The PLIB3.DBTOOLS package provides an easier to use interface to DB-API compliant databases.

Note: PLIB3.DBTOOLS works with Python 3. If you are using Python 2.7, see the PLIB.DBTOOLS package, available at

The script for PLIB3.DBTOOLS uses the setuputils helper module, which helps to automate away much of the boilerplate in Python 3 setup scripts. This module is available as a separate release at


The following classes and functions are available in the plib.dbtools namespace:

  • The DBInterface class provides a simple API for working with databases. It assumes that the underlying database engine conforms to the Python DB API. This class itself does not know about any particular database engine; a customized class should be derived from it for each particular engine.
  • The SQLite3DBInterface class customizes DBInterface to work with an SQLite version 3 database. It uses the sqlite3 module that comes with Python as its database engine. This class is in the plib.dbtools.sqlite sub-package.
  • The get_db_interface function is a convenience function that returns an instance of the appropriate DBInterface subclass for the database type passed to it.
  • The get_db_interface_class and get_db_interface_args functions factor out key portions of get_db_interface so that they can be used separately if desired. For example, some databases might require subclassing the interface class further, or customizing the interface arguments.


To install PLIB3.DBTOOLS, you can simply run:

$ python3 install

at a shell prompt from the directory into which you unzipped the source tarball (the same directory that this README file is in). This will install PLIB3 and then run each of the post-install scripts in the scripts directory.

The Zen of PLIB3

There is no single unifying purpose or theme to PLIB3, but like Python itself, it does have a 'Zen' of sorts:

  • Express everything possible in terms of built-in Python data structures.
  • Once you've expressed it that way, what the code is going to do with it should be obvious.
  • Avoid boilerplate code, and boilerplate data. Every piece of data your program needs should have one and only one source.