Encapsulate sqlalchemy modelling infrastructure.
risclog.sqlalchemy encapsulates sqlalchemy modelling infrastructure. It eases the setup of (multi-)database environments and the usage of sqlalchemy models by providing often used functionality on special classes.
This package is compatible with Python version 2.7, 3.4 up to 3.6, PyPy2 and PyPy3.
|Current change log:|
This package ships with a minimum buildout config which allows to run the tests:
$ python bootstrap.py $ bin/buildout $ bin/test
risclog.sqlalchemy ships with a tox configuration that allows to run all tests against all supported python versions. You will need tox to run those tests:
$ tox py27 develop-inst-nodeps... ... py33 develop-inst-nodeps...
In order to build the Sphinx documentation, run the following command with a python, where sphinx is installed:
$ python setup.py build_sphinx
In order to upload the Sphinx documentation, run the following command with a python, where sphinx-pypi-upload is installed:
$ python setup.py upload_sphinx
Change log for risclog.sqlalchemy
- Drop support for Python 3.3.
- Reintroduce RoutingSession.using_bind as it is needed for executing queries in multi database environments. It was removed in version 2.0.
- Pin to zope.component < 4.4.1 for now as we are using a private variable from this package which has been removed lately.
- Add the fixture .fixtures.database__selenium_testing to switch SQLAlchemy session handling to into the same way the live server does it. Using this fixture leads to DetachedInstanceError exceptions when using a database object after the commit. It has to be fetched again from the database.
- Omit files from sdist which are related to buildout, testing or mercurial.
- Support Python 3.6 and PyPy3.
- Fix setup.py to no longer use absolute paths.
Add compatibility to pypy.
Also removed explicit dependency to psycopg2 as it’s a PostgreSQL specific dependency and not compatible with pypy. You will need psycopg2cffi in order to run risclog.sqlalchemy with pypy.
- Removed RoutingSession.using_bind, require SQLAlchemy >= 1.0. (#13968)
- Updated API calls to alembic, require alembic >= 0.7. (#13968)
- Added documentation. (#13952)
- Officially supporting Python 3.3 up to 3.5.
- Set maximum supported version numbers of alembic and SQLALchemy in setup.py
- Fix bug in testing mode detection so it works with sqlite, too.
- Have .model.Object.create use the **kw constructor instead of performing setattr itself.
- Made commit in db.empty optional.
- Make test setup/teardown a bit more independent from the IDatabase utility.
- Made pyramid dependent code optional, because it is an optional requirement.
- Moved the project to bitbucket.org and prepared for public release.
- Adjust test dependency for external projects, so pyramid is not required. (#1458)
- Do not truncate spatial_ref_sys when emptying entire database. (#13144)
- Do not truncate alembic_versions when emptying entire database (using db.empty(engine))
- Fix create_defaults, so it can be used with multiple engines. (#1172)
- Fixed setup.py
- Allow .db.Database.empty() to not restart sequences. (This can be necassary if the user is not allowed to call ALTER SEQUENCE.)
- Create factory for JSON renderer, so it can be customized via arguments. (#1037)
- CAUTION: Backward incompatible changes to provide multiple database
support. You need to change:
- Use the pyramid extra to use the pyramid serializers and call .serializer.patch() by yourself.
- Changes in .model:
- Use .model.declarative_base(cls) to register a class as SQLALchemy declarative_base.
- Use your own declarative_base as .model.Object has been dropped.
- Create your own ReflectedObject as .model.ReflectedObject has been dropped.
- Changes in .db.Database:
- To get an instance use .db.get_database(testing=<True|False>).
- To register a database with the utility use register_engine.
- empty now expects to get the engine as first argument and allows to cadcade via (cascade=True).
- _verify was removed, use _verify_engine now.
- setup_utility was removed, it is now done in __init__.
- To access former engine` attribute use get_engine.
- close was renamed to drop_engine.
- To create all tables for a database use create_all.
- Changes in .testing:
- setUpDB lost its first argument as it was not used.
- setUp now expects a dict or None, see its docstring.
- Provide a way to insert default values for new created models. (#1137)
- Added support for schema migrations with alembic.
- Dropped support for Python 2.6.
- Added convenience functions to create py.test database fixtures.
- Declared pyramid as test dependency so tests only need the test extra.
- Declared testing dependency on plone.testing[test] since we use plone.testing.zca which imports that test extra’s dependencies.
- Added convenience methods on .db.Database for teardown in tests.
- Updated to Buildout 2.
- Using py.test as testrunner.
- Added Python 3.3 compatibility.
- Add Database methods for setting up a database utility and emptying tables.
- Add testing layer that sets up a database utility (requires plone.testing).
- Add helper to provide transparent json encoding of sqlalchemy mapped objects, dates, datetimes and decimals. (#39)
- Add create_defaults class method on base model which can be used to create test data into the database.
- Repair db setup for reflected and unreflected objects.
- Add unreflected Object model next to the reflected Object model.