : public re_syntax_base -{ - unsigned char _map[1 << CHAR_BIT]; -}; - -/*** struct re_jump *************************************************** -Jump to a new location in the machine (not next). -***********************************************************************/ -struct re_jump : public re_syntax_base -{ - offset_type alt; // location to jump to -}; - -/*** struct re_alt *************************************************** -Jump to a new location in the machine (possibly next). -***********************************************************************/ -struct re_alt : public re_jump -{ - unsigned char _map[1 << CHAR_BIT]; // which characters can take the jump - unsigned int can_be_null; // true if we match a NULL string -}; - -/*** struct re_repeat ************************************************* -Repeat a section of the machine -***********************************************************************/ -struct re_repeat : public re_alt -{ - std::size_t min, max; // min and max allowable repeats - int state_id; // Unique identifier for this repeat - bool leading; // True if this repeat is at the start of the machine (lets us optimize some searches) - bool greedy; // True if this is a greedy repeat -}; - -/*** struct re_recurse ************************************************ -Recurse to a particular subexpression. -**********************************************************************/ -struct re_recurse : public re_jump -{ - int state_id; // identifier of first nested repeat within the recursion. -}; - -/*** enum re_jump_size_type ******************************************* -Provides compiled size of re_jump structure (allowing for trailing alignment). -We provide this so we know how manybytes to insert when constructing the machine -(The value of padding_mask is defined in regex_raw_buffer.hpp). -***********************************************************************/ -enum re_jump_size_type -{ - re_jump_size = (sizeof(re_jump) + padding_mask) & ~(padding_mask), - re_repeater_size = (sizeof(re_repeat) + padding_mask) & ~(padding_mask), - re_alt_size = (sizeof(re_alt) + padding_mask) & ~(padding_mask) -}; - -/*** proc re_is_set_member ********************************************* -Forward declaration: we'll need this one later... -***********************************************************************/ - -template -struct regex_data; - -template -iterator BOOST_REGEX_CALL re_is_set_member(iterator next, - iterator last, - const re_set_long* set_, - const regex_data& e, bool icase); - -} // namespace re_detail - -} // namespace cutl_details_boost - -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable: 4103) -#endif -#ifdef BOOST_HAS_ABI_HEADERS -# include BOOST_ABI_SUFFIX -#endif -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif - -#endif - - diff --git a/cutl/details/boost/regex/v4/sub_match.hpp b/cutl/details/boost/regex/v4/sub_match.hpp deleted file mode 100644 index 53f0b27..0000000 --- a/cutl/details/boost/regex/v4/sub_match.hpp +++ /dev/null @@ -1,512 +0,0 @@ -/* - * - * Copyright (c) 1998-2002 - * John Maddock - * - * Use, modification and distribution are subject to the - * Boost Software License, Version 1.0. (See accompanying file - * LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt) - * - */ - - /* - * LOCATION: see http://www.boost.org for most recent version. - * FILE sub_match.cpp - * VERSION see - * DESCRIPTION: Declares template class sub_match. - */ - -#ifndef BOOST_REGEX_V4_SUB_MATCH_HPP -#define BOOST_REGEX_V4_SUB_MATCH_HPP - -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable: 4103) -#endif -#ifdef BOOST_HAS_ABI_HEADERS -# include BOOST_ABI_PREFIX -#endif -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif - -namespace cutl_details_boost{ - -template -struct sub_match : public std::pair -{ - typedef typename re_detail::regex_iterator_traits::value_type value_type; -#if defined(BOOST_NO_STD_ITERATOR_TRAITS) || defined(BOOST_NO_TEMPLATE_PARTIAL_SPECIALIZATION) - typedef std::ptrdiff_t difference_type; -#else - typedef typename re_detail::regex_iterator_traits::difference_type difference_type; -#endif - typedef BidiIterator iterator_type; - typedef BidiIterator iterator; - typedef BidiIterator const_iterator; - - bool matched; - - sub_match() : std::pair(), matched(false) {} - sub_match(BidiIterator i) : std::pair(i, i), matched(false) {} -#if !defined(BOOST_NO_TEMPLATED_ITERATOR_CONSTRUCTORS)\ - && !BOOST_WORKAROUND(BOOST_MSVC, < 1310)\ - && !BOOST_WORKAROUND(__BORLANDC__, <= 0x0551)\ - && !BOOST_WORKAROUND(__DECCXX_VER, BOOST_TESTED_AT(60590042)) - template - operator std::basic_string ()const - { - return matched ? std::basic_string(this->first, this->second) : std::basic_string(); - } -#else - operator std::basic_string ()const - { - return str(); - } -#endif - difference_type BOOST_REGEX_CALL length()const - { - difference_type n = matched ? ::cutl_details_boost::re_detail::distance((BidiIterator)this->first, (BidiIterator)this->second) : 0; - return n; - } - std::basic_string str()const - { - std::basic_string result; - if(matched) - { - std::size_t len = ::cutl_details_boost::re_detail::distance((BidiIterator)this->first, (BidiIterator)this->second); - result.reserve(len); - BidiIterator i = this->first; - while(i != this->second) - { - result.append(1, *i); - ++i; - } - } - return result; - } - int compare(const sub_match& s)const - { - if(matched != s.matched) - return static_cast(matched) - static_cast(s.matched); - return str().compare(s.str()); - } - int compare(const std::basic_string& s)const - { - return str().compare(s); - } - int compare(const value_type* p)const - { - return str().compare(p); - } - - bool operator==(const sub_match& that)const - { return compare(that) == 0; } - bool BOOST_REGEX_CALL operator !=(const sub_match& that)const - { return compare(that) != 0; } - bool operator<(const sub_match& that)const - { return compare(that) < 0; } - bool operator>(const sub_match& that)const - { return compare(that) > 0; } - bool operator<=(const sub_match& that)const - { return compare(that) <= 0; } - bool operator>=(const sub_match& that)const - { return compare(that) >= 0; } - -#ifdef BOOST_REGEX_MATCH_EXTRA - typedef std::vector > capture_sequence_type; - - const capture_sequence_type& captures()const - { - if(!m_captures) - m_captures.reset(new capture_sequence_type()); - return *m_captures; - } - // - // Private implementation API: DO NOT USE! - // - capture_sequence_type& get_captures()const - { - if(!m_captures) - m_captures.reset(new capture_sequence_type()); - return *m_captures; - } - -private: - mutable cutl_details_boost::scoped_ptr m_captures; -public: - -#endif - sub_match(const sub_match& that, bool -#ifdef BOOST_REGEX_MATCH_EXTRA - deep_copy -#endif - = true - ) - : std::pair(that), - matched(that.matched) - { -#ifdef BOOST_REGEX_MATCH_EXTRA - if(that.m_captures) - if(deep_copy) - m_captures.reset(new capture_sequence_type(*(that.m_captures))); -#endif - } - sub_match& operator=(const sub_match& that) - { - this->first = that.first; - this->second = that.second; - matched = that.matched; -#ifdef BOOST_REGEX_MATCH_EXTRA - if(that.m_captures) - get_captures() = *(that.m_captures); -#endif - return *this; - } - - -#ifdef BOOST_OLD_REGEX_H - // - // the following are deprecated, do not use!! - // - operator int()const; - operator unsigned int()const; - operator short()const - { - return (short)(int)(*this); - } - operator unsigned short()const - { - return (unsigned short)(unsigned int)(*this); - } -#endif -}; - -typedef sub_match csub_match; -typedef sub_match ssub_match; -#ifndef BOOST_NO_WREGEX -typedef sub_match wcsub_match; -typedef sub_match wssub_match; -#endif - -// comparison to std::basic_string<> part 1: -template -inline bool operator == (const std::basic_string::value_type, traits, Allocator>& s, - const sub_match& m) -{ return s.compare(m.str()) == 0; } -template -inline bool operator != (const std::basic_string::value_type, traits, Allocator>& s, - const