dbkit is a library that abstracts away at least part of the pain
involved in dealing with DB-API 2 compatible database drivers.
Here’s an example:
from dbkit import connect, query
from contextlib import closing
with connect(sqlite3, 'counters.db') as ctx, closing(ctx):
for counter, value in query('SELECT counter, value FROM counters'):
print "%s: %d" % (counter, value)
dbkit is intended to be used in circumstances where it is impractical
or overkill to use an ORM such as SQLObject or SQLAlchemy, but it
would be useful to at least abstract away some of the pain involved in
dealing with the database.
- Rather than passing around database connections, statements are executed
within a database context, thus helping to decouple modules that
interface with the database from the database itself and its connection
- Database contexts contain references to the exceptions exposed by the
database driver, thus decoupling exception handling from the database
- Easier to use transaction handling.
- Easier iteration over resultsets.
- Connection pooling. In addition, any code using pooled connections has
no need to know connection pooling is in place.
- Query logging.
- Abstraction of SQL statements. The idea is to get rid of the more
annoying but necessary boilerplate code involved in dealing with
DB-API 2 drivers, not to totally abstract away SQL itself.
dbkit can be installed with either either pip (recommended):
$ pip install dbkit
Or with easy_install:
$ easy_install dbkit
The repos can be found on Github while
the documentation can be found on Read the Docs.
- Context.cursor() now always creates a transaction. The lack of this outer
transaction meant that PostgreSQL would end up with a large number of idle
transactions that had neither been committed or rolled back.
- Scrap unindent_statement().
- Derive all dbkit exceptions from Exception.
- Clean up connection pinging code.
- Add make_placeholders() for generating statement placeholders safely.
- Add to_dict() for converting resultsets to dicts mapped off of a
- Add last_row_id().
- Pools now can have custom mediators.
- Cursors are now tracked.
- Pooled connections are no longer closed prematurely.
- Row factories are now usable outside of context safely.
- execute*() now returns the number of affected rows.
- Add last_row_count and last_row_id to Context.
- Remove DummyPool and ThreadAffinePool, though the latter may be
- Stablise the behaviour of Pool when dealing with expired connections.
- Documentation version is now pegged directly to the library.
- Initial revision with a changelog.
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.