Age | Commit message (Collapse) | Author | Files | Lines |
|
|
|
|
|
In particular, this means that the primary key will point to the base
table, not root table and we can figure out the inheritance chain based
on tables by traversing their foreign keys.
|
|
|
|
|
|
|
|
|
|
Sections are an optimization mechanism that allows the partitioning of
data members of a persistent class into groups that can be separately
loaded and/or updated.
|
|
|
|
ODB now supports "smart" ordered containers. Such containers get extra
functions for updating and deleting individual elements. Based on this
functionality implement two change-tracking containers: odb::vector
(equivalent to std::vector) and QOdbList (equivalent to QList). New
tests: common/container/change-tracking and qt/common/container/change-
tracking.
|
|
|
|
|
|
Add new options (--multi-database, --default-database). Generate common
code to -odb.?xx files and database-specific to -odb-<db>.?xx.
|
|
|
|
|
|
Too much effort to maintain.
|
|
The include, types, and pragma tests have been moved to the common/
directory while the object test has been merged into common/lifecycle.
The transaction test will be re-implemented as common/ test as soon
as SQL statement tracing support is merged.
|
|
We now first create the so-called database model from C++ model and
then use that to generate the database schema. The new approach also
adds more general support for primary/foreign keys, including multi-
column keys. Finally, for MySQL we now generate out-of-line foreign
key definitions. Because MySQL does not support deferred constraints
checking, deferred foreign keys are written commented out, for
documentation.
|
|
A more generic name is required since it is now used in both generator
and processor (and in the future may be used validator).
|
|
|