sub_match& m) -{ return s.compare(m.str()) != 0; } -template -inline bool operator < (const std::basic_string::value_type, traits, Allocator>& s, - const sub_match& m) -{ return s.compare(m.str()) < 0; } -template -inline bool operator <= (const std::basic_string::value_type, traits, Allocator>& s, - const sub_match& m) -{ return s.compare(m.str()) <= 0; } -template -inline bool operator >= (const std::basic_string::value_type, traits, Allocator>& s, - const sub_match& m) -{ return s.compare(m.str()) >= 0; } -template -inline bool operator > (const std::basic_string::value_type, traits, Allocator>& s, - const sub_match& m) -{ return s.compare(m.str()) > 0; } -// comparison to std::basic_string<> part 2: -template -inline bool operator == (const sub_match& m, - const std::basic_string::value_type, traits, Allocator>& s) -{ return m.str().compare(s) == 0; } -template -inline bool operator != (const sub_match& m, - const std::basic_string::value_type, traits, Allocator>& s) -{ return m.str().compare(s) != 0; } -template -inline bool operator < (const sub_match& m, - const std::basic_string::value_type, traits, Allocator>& s) -{ return m.str().compare(s) < 0; } -template -inline bool operator > (const sub_match& m, - const std::basic_string::value_type, traits, Allocator>& s) -{ return m.str().compare(s) > 0; } -template -inline bool operator <= (const sub_match& m, - const std::basic_string::value_type, traits, Allocator>& s) -{ return m.str().compare(s) <= 0; } -template -inline bool operator >= (const sub_match& m, - const std::basic_string::value_type, traits, Allocator>& s) -{ return m.str().compare(s) >= 0; } -// comparison to const charT* part 1: -template -inline bool operator == (const sub_match& m, - typename re_detail::regex_iterator_traits::value_type const* s) -{ return m.str().compare(s) == 0; } -template -inline bool operator != (const sub_match& m, - typename re_detail::regex_iterator_traits::value_type const* s) -{ return m.str().compare(s) != 0; } -template -inline bool operator > (const sub_match& m, - typename re_detail::regex_iterator_traits::value_type const* s) -{ return m.str().compare(s) > 0; } -template -inline bool operator < (const sub_match& m, - typename re_detail::regex_iterator_traits::value_type const* s) -{ return m.str().compare(s) < 0; } -template -inline bool operator >= (const sub_match& m, - typename re_detail::regex_iterator_traits::value_type const* s) -{ return m.str().compare(s) >= 0; } -template -inline bool operator <= (const sub_match& m, - typename re_detail::regex_iterator_traits::value_type const* s) -{ return m.str().compare(s) <= 0; } -// comparison to const charT* part 2: -template -inline bool operator == (typename re_detail::regex_iterator_traits::value_type const* s, - const sub_match& m) -{ return m.str().compare(s) == 0; } -template -inline bool operator != (typename re_detail::regex_iterator_traits::value_type const* s, - const sub_match& m) -{ return m.str().compare(s) != 0; } -template -inline bool operator < (typename re_detail::regex_iterator_traits::value_type const* s, - const sub_match& m) -{ return m.str().compare(s) > 0; } -template -inline bool operator > (typename re_detail::regex_iterator_traits::value_type const* s, - const sub_match& m) -{ return m.str().compare(s) < 0; } -template -inline bool operator <= (typename re_detail::regex_iterator_traits::value_type const* s, - const sub_match& m) -{ return m.str().compare(s) >= 0; } -template -inline bool operator >= (typename re_detail::regex_iterator_traits::value_type const* s, - const sub_match& m) -{ return m.str().compare(s) <= 0; } - -// comparison to const charT& part 1: -template -inline bool operator == (const sub_match& m, - typename re_detail::regex_iterator_traits::value_type const& s) -{ return m.str().compare(0, m.length(), &s, 1) == 0; } -template -inline bool operator != (const sub_match& m, - typename re_detail::regex_iterator_traits::value_type const& s) -{ return m.str().compare(0, m.length(), &s, 1) != 0; } -template -inline bool operator > (const sub_match& m, - typename re_detail::regex_iterator_traits::value_type const& s) -{ return m.str().compare(0, m.length(), &s, 1) > 0; } -template -inline bool operator < (const sub_match& m, - typename re_detail::regex_iterator_traits::value_type const& s) -{ return m.str().compare(0, m.length(), &s, 1) < 0; } -template -inline bool operator >= (const sub_match& m, - typename re_detail::regex_iterator_traits::value_type const& s) -{ return m.str().compare(0, m.length(), &s, 1) >= 0; } -template -inline bool operator <= (const sub_match& m, - typename re_detail::regex_iterator_traits::value_type const& s) -{ return m.str().compare(0, m.length(), &s, 1) <= 0; } -// comparison to const charT* part 2: -template -inline bool operator == (typename re_detail::regex_iterator_traits::value_type const& s, - const sub_match& m) -{ return m.str().compare(0, m.length(), &s, 1) == 0; } -template -inline bool operator != (typename re_detail::regex_iterator_traits::value_type const& s, - const sub_match& m) -{ return m.str().compare(0, m.length(), &s, 1) != 0; } -template -inline bool operator < (typename re_detail::regex_iterator_traits::value_type const& s, - const sub_match& m) -{ return m.str().compare(0, m.length(), &s, 1) > 0; } -template -inline bool operator > (typename re_detail::regex_iterator_traits::value_type const& s, - const sub_match& m) -{ return m.str().compare(0, m.length(), &s, 1) < 0; } -template -inline bool operator <= (typename re_detail::regex_iterator_traits::value_type const& s, - const sub_match& m) -{ return m.str().compare(0, m.length(), &s, 1) >= 0; } -template -inline bool operator >= (typename re_detail::regex_iterator_traits::value_type const& s, - const sub_match& m) -{ return m.str().compare(0, m.length(), &s, 1) <= 0; } - -// addition operators: -template -inline std::basic_string::value_type, traits, Allocator> -operator + (const std::basic_string::value_type, traits, Allocator>& s, - const sub_match& m) -{ - std::basic_string::value_type, traits, Allocator> result; - result.reserve(s.size() + m.length() + 1); - return result.append(s).append(m.first, m.second); -} -template -inline std::basic_string::value_type, traits, Allocator> -operator + (const sub_match& m, - const std::basic_string::value_type, traits, Allocator>& s) -{ - std::basic_string::value_type, traits, Allocator> result; - result.reserve(s.size() + m.length() + 1); - return result.append(m.first, m.second).append(s); -} -#if !(defined(__GNUC__) && defined(BOOST_NO_STD_LOCALE)) -template -inline std::basic_string::value_type> -operator + (typename re_detail::regex_iterator_traits::value_type const* s, - const sub_match& m) -{ - std::basic_string::value_type> result; - result.reserve(std::char_traits::value_type>::length(s) + m.length() + 1); - return result.append(s).append(m.first, m.second); -} -template -inline std::basic_string::value_type> -operator + (const sub_match& m, - typename re_detail::regex_iterator_traits::value_type const * s) -{ - std::basic_string::value_type> result; - result.reserve(std::char_traits::value_type>::length(s) + m.length() + 1); - return result.append(m.first, m.second).append(s); -} -#else -// worwaround versions: -template -inline std::basic_string::value_type> -operator + (typename re_detail::regex_iterator_traits::value_type const* s, - const sub_match& m) -{ - return s + m.str(); -} -template -inline std::basic_string::value_type> -operator + (const sub_match& m, - typename re_detail::regex_iterator_traits::value_type const * s) -{ - return m.str() + s; -} -#endif -template -inline std::basic_string::value_type> -operator + (typename re_detail::regex_iterator_traits::value_type const& s, - const sub_match& m) -{ - std::basic_string::value_type> result; - result.reserve(m.length() + 2); - return result.append(1, s).append(m.first, m.second); -} -template -inline std::basic_string::value_type> -operator + (const sub_match& m, - typename