Dejavu 1.4.0 http://projects.amor.org/dejavu/
I'm extremely pleased to announce the release of Dejavu 1.4.0, a pure-Python Object-Relational Mapping library. Dejavu allows you to create, query, and manage persistent data using your existing knowledge of Python programming.
* Data queries are expressed using pure Python; no SQL, no operator hacks, and no need to wrap code in strings. * Data may be transparently stored in PostgreSQL, MySQL, SQLite, Access, or SQL Server databases, as well as in flat files (using shelve), and caching proxies. You can create and combine custom storage systems for your own integration and performance needs. * Easy associations between Unit classes. * Full thread-safety for reliable use in web applications and other concurrent environments. * Views, sorting, cross-tabs, and other analysis tools. * Unit Collections, plus Engines and Rules to form powerful end-user query and reporting interfaces.
What's new in 1.4 -----------------
* Full LEFT, RIGHT, and INNER JOIN support. * Optimized and introspectable To-One and To-Many associations. * Arbitrary names for Unit.ID's (primary keys). * Support for multiple ID's (primary keys). * A new Schema class and other upgrade-management tools. * Default values for UnitProperties. * New logging hooks to help debug SQL and other storage issues. * Fixes to support Python 2.4 bytecode and other changes. * Inheritance support (all subclasses are recalled). * Vastly-improved test suite. * New recur module included, with a threaded Worker class. * Better support for update triggers.
Dejavu is in the Public Domain, and you may use it anywhere with no obligation whatsoever.
User documentation and a full Trac site are available at: http://projects.amor.org/dejavu/
Robert Brewer System Architect Amor Ministries firstname.lastname@example.org