re_detail::regex_iterator_traits::value_type const& s) -{ - std::basic_string::value_type> result; - result.reserve(m.length() + 2); - return result.append(m.first, m.second).append(1, s); -} -template -inline std::basic_string::value_type> -operator + (const sub_match& m1, - const sub_match& m2) -{ - std::basic_string::value_type> result; - result.reserve(m1.length() + m2.length() + 1); - return result.append(m1.first, m1.second).append(m2.first, m2.second); -} -#ifndef BOOST_NO_STD_LOCALE -template -std::basic_ostream& - operator << (std::basic_ostream& os, - const sub_match& s) -{ - return (os << s.str()); -} -#else -template -std::ostream& operator << (std::ostream& os, - const sub_match& s) -{ - return (os << s.str()); -} -#endif - -#ifdef BOOST_OLD_REGEX_H -namespace re_detail{ -template -int do_toi(BidiIterator i, BidiIterator j, char c, int radix) -{ - std::string s(i, j); - char* p; - int result = std::strtol(s.c_str(), &p, radix); - if(*p)raise_regex_exception("Bad sub-expression"); - return result; -} - -// -// helper: -template -int do_toi(I& i, I j, charT c) -{ - int result = 0; - while((i != j) && (isdigit(*i))) - { - result = result*10 + (*i - '0'); - ++i; - } - return result; -} -} - - -template -sub_match::operator int()const -{ - BidiIterator i = first; - BidiIterator j = second; - if(i == j)raise_regex_exception("Bad sub-expression"); - int neg = 1; - if((i != j) && (*i == '-')) - { - neg = -1; - ++i; - } - neg *= re_detail::do_toi(i, j, *i); - if(i != j)raise_regex_exception("Bad sub-expression"); - return neg; -} -template -sub_match::operator unsigned int()const -{ - BidiIterator i = first; - BidiIterator j = second; - if(i == j) - raise_regex_exception("Bad sub-expression"); - return re_detail::do_toi(i, j, *first); -} -#endif - -} // namespace cutl_details_boost - -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable: 4103) -#endif -#ifdef BOOST_HAS_ABI_HEADERS -# include BOOST_ABI_SUFFIX -#endif -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif - -#endif - diff --git a/cutl/details/boost/regex/v4/syntax_type.hpp b/cutl/details/boost/regex/v4/syntax_type.hpp deleted file mode 100644 index 4794c32..0000000 --- a/cutl/details/boost/regex/v4/syntax_type.hpp +++ /dev/null @@ -1,105 +0,0 @@ -/* - * - * Copyright (c) 2003 - * John Maddock - * - * Use, modification and distribution are subject to the - * Boost Software License, Version 1.0. (See accompanying file - * LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt) - * - */ - - /* - * LOCATION: see http://www.boost.org for most recent version. - * FILE syntax_type.hpp - * VERSION see - * DESCRIPTION: Declares regular expression synatx type enumerator. - */ - -#ifndef BOOST_REGEX_SYNTAX_TYPE_HPP -#define BOOST_REGEX_SYNTAX_TYPE_HPP - -namespace cutl_details_boost{ -namespace regex_constants{ - -typedef unsigned char syntax_type; - -// -// values chosen are binary compatible with previous version: -// -static const syntax_type syntax_char = 0; -static const syntax_type syntax_open_mark = 1; -static const syntax_type syntax_close_mark = 2; -static const syntax_type syntax_dollar = 3; -static const syntax_type syntax_caret = 4; -static const syntax_type syntax_dot = 5; -static const syntax_type syntax_star = 6; -static const syntax_type syntax_plus = 7; -static const syntax_type syntax_question = 8; -static const syntax_type syntax_open_set = 9; -static const syntax_type syntax_close_set = 10; -static const syntax_type syntax_or = 11; -static const syntax_type syntax_escape = 12; -static const syntax_type syntax_dash = 14; -static const syntax_type syntax_open_brace = 15; -static const syntax_type syntax_close_brace = 16; -static const syntax_type syntax_digit = 17; -static const syntax_type syntax_comma = 27; -static const syntax_type syntax_equal = 37; -static const syntax_type syntax_colon = 36; -static const syntax_type syntax_not = 53; - -// extensions: - -static const syntax_type syntax_hash = 13; -static const syntax_type syntax_newline = 26; - -// escapes: - -typedef syntax_type escape_syntax_type; - -static const escape_syntax_type escape_type_word_assert = 18; -static const escape_syntax_type escape_type_not_word_assert = 19; -static const escape_syntax_type escape_type_control_f = 29; -static const escape_syntax_type escape_type_control_n = 30; -static const escape_syntax_type escape_type_control_r = 31; -static const escape_syntax_type escape_type_control_t = 32; -static const escape_syntax_type escape_type_control_v = 33; -static const escape_syntax_type escape_type_ascii_control = 35; -static const escape_syntax_type escape_type_hex = 34; -static const escape_syntax_type escape_type_unicode = 0; // not used -static const escape_syntax_type escape_type_identity = 0; // not used -static const escape_syntax_type escape_type_backref = syntax_digit; -static const escape_syntax_type escape_type_decimal = syntax_digit; // not used -static const escape_syntax_type escape_type_class = 22; -static const escape_syntax_type escape_type_not_class = 23; - -// extensions: - -static const escape_syntax_type escape_type_left_word = 20; -static const escape_syntax_type escape_type_right_word = 21; -static const escape_syntax_type escape_type_start_buffer = 24; // for \` -static const escape_syntax_type escape_type_end_buffer = 25; // for \' -static const escape_syntax_type escape_type_control_a = 28; // for \a -static const escape_syntax_type escape_type_e = 38; // for \e -static const escape_syntax_type escape_type_E = 47; // for \Q\E -static const escape_syntax_type escape_type_Q = 48; // for \Q\E -static const escape_syntax_type escape_type_X = 49; // for \X -static const escape_syntax_type escape_type_C = 50; // for \C -static const escape_syntax_type escape_type_Z = 51; // for \Z -static const escape_syntax_type escape_type_G = 52; // for \G - -static const escape_syntax_type escape_type_property = 54; // for \p -static const escape_syntax_type escape_type_not_property = 55; // for \P -static const escape_syntax_type escape_type_named_char = 56; // for \N -static const escape_syntax_type escape_type_extended_backref = 57; // for \g -static const escape_syntax_type escape_type_reset_start_mark = 58; // for \K -static const escape_syntax_type escape_type_line_ending = 59; // for \R - -static const escape_syntax_type syntax_max = 60; - -} -} - - -#endif diff --git a/cutl/details/boost/regex/v4/u32regex_iterator.hpp b/cutl/details/boost/regex/v4/u32regex_iterator.hpp deleted file mode 100644 index 4a3068b..0000000 --- a/cutl/details/boost/regex/v4/u32regex_iterator.hpp +++ /dev/null @@ -1,193 +0,0 @@ -/* - * - * Copyright (c) 2003 - * John Maddock - * - * Use, modification and distribution are subject to the - * Boost Software License, Version 1.0. (See accompanying file - * LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt) - * - */ - - /* - * LOCATION: see http://www.boost.org for most recent version. - * FILE u32regex_iterator.hpp - * VERSION see - * DESCRIPTION: Provides u32regex_iterator implementation. - */ - -#ifndef BOOST_REGEX_V4_U32REGEX_ITERATOR_HPP -#define BOOST_REGEX_V4_U32REGEX_ITERATOR_HPP - -namespace cutl_details_boost{ - -#ifdef BOOST_HAS_ABI_HEADERS -# include BOOST_ABI_PREFIX -#endif - -template -class u32regex_iterator_implementation -{ - typedef u32regex regex_type; - - match_results what; // current match - BidirectionalIterator base; // start of sequence - BidirectionalIterator end; // end of sequence - const regex_type re; // the expression - match_flag_type flags; // flags for matching - -public: - u32regex_iterator_implementation(const regex_type* p, BidirectionalIterator last, match_flag_type f) - : base(), end(last), re(*p), flags(f){} - bool init(BidirectionalIterator first) - { - base = first; - return u32regex_search(first, end, what, re, flags, base); - } - bool compare(const u32regex_iterator_implementation& that) - { - if(this == &that) return true; - return (&re.get_data() == &that.re.get_data()) && (end == that.end) && (flags == that.flags) && (what[0].first == that.what[0].first) && (what[0].second == that.what[0].second); - } - const match_results& get() - { return what; } - bool next() - { - //if(what.prefix().first != what[0].second) - // flags |= match_prev_avail; - BidirectionalIterator next_start = what[0].second; - match_flag_type f(flags); - if(!what.length()) - f |= regex_constants::match_not_initial_null; - //if(base != next_start) - // f |= regex_constants::match_not_bob; - bool result = u32regex_search(next_start, end, what, re, f, base); - if(result) - what.set_base(base); - return result; - } -private: - u32regex_iterator_implementation& operator=(const u32regex_iterator_implementation&); -}; - -template -class u32regex_iterator -#ifndef BOOST_NO_STD_ITERATOR - : public std::iterator< - std::forward_iterator_tag, - match_results, - typename re_detail::regex_iterator_traits::difference_type, - const match_results*, - const match_results& > -#endif -{ -private: - typedef u32regex_iterator_implementation impl; - typedef shared_ptr pimpl; -public: - typedef u32regex regex_type; - typedef match_results value_type; - typedef typename re_detail::regex_iterator_traits::difference_type - difference_type; - typedef const value_type* pointer; - typedef const value_type& reference; - typedef std::forward_iterator_tag iterator_category; - - u32regex_iterator(){} - u32regex_iterator(BidirectionalIterator a, BidirectionalIterator b, - const regex_type& re, - match_flag_type m = match_default) - : pdata(new impl(&re, b, m)) - { - if(!pdata->init(a)) - { - pdata.reset(); - } - } - u32regex_iterator(const u32regex_iterator& that) - : pdata(that.pdata) {} - u32regex_iterator& operator=(const u32regex_iterator& that) - { - pdata = that.pdata; - return *this; - } - bool operator==(const u32regex_iterator& that)const - { - if((pdata.get() == 0) || (that.pdata.get() == 0)) - return pdata.get() == that.pdata.get(); - return pdata->compare(*(that.pdata.get())); - } - bool operator!=(const u32regex_iterator& that)const - { return !(*this == that); } - const value_type& operator*()const - { return pdata->get(); } - const value_type* operator->()const - { return &(pdata->get()); } - u32regex_iterator& operator++() - { - cow(); - if(0 == pdata->next()) - { - pdata.reset(); - } - return *this; - } - u32regex_iterator operator++(int) - { - u32regex_iterator result(*this); - ++(*this); - return result; - } -private: - - pimpl pdata; - - void cow() - { - // copy-on-write - if(pdata.get() && !pdata.unique()) - { - pdata.reset(new impl(*(pdata.get()))); - } - } -}; - -typedef u32regex_iterator utf8regex_iterator; -typedef u32regex_iterator utf16regex_iterator; -typedef u32regex_iterator utf32regex_iterator; - -inline u32regex_iterator make_u32regex_iterator(const char* p, const u32regex& e, regex_constants::match_flag_type m = regex_constants::match_default) -{ - return u32regex_iterator(p, p+std::strlen(p), e, m); -} -#ifndef BOOST_NO_WREGEX -inline u32regex_iterator make_u32regex_iterator(const wchar_t* p, const u32regex& e, regex_constants::match_flag_type m = regex_constants::match_default) -{ - return u32regex_iterator(p, p+std::wcslen(p), e, m); -} -#endif -#if !defined(U_WCHAR_IS_UTF16) && (U_SIZEOF_WCHAR_T != 2) -inline u32regex_iterator make_u32regex_iterator(const UChar* p, const u32regex& e, regex_constants::match_flag_type m = regex_constants::match_default) -{ - return u32regex_iterator(p, p+u_strlen(p), e, m); -} -#endif -template -inline u32regex_iterator::const_iterator> make_u32regex_iterator(const std::basic_string& p, const u32regex& e, regex_constants::match_flag_type m = regex_constants::match_default) -{ - typedef typename std::basic_string::const_iterator iter_type; - return u32regex_iterator(p.begin(), p.end(), e, m); -} -inline u32regex_iterator make_u32regex_iterator(const U_NAMESPACE_QUALIFIER UnicodeString& s, const u32regex& e, regex_constants::match_flag_type m = regex_constants::match_default) -{ - return u32regex_iterator(s.getBuffer(), s.getBuffer() + s.length(), e, m); -} - -#ifdef BOOST_HAS_ABI_HEADERS -# include BOOST_ABI_SUFFIX -#endif - -} // namespace cutl_details_boost - -#endif // BOOST_REGEX_V4_REGEX_ITERATOR_HPP - diff --git a/cutl/details/boost/regex/v4/u32regex_token_iterator.hpp b/cutl/details/boost/regex/v4/u32regex_token_iterator.hpp deleted file mode 100644 index 0c7a533..0000000 --- a/cutl/details/boost/regex/v4/u32regex_token_iterator.hpp +++ /dev/null @@ -1,377 +0,0 @@ -/* - * - * Copyright (c) 2003 - * John Maddock - * - * Use, modification and distribution are subject to the - * Boost Software License, Version 1.0. (See accompanying file - * LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt) - * - */ - - /* - * LOCATION: see http://www.boost.org for most recent version. - * FILE u32regex_token_iterator.hpp - * VERSION see - * DESCRIPTION: Provides u32regex_token_iterator implementation. - */ - -#ifndef BOOST_REGEX_V4_U32REGEX_TOKEN_ITERATOR_HPP -#define BOOST_REGEX_V4_U32REGEX_TOKEN_ITERATOR_HPP - -#if (BOOST_WORKAROUND(__BORLANDC__, >= 0x560) && BOOST_WORKAROUND(__BORLANDC__, BOOST_TESTED_AT(0x570)))\ - || BOOST_WORKAROUND(BOOST_MSVC, < 1300) \ - || BOOST_WORKAROUND(__MWERKS__, BOOST_TESTED_AT(0x3003)) -// -// Borland C++ Builder 6, and Visual C++ 6, -// can't cope with the array template constructor -// so we have a template member that will accept any type as -// argument, and then assert that is really is an array: -// -#include -#include -#endif - -namespace cutl_details_boost{ - -#ifdef BOOST_HAS_ABI_HEADERS -# include BOOST_ABI_PREFIX -#endif -#if BOOST_WORKAROUND(BOOST_MSVC, > 1300) -# pragma warning(push) -# pragma warning(disable:4700) -#endif - -template -class u32regex_token_iterator_implementation -{ - typedef u32regex regex_type; - typedef sub_match value_type; - - match_results what; // current match - BidirectionalIterator end; // end of search area - BidirectionalIterator base; // start of search area - const regex_type re; // the expression - match_flag_type flags; // match flags - value_type result; // the current string result - int N; // the current sub-expression being enumerated - std::vector subs; // the sub-expressions to enumerate - -public: - u32regex_token_iterator_implementation(const regex_type* p, BidirectionalIterator last, int sub, match_flag_type f) - : end(last), re(*p), flags(f){ subs.push_back(sub); } - u32regex_token_iterator_implementation(const regex_type* p, BidirectionalIterator last, const std::vector& v, match_flag_type f) - : end(last), re(*p), flags(f), subs(v){} -#if BOOST_WORKAROUND(BOOST_MSVC, < 1300) - // can't reliably get this to work.... -#elif (BOOST_WORKAROUND(__BORLANDC__, >= 0x560) && BOOST_WORKAROUND(__BORLANDC__, BOOST_TESTED_AT(0x570)))\ - || BOOST_WORKAROUND(BOOST_MSVC, < 1300) \ - || BOOST_WORKAROUND(__MWERKS__, BOOST_TESTED_AT(0x3003)) \ - || BOOST_WORKAROUND(__HP_aCC, < 60700) - template - u32regex_token_iterator_implementation(const regex_type* p, BidirectionalIterator last, const T& submatches, match_flag_type f) - : end(last), re(*p), flags(f) - { - // assert that T really is an array: - BOOST_STATIC_ASSERT(::cutl_details_boost::is_array::value); - const std::size_t array_size = sizeof(T) / sizeof(submatches[0]); - for(std::size_t i = 0; i < array_size; ++i) - { - subs.push_back(submatches[i]); - } - } -#else - template - u32regex_token_iterator_implementation(const regex_type* p, BidirectionalIterator last, const int (&submatches)[CN], match_flag_type f) - : end(last), re(*p), flags(f) - { - for(std::size_t i = 0; i < CN; ++i) - { - subs.push_back(submatches[i]); - } - } -#endif - - bool init(BidirectionalIterator first) - { - base = first; - N = 0; - if(u32regex_search(first, end, what, re, flags, base) == true) - { - N = 0; - result = ((subs[N] == -1) ? what.prefix() : what[(int)subs[N]]); - return true; - } - else if((subs[N] == -1) && (first != end)) - { - result.first = first; - result.second = end; - result.matched = (first != end); - N = -1; - return true; - } - return false; - } - bool compare(const u32regex_token_iterator_implementation& that) - { - if(this == &that) return true; - return (&re.get_data() == &that.re.get_data()) - && (end == that.end) - && (flags == that.flags) - && (N == that.N) - && (what[0].first == that.what[0].first) - && (what[0].second == that.what[0].second); - } - const value_type& get() - { return result; } - bool next() - { - if(N == -1) - return false; - if(N+1 < (int)subs.size()) - { - ++N; - result =((subs[N] == -1) ? what.prefix() : what[subs[N]]); - return true; - } - //if(what.prefix().first != what[0].second) - // flags |= match_prev_avail | regex_constants::match_not_bob; - BidirectionalIterator last_end(what[0].second); - if(u32regex_search(last_end, end, what, re, ((what[0].first == what[0].second) ? flags | regex_constants::match_not_initial_null : flags), base)) - { - N =0; - result =((subs[N] == -1) ? what.prefix() : what[subs[N]]); - return true; - } - else if((last_end != end) && (subs[0] == -1)) - { - N =-1; - result.first = last_end; - result.second = end; - result.matched = (last_end != end); - return true; - } - return false; - } -private: - u32regex_token_iterator_implementation& operator=(const u32regex_token_iterator_implementation&); -}; - -template -class u32regex_token_iterator -#ifndef BOOST_NO_STD_ITERATOR - : public std::iterator< - std::forward_iterator_tag, - sub_match, - typename re_detail::regex_iterator_traits::difference_type, - const sub_match*, - const sub_match& > -#endif -{ -private: - typedef u32regex_token_iterator_implementation impl; - typedef shared_ptr pimpl; -public: - typedef u32regex regex_type; - typedef sub_match value_type; - typedef typename re_detail::regex_iterator_traits::difference_type - difference_type; - typedef const value_type* pointer; - typedef const value_type& reference; - typedef std::forward_iterator_tag iterator_category; - - u32regex_token_iterator(){} - u32regex_token_iterator(BidirectionalIterator a, BidirectionalIterator b, const regex_type& re, - int submatch = 0, match_flag_type m = match_default) - : pdata(new impl(&re, b, submatch, m)) - { - if(!pdata->init(a)) - pdata.reset(); - } - u32regex_token_iterator(BidirectionalIterator a, BidirectionalIterator b, const regex_type& re, - const std::vector& submatches, match_flag_type m = match_default) - : pdata(new impl(&re, b, submatches, m)) - { - if(!pdata->init(a)) - pdata.reset(); - } -#if BOOST_WORKAROUND(BOOST_MSVC, < 1300) - // can't reliably get this to work.... -#elif (BOOST_WORKAROUND(__BORLANDC__, >= 0x560) && BOOST_WORKAROUND(__BORLANDC__, BOOST_TESTED_AT(0x570)))\ - || BOOST_WORKAROUND(BOOST_MSVC, < 1300) \ - || BOOST_WORKAROUND(__MWERKS__, BOOST_TESTED_AT(0x3003)) \ - || BOOST_WORKAROUND(__HP_aCC, < 60700) - template - u32regex_token_iterator(BidirectionalIterator a, BidirectionalIterator b, const regex_type& re, - const T& submatches, match_flag_type m = match_default) - : pdata(new impl(&re, b, submatches, m)) - { - if(!pdata->init(a)) - pdata.reset(); - } -#else - template - u32regex_token_iterator(BidirectionalIterator a, BidirectionalIterator b, const regex_type& re, - const int (&submatches)[N], match_flag_type m = match_default) - : pdata(new impl(&re, b, submatches, m)) - { - if(!pdata->init(a)) - pdata.reset(); - } -#endif - u32regex_token_iterator(const u32regex_token_iterator& that) - : pdata(that.pdata) {} - u32regex_token_iterator& operator=(const u32regex_token_iterator& that) - { - pdata = that.pdata; - return *this; - } - bool operator==(const u32regex_token_iterator& that)const - { - if((pdata.get() == 0) || (that.pdata.get() == 0)) - return pdata.get() == that.pdata.get(); - return pdata->compare(*(that.pdata.get())); - } - bool operator!=(const u32regex_token_iterator& that)const - { return !(*this == that); } - const value_type& operator*()const - { return pdata->get(); } - const value_type* operator->()const - { return &(pdata->get()); } - u32regex_token_iterator& operator++() - { - cow(); - if(0 == pdata->next()) - { - pdata.reset(); - } - return *this; - } - u32regex_token_iterator operator++(int) - { - u32regex_token_iterator result(*this); - ++(*this); - return result; - } -private: - - pimpl pdata; - - void cow() - { - // copy-on-write - if(pdata.get() && !pdata.unique()) - { - pdata.reset(new impl(*(pdata.get()))); - } - } -}; - -typedef u32regex_token_iterator utf8regex_token_iterator; -typedef u32regex_token_iterator utf16regex_token_iterator; -typedef u32regex_token_iterator utf32regex_token_iterator; - -// construction from an integral sub_match state_id: -inline u32regex_token_iterator make_u32regex_token_iterator(const char* p, const u32regex& e, int submatch = 0, regex_constants::match_flag_type m = regex_constants::match_default) -{ - return u32regex_token_iterator(p, p+std::strlen(p), e, submatch, m); -} -#ifndef BOOST_NO_WREGEX -inline u32regex_token_iterator make_u32regex_token_iterator(const wchar_t* p, const u32regex& e, int submatch = 0, regex_constants::match_flag_type m = regex_constants::match_default) -{ - return u32regex_token_iterator(p, p+std::wcslen(p), e, submatch, m); -} -#endif -#if !defined(U_WCHAR_IS_UTF16) && (U_SIZEOF_WCHAR_T != 2) -inline u32regex_token_iterator make_u32regex_token_iterator(const UChar* p, const u32regex& e, int submatch = 0, regex_constants::match_flag_type m = regex_constants::match_default) -{ - return u32regex_token_iterator(p, p+u_strlen(p), e, submatch, m); -} -#endif -template -inline u32regex_token_iterator::const_iterator> make_u32regex_token_iterator(const std::basic_string& p, const u32regex& e, int submatch = 0, regex_constants::match_flag_type m = regex_constants::match_default) -{ - typedef typename std::basic_string::const_iterator iter_type; - return u32regex_token_iterator(p.begin(), p.end(), e, submatch, m); -} -inline u32regex_token_iterator make_u32regex_token_iterator(const U_NAMESPACE_QUALIFIER UnicodeString& s, const u32regex& e, int submatch = 0, regex_constants::match_flag_type m = regex_constants::match_default) -{ - return u32regex_token_iterator(s.getBuffer(), s.getBuffer() + s.length(), e, submatch, m); -} - -#if !BOOST_WORKAROUND(BOOST_MSVC, < 1300) -// construction from a reference to an array: -template -inline u32regex_token_iterator make_u32regex_token_iterator(const char* p, const u32regex& e, const int (&submatch)[N], regex_constants::match_flag_type m = regex_constants::match_default) -{ - return u32regex_token_iterator(p, p+std::strlen(p), e, submatch, m); -} -#ifndef BOOST_NO_WREGEX -template -inline u32regex_token_iterator make_u32regex_token_iterator(const wchar_t* p, const u32regex& e, const int (&submatch)[N], regex_constants::match_flag_type m = regex_constants::match_default) -{ - return u32regex_token_iterator(p, p+std::wcslen(p), e, submatch, m); -} -#endif -#if !defined(U_WCHAR_IS_UTF16) && (U_SIZEOF_WCHAR_T != 2) -template -inline u32regex_token_iterator make_u32regex_token_iterator(const UChar* p, const u32regex& e, const int (&submatch)[N], regex_constants::match_flag_type m = regex_constants::match_default) -{ - return u32regex_token_iterator(p, p+u_strlen(p), e, submatch, m); -} -#endif -template -inline u32regex_token_iterator::const_iterator> make_u32regex_token_iterator(const std::basic_string& p, const u32regex& e, const int (&submatch)[N], regex_constants::match_flag_type m = regex_constants::match_default) -{ - typedef typename std::basic_string::const_iterator iter_type; - return u32regex_token_iterator(p.begin(), p.end(), e, submatch, m); -} -template -inline u32regex_token_iterator make_u32regex_token_iterator(const U_NAMESPACE_QUALIFIER UnicodeString& s, const u32regex& e, const int (&submatch)[N], regex_constants::match_flag_type m = regex_constants::match_default) -{ - return u32regex_token_iterator(s.getBuffer(), s.getBuffer() + s.length(), e, submatch, m); -} -#endif // BOOST_MSVC < 1300 - -// construction from a vector of sub_match state_id's: -inline u32regex_token_iterator make_u32regex_token_iterator(const char* p, const u32regex& e, const std::vector& submatch, regex_constants::match_flag_type m = regex_constants::match_default) -{ - return u32regex_token_iterator(p, p+std::strlen(p), e, submatch, m); -} -#ifndef BOOST_NO_WREGEX -inline u32regex_token_iterator make_u32regex_token_iterator(const wchar_t* p, const u32regex& e, const std::vector& submatch, regex_constants::match_flag_type m = regex_constants::match_default) -{ - return u32regex_token_iterator(p, p+std::wcslen(p), e, submatch, m); -} -#endif -#if !defined(U_WCHAR_IS_UTF16) && (U_SIZEOF_WCHAR_T != 2) -inline u32regex_token_iterator make_u32regex_token_iterator(const UChar* p, const u32regex& e, const std::vector& submatch, regex_constants::match_flag_type m = regex_constants::match_default) -{ - return u32regex_token_iterator(p, p+u_strlen(p), e, submatch, m); -} -#endif -template -inline u32regex_token_iterator::const_iterator> make_u32regex_token_iterator(const std::basic_string& p, const u32regex& e, const std::vector& submatch, regex_constants::match_flag_type m = regex_constants::match_default) -{ - typedef typename std::basic_string::const_iterator iter_type; - return u32regex_token_iterator(p.begin(), p.end(), e, submatch, m); -} -inline u32regex_token_iterator make_u32regex_token_iterator(const U_NAMESPACE_QUALIFIER UnicodeString& s, const u32regex& e, const std::vector& submatch, regex_constants::match_flag_type m = regex_constants::match_default) -{ - return u32regex_token_iterator(s.getBuffer(), s.getBuffer() + s.length(), e, submatch, m); -} - -#if BOOST_WORKAROUND(BOOST_MSVC, > 1300) -# pragma warning(pop) -#endif -#ifdef BOOST_HAS_ABI_HEADERS -# include BOOST_ABI_SUFFIX -#endif - -} // namespace cutl_details_boost - -#endif // BOOST_REGEX_V4_REGEX_TOKEN_ITERATOR_HPP - - - - diff --git a/cutl/details/boost/regex/v4/w32_regex_traits.hpp b/cutl/details/boost/regex/v4/w32_regex_traits.hpp deleted file mode 100644 index 454c915..0000000 --- a/cutl/details/boost/regex/v4/w32_regex_traits.hpp +++ /dev/null @@ -1,741 +0,0 @@ -/* - * - * Copyright (c) 2004 - * John Maddock - * - * Use, modification and distribution are subject to the - * Boost Software License, Version 1.0. (See accompanying file - * LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt) - * - */ - - /* - * LOCATION: see http://www.boost.org for most recent version. - * FILE w32_regex_traits.hpp - * VERSION see - * DESCRIPTION: Declares regular expression traits class w32_regex_traits. - */ - -#ifndef BOOST_W32_REGEX_TRAITS_HPP_INCLUDED -#define BOOST_W32_REGEX_TRAITS_HPP_INCLUDED - -#ifndef BOOST_RE_PAT_EXCEPT_HPP -#include -#endif -#ifndef BOOST_REGEX_TRAITS_DEFAULTS_HPP_INCLUDED -#include -#endif -#ifdef BOOST_HAS_THREADS -#include -#endif -#ifndef BOOST_REGEX_PRIMARY_TRANSFORM -#include -#endif -#ifndef BOOST_REGEX_OBJECT_CACHE_HPP -#include -#endif - -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable: 4103) -#endif -#ifdef BOOST_HAS_ABI_HEADERS -# include BOOST_ABI_PREFIX -#endif -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif - -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable:4786) -#pragma warning(disable:4800) -#endif - -namespace cutl_details_boost{ - -// -// forward declaration is needed by some compilers: -// -template -class w32_regex_traits; - -namespace re_detail{ - -// -// start by typedeffing the types we'll need: -// -typedef ::cutl_details_boost::uint32_t lcid_type; // placeholder for LCID. -typedef ::cutl_details_boost::shared_ptr cat_type; // placeholder for dll HANDLE. - -// -// then add wrappers around the actual Win32 API's (ie implementation hiding): -// -BOOST_REGEX_DECL lcid_type BOOST_REGEX_CALL w32_get_default_locale(); -BOOST_REGEX_DECL bool BOOST_REGEX_CALL w32_is_lower(char, lcid_type); -#ifndef BOOST_NO_WREGEX -BOOST_REGEX_DECL bool BOOST_REGEX_CALL w32_is_lower(wchar_t, lcid_type); -#ifdef BOOST_REGEX_HAS_OTHER_WCHAR_T -BOOST_REGEX_DECL bool BOOST_REGEX_CALL w32_is_lower(unsigned short ca, lcid_type state_id); -#endif -#endif -BOOST_REGEX_DECL bool BOOST_REGEX_CALL w32_is_upper(char, lcid_type); -#ifndef BOOST_NO_WREGEX -BOOST_REGEX_DECL bool BOOST_REGEX_CALL w32_is_upper(wchar_t, lcid_type); -#ifdef BOOST_REGEX_HAS_OTHER_WCHAR_T -BOOST_REGEX_DECL bool BOOST_REGEX_CALL w32_is_upper(unsigned short ca, lcid_type state_id); -#endif -#endif -BOOST_REGEX_DECL cat_type BOOST_REGEX_CALL w32_cat_open(const std::string& name); -BOOST_REGEX_DECL std::string BOOST_REGEX_CALL w32_cat_get(const cat_type& cat, lcid_type state_id, int i, const std::string& def); -#ifndef BOOST_NO_WREGEX -BOOST_REGEX_DECL std::wstring BOOST_REGEX_CALL w32_cat_get(const cat_type& cat, lcid_type state_id, int i, const std::wstring& def); -#ifdef BOOST_REGEX_HAS_OTHER_WCHAR_T -BOOST_REGEX_DECL std::basic_string BOOST_REGEX_CALL w32_cat_get(const cat_type& cat, lcid_type, int i, const std::basic_string& def); -#endif -#endif -BOOST_REGEX_DECL std::string BOOST_REGEX_CALL w32_transform(lcid_type state_id, const char* p1, const char* p2); -#ifndef BOOST_NO_WREGEX -BOOST_REGEX_DECL std::wstring BOOST_REGEX_CALL w32_transform(lcid_type state_id, const wchar_t* p1, const wchar_t* p2); -#ifdef BOOST_REGEX_HAS_OTHER_WCHAR_T -BOOST_REGEX_DECL std::basic_string BOOST_REGEX_CALL w32_transform(lcid_type state_id, const unsigned short* p1, const unsigned short* p2); -#endif -#endif -BOOST_REGEX_DECL char BOOST_REGEX_CALL w32_tolower(char c, lcid_type); -#ifndef BOOST_NO_WREGEX -BOOST_REGEX_DECL wchar_t BOOST_REGEX_CALL w32_tolower(wchar_t c, lcid_type); -#ifdef BOOST_REGEX_HAS_OTHER_WCHAR_T -BOOST_REGEX_DECL unsigned short BOOST_REGEX_CALL w32_tolower(unsigned short c, lcid_type state_id); -#endif -#endif -BOOST_REGEX_DECL char BOOST_REGEX_CALL w32_toupper(char c, lcid_type); -#ifndef BOOST_NO_WREGEX -BOOST_REGEX_DECL wchar_t BOOST_REGEX_CALL w32_toupper(wchar_t c, lcid_type); -#endif -BOOST_REGEX_DECL bool BOOST_REGEX_CALL w32_is(lcid_type, cutl_details_boost::uint32_t mask, char c); -#ifndef BOOST_NO_WREGEX -BOOST_REGEX_DECL bool BOOST_REGEX_CALL w32_is(lcid_type, cutl_details_boost::uint32_t mask, wchar_t c); -#ifdef BOOST_REGEX_HAS_OTHER_WCHAR_T -BOOST_REGEX_DECL bool BOOST_REGEX_CALL w32_is(lcid_type state_id, cutl_details_boost::uint32_t m, unsigned short c); -#endif -#endif -// -// class w32_regex_traits_base: -// acts as a container for locale and the facets we are using. -// -template -struct w32_regex_traits_base -{ - w32_regex_traits_base(lcid_type l) - { imbue(l); } - lcid_type imbue(lcid_type l); - - lcid_type m_locale; -}; - -template -inline lcid_type w32_regex_traits_base::imbue(lcid_type l) -{ - lcid_type result(m_locale); - m_locale = l; - return result; -} - -// -// class w32_regex_traits_char_layer: -// implements methods that require specialisation for narrow characters: -// -template -class w32_regex_traits_char_layer : public w32_regex_traits_base -{ - typedef std::basic_string string_type; - typedef std::map map_type; - typedef typename map_type::const_iterator map_iterator_type; -public: - w32_regex_traits_char_layer(const lcid_type l); - - regex_constants::syntax_type syntax_type(charT c)const - { - map_iterator_type i = m_char_map.find(c); - return ((i == m_char_map.end()) ? 0 : i->second); - } - regex_constants::escape_syntax_type escape_syntax_type(charT c) const - { - map_iterator_type i = m_char_map.find(c); - if(i == m_char_map.end()) - { - if(::cutl_details_boost::re_detail::w32_is_lower(c, this->m_locale)) return regex_constants::escape_type_class; - if(::cutl_details_boost::re_detail::w32_is_upper(c, this->m_locale)) return regex_constants::escape_type_not_class; - return 0; - } - return i->second; - } - charT tolower(charT c)const - { - return ::cutl_details_boost::re_detail::w32_tolower(c, this->m_locale); - } - bool isctype(cutl_details_boost::uint32_t mask, charT c)const - { - return ::cutl_details_boost::re_detail::w32_is(this->m_locale, mask, c); - } - -private: - string_type get_default_message(regex_constants::syntax_type); - // TODO: use a hash table when available! - map_type m_char_map; -}; - -template -w32_regex_traits_char_layer::w32_regex_traits_char_layer(::cutl_details_boost::re_detail::lcid_type l) - : w32_regex_traits_base(l) -{ - // we need to start by initialising our syntax map so we know which - // character is used for which purpose: - cat_type cat; - std::string cat_name(w32_regex_traits::get_catalog_name()); - if(cat_name.size()) - { - cat = ::cutl_details_boost::re_detail::w32_cat_open(cat_name); - if(!cat) - { - std::string m("Unable to open message catalog: "); - std::runtime_error err(m + cat_name); - cutl_details_boost::re_detail::raise_runtime_error(err); - } - } - // - // if we have a valid catalog then load our messages: - // - if(cat) - { - for(regex_constants::syntax_type i = 1; i < regex_constants::syntax_max; ++i) - { - string_type mss = ::cutl_details_boost::re_detail::w32_cat_get(cat, this->m_locale, i, get_default_message(i)); - for(typename string_type::size_type j = 0; j < mss.size(); ++j) - { - this->m_char_map[mss[j]] = i; - } - } - } - else - { - for(regex_constants::syntax_type i = 1; i < regex_constants::syntax_max; ++i) - { - const char* ptr = get_default_syntax(i); - while(ptr && *ptr) - { - this->m_char_map[static_cast(*ptr)] = i; - ++ptr; - } - } - } -} - -template -typename w32_regex_traits_char_layer::string_type - w32_regex_traits_char_layer::get_default_message(regex_constants::syntax_type i) -{ - const char* ptr = get_default_syntax(i); - string_type result; - while(ptr && *ptr) - { - result.append(1, static_cast(*ptr)); - ++ptr; - } - return result; -} - -// -// specialised version for narrow characters: -// -template <> -class BOOST_REGEX_DECL w32_regex_traits_char_layer : public w32_regex_traits_base -{ - typedef std::string string_type; -public: - w32_regex_traits_char_layer(::cutl_details_boost::re_detail::lcid_type l) - : w32_regex_traits_base(l) - { - init(); - } - - regex_constants::syntax_type syntax_type(char c)const - { - return m_char_map[static_cast(c)]; - } - regex_constants::escape_syntax_type escape_syntax_type(char c) const - { - return m_char_map[static_cast(c)]; - } - char tolower(char c)const - { - return m_lower_map[static_cast(c)]; - } - bool isctype(cutl_details_boost::uint32_t mask, char c)const - { - return m_type_map[static_cast(c)] & mask; - } - -private: - regex_constants::syntax_type m_char_map[1u << CHAR_BIT]; - char m_lower_map[1u << CHAR_BIT]; - cutl_details_boost::uint16_t m_type_map[1u << CHAR_BIT]; - void init(); -}; - -// -// class w32_regex_traits_implementation: -// provides pimpl implementation for w32_regex_traits. -// -template -class w32_regex_traits_implementation : public w32_regex_traits_char_layer -{ -public: - typedef typename w32_regex_traits::char_class_type char_class_type; - BOOST_STATIC_CONSTANT(char_class_type, mask_word = 0x0400); // must be C1_DEFINED << 1 - BOOST_STATIC_CONSTANT(char_class_type, mask_unicode = 0x0800); // must be C1_DEFINED << 2 - BOOST_STATIC_CONSTANT(char_class_type, mask_horizontal = 0x1000); // must be C1_DEFINED << 3 - BOOST_STATIC_CONSTANT(char_class_type, mask_vertical = 0x2000); // must be C1_DEFINED << 4 - BOOST_STATIC_CONSTANT(char_class_type, mask_base = 0x3ff); // all the masks used by the CT_CTYPE1 group - - typedef std::basic_string string_type; - typedef charT char_type; - w32_regex_traits_implementation(::cutl_details_boost::re_detail::lcid_type l); - std::string error_string(regex_constants::error_type n) const - { - if(!m_error_strings.empty()) - { - std::map::const_iterator p = m_error_strings.find(n); - return (p == m_error_strings.end()) ? std::string(get_default_error_string(n)) : p->second; - } - return get_default_error_string(n); - } - char_class_type lookup_classname(const charT* p1, const charT* p2) const - { - char_class_type result = lookup_classname_imp(p1, p2); - if(result == 0) - { - typedef typename string_type::size_type size_type; - string_type temp(p1, p2); - for(size_type i = 0; i < temp.size(); ++i) - temp[i] = this->tolower(temp[i]); - result = lookup_classname_imp(&*temp.begin(), &*temp.begin() + temp.size()); - } - return result; - } - string_type lookup_collatename(const charT* p1, const charT* p2) const; - string_type transform_primary(const charT* p1, const charT* p2) const; - string_type transform(const charT* p1, const charT* p2) const - { - return ::cutl_details_boost::re_detail::w32_transform(this->m_locale, p1, p2); - } -private: - std::map m_error_strings; // error messages indexed by numberic ID - std::map m_custom_class_names; // character class names - std::map m_custom_collate_names; // collating element names - unsigned m_collate_type; // the form of the collation string - charT m_collate_delim; // the collation group delimiter - // - // helpers: - // - char_class_type lookup_classname_imp(const charT* p1, const charT* p2) const; -}; - -template -typename w32_regex_traits_implementation::string_type - w32_regex_traits_implementation::transform_primary(const charT* p1, const charT* p2) const -{ - string_type result; - // - // What we do here depends upon the format of the sort key returned by - // sort key returned by this->transform: - // - switch(m_collate_type) - { - case sort_C: - case sort_unknown: - // the best we can do is translate to lower case, then get a regular sort key: - { - result.assign(p1, p2); - typedef typename string_type::size_type size_type; - for(size_type i = 0; i < result.size(); ++i) - result[i] = this->tolower(result[i]); - result = this->transform(&*result.begin(), &*result.begin() + result.size()); - break; - } - case sort_fixed: - { - // get a regular sort key, and then truncate it: - result.assign(this->transform(p1, p2)); - result.erase(this->m_collate_delim); - break; - } - case sort_delim: - // get a regular sort key, and then truncate everything after the delim: - result.assign(this->transform(p1, p2)); - std::size_t i; - for(i = 0; i < result.size(); ++i) - { - if(result[i] == m_collate_delim) - break; - } - result.erase(i); - break; - } - if(result.empty()) - result = string_type(1, charT(0)); - return result; -} - -template -typename w32_regex_traits_implementation::string_type - w32_regex_traits_implementation::lookup_collatename(const charT* p1, const charT* p2) const -{ - typedef typename std::map::const_iterator iter_type; - if(m_custom_collate_names.size()) - { - iter_type pos = m_custom_collate_names.find(string_type(p1, p2)); - if(pos != m_custom_collate_names.end()) - return pos->second; - } -#if !defined(BOOST_NO_TEMPLATED_ITERATOR_CONSTRUCTORS)\ - && !BOOST_WORKAROUND(BOOST_MSVC, < 1300)\ - && !BOOST_WORKAROUND(__BORLANDC__, <= 0x0551) - std::string name(p1, p2); -#else - std::string name; - const charT* p0 = p1; - while(p0 != p2) - name.append(1, char(*p0++)); -#endif - name = lookup_default_collate_name(name); -#if !defined(BOOST_NO_TEMPLATED_ITERATOR_CONSTRUCTORS)\ - && !BOOST_WORKAROUND(BOOST_MSVC, < 1300)\ - && !BOOST_WORKAROUND(__BORLANDC__, <= 0x0551) - if(name.size()) - return string_type(name.begin(), name.end()); -#else - if(name.size()) - { - string_type result; - typedef std::string::const_iterator iter; - iter b = name.begin(); - iter e = name.end(); - while(b != e) - result.append(1, charT(*b++)); - return result; - } -#endif - if(p2 - p1 == 1) - return string_type(1, *p1); - return string_type(); -} - -template -w32_regex_traits_implementation::w32_regex_traits_implementation(::cutl_details_boost::re_detail::lcid_type l) -: w32_regex_traits_char_layer(l) -{ - cat_type cat; - std::string cat_name(w32_regex_traits::get_catalog_name()); - if(cat_name.size()) - { - cat = ::cutl_details_boost::re_detail::w32_cat_open(cat_name); - if(!cat) - { - std::string m("Unable to open message catalog: "); - std::runtime_error err(m + cat_name); - cutl_details_boost::re_detail::raise_runtime_error(err); - } - } - // - // if we have a valid catalog then load our messages: - // - if(cat) - { - // - // Error messages: - // - for(cutl_details_boost::regex_constants::error_type i = static_cast(0); - i <= cutl_details_boost::regex_constants::error_unknown; - i = static_cast(i + 1)) - { - const char* p = get_default_error_string(i); - string_type default_message; - while(*p) - { - default_message.append(1, static_cast(*p)); - ++p; - } - string_type s = ::cutl_details_boost::re_detail::w32_cat_get(cat, this->m_locale, i+200, default_message); - std::string result; - for(std::string::size_type j = 0; j < s.size(); ++j) - { - result.append(1, static_cast(s[j])); - } - m_error_strings[i] = result; - } - // - // Custom class names: - // - static const char_class_type masks[14] = - { - 0x0104u, // C1_ALPHA | C1_DIGIT - 0x0100u, // C1_ALPHA - 0x0020u, // C1_CNTRL - 0x0004u, // C1_DIGIT - (~(0x0020u|0x0008u) & 0x01ffu) | 0x0400u, // not C1_CNTRL or C1_SPACE - 0x0002u, // C1_LOWER - (~0x0020u & 0x01ffu) | 0x0400, // not C1_CNTRL - 0x0010u, // C1_PUNCT - 0x0008u, // C1_SPACE - 0x0001u, // C1_UPPER - 0x0080u, // C1_XDIGIT - 0x0040u, // C1_BLANK - w32_regex_traits_implementation::mask_word, - w32_regex_traits_implementation::mask_unicode, - }; - static const string_type null_string; - for(unsigned int j = 0; j <= 13; ++j) - { - string_type s(::cutl_details_boost::re_detail::w32_cat_get(cat, this->m_locale, j+300, null_string)); - if(s.size()) - this->m_custom_class_names[s] = masks[j]; - } - } - // - // get the collation format used by m_pcollate: - // - m_collate_type = re_detail::find_sort_syntax(this, &m_collate_delim); -} - -template -typename w32_regex_traits_implementation::char_class_type - w32_regex_traits_implementation::lookup_classname_imp(const charT* p1, const charT* p2) const -{ - static const char_class_type masks[22] = - { - 0, - 0x0104u, // C1_ALPHA | C1_DIGIT - 0x0100u, // C1_ALPHA - 0x0040u, // C1_BLANK - 0x0020u, // C1_CNTRL - 0x0004u, // C1_DIGIT - 0x0004u, // C1_DIGIT - (~(0x0020u|0x0008u|0x0040) & 0x01ffu) | 0x0400u, // not C1_CNTRL or C1_SPACE or C1_BLANK - w32_regex_traits_implementation::mask_horizontal, - 0x0002u, // C1_LOWER - 0x0002u, // C1_LOWER - (~0x0020u & 0x01ffu) | 0x0400, // not C1_CNTRL - 0x0010u, // C1_PUNCT - 0x0008u, // C1_SPACE - 0x0008u, // C1_SPACE - 0x0001u, // C1_UPPER - w32_regex_traits_implementation::mask_unicode, - 0x0001u, // C1_UPPER - w32_regex_traits_implementation::mask_vertical, - 0x0104u | w32_regex_traits_implementation::mask_word, - 0x0104u | w32_regex_traits_implementation::mask_word, - 0x0080u, // C1_XDIGIT - }; - if(m_custom_class_names.size()) - { - typedef typename std::map, char_class_type>::const_iterator map_iter; - map_iter pos = m_custom_class_names.find(string_type(p1, p2)); - if(pos != m_custom_class_names.end()) - return pos->second; - } - std::size_t state_id = 1 + re_detail::get_default_class_id(p1, p2); - if(state_id < sizeof(masks) / sizeof(masks[0])) - return masks[state_id]; - return masks[0]; -} - - -template -cutl_details_boost::shared_ptr > create_w32_regex_traits(::cutl_details_boost::re_detail::lcid_type l BOOST_APPEND_EXPLICIT_TEMPLATE_TYPE(charT)) -{ - // TODO: create a cache for previously constructed objects. - return cutl_details_boost::object_cache< ::cutl_details_boost::re_detail::lcid_type, w32_regex_traits_implementation >::get(l, 5); -} - -} // re_detail - -template -class w32_regex_traits -{ -public: - typedef charT char_type; - typedef std::size_t size_type; - typedef std::basic_string string_type; - typedef ::cutl_details_boost::re_detail::lcid_type locale_type; - typedef cutl_details_boost::uint_least32_t char_class_type; - - struct boost_extensions_tag{}; - - w32_regex_traits() - : m_pimpl(re_detail::create_w32_regex_traits(::cutl_details_boost::re_detail::w32_get_default_locale())) - { } - static size_type length(const char_type* p) - { - return std::char_traits::length(p); - } - regex_constants::syntax_type syntax_type(charT c)const - { - return m_pimpl->syntax_type(c); - } - regex_constants::escape_syntax_type escape_syntax_type(charT c) const - { - return m_pimpl->escape_syntax_type(c); - } - charT translate(charT c) const - { - return c; - } - charT translate_nocase(charT c) const - { - return this->m_pimpl->tolower(c); - } - charT translate(charT c, bool icase) const - { - return icase ? this->m_pimpl->tolower(c) : c; - } - charT tolower(charT c) const - { - return this->m_pimpl->tolower(c); - } - charT toupper(charT c) const - { - return ::cutl_details_boost::re_detail::w32_toupper(c, this->m_pimpl->m_locale); - } - string_type transform(const charT* p1, const charT* p2) const - { - return ::cutl_details_boost::re_detail::w32_transform(this->m_pimpl->m_locale, p1, p2); - } - string_type transform_primary(const charT* p1, const charT* p2) const - { - return m_pimpl->transform_primary(p1, p2); - } - char_class_type lookup_classname(const charT* p1, const charT* p2) const - { - return m_pimpl->lookup_classname(p1, p2); - } - string_type lookup_collatename(const charT* p1, const charT* p2) const - { - return m_pimpl->lookup_collatename(p1, p2); - } - bool isctype(charT c, char_class_type f) const - { - if((f & re_detail::w32_regex_traits_implementation::mask_base) - && (this->m_pimpl->isctype(f & re_detail::w32_regex_traits_implementation::mask_base, c))) - return true; - else if((f & re_detail::w32_regex_traits_implementation::mask_unicode) && re_detail::is_extended(c)) - return true; - else if((f & re_detail::w32_regex_traits_implementation::mask_word) && (c == '_')) - return true; - else if((f & re_detail::w32_regex_traits_implementation::mask_vertical) - && (::cutl_details_boost::re_detail::is_separator(c) || (c == '\v'))) - return true; - else if((f & re_detail::w32_regex_traits_implementation::mask_horizontal) - && this->isctype(c, 0x0008u) && !this->isctype(c, re_detail::w32_regex_traits_implementation::mask_vertical)) - return true; - return false; - } - int toi(const charT*& p1, const charT* p2, int radix)const - { - return ::cutl_details_boost::re_detail::global_toi(p1, p2, radix, *this); - } - int value(charT c, int radix)const - { - int result = ::cutl_details_boost::re_detail::global_value(c); - return result < radix ? result : -1; - } - locale_type imbue(locale_type l) - { - ::cutl_details_boost::re_detail::lcid_type result(getloc()); - m_pimpl = re_detail::create_w32_regex_traits(l); - return result; - } - locale_type getloc()const - { - return m_pimpl->m_locale; - } - std::string error_string(regex_constants::error_type n) const - { - return m_pimpl->error_string(n); - } - - // - // extension: - // set the name of the message catalog in use (defaults to "boost_regex"). - // - static std::string catalog_name(const std::string& name); - static std::string get_catalog_name(); - -private: - cutl_details_boost::shared_ptr > m_pimpl; - // - // catalog name handler: - // - static std::string& get_catalog_name_inst(); - -#ifdef BOOST_HAS_THREADS - static static_mutex& get_mutex_inst(); -#endif -}; - -template -std::string w32_regex_traits::catalog_name(const std::string& name) -{ -#ifdef BOOST_HAS_THREADS - static_mutex::scoped_lock lk(get_mutex_inst()); -#endif - std::string result(get_catalog_name_inst()); - get_catalog_name_inst() = name; - return result; -} - -template -std::string& w32_regex_traits::get_catalog_name_inst() -{ - static std::string s_name; - return s_name; -} - -template -std::string w32_regex_traits::get_catalog_name() -{ -#ifdef BOOST_HAS_THREADS - static_mutex::scoped_lock lk(get_mutex_inst()); -#endif - std::string result(get_catalog_name_inst()); - return result; -} - -#ifdef BOOST_HAS_THREADS -template -static_mutex& w32_regex_traits::get_mutex_inst() -{ - static static_mutex s_mutex = BOOST_STATIC_MUTEX_INIT; - return s_mutex; -} -#endif - - -} // boost - -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif - -#ifdef BOOST_MSVC -#pragma warning(push) -#pragma warning(disable: 4103) -#endif -#ifdef BOOST_HAS_ABI_HEADERS -# include BOOST_ABI_SUFFIX -#endif -#ifdef BOOST_MSVC -#pragma warning(pop) -#endif - -#endif -- cgit v1.1