%META:TOPICINFO{author="MarkusDoering" date="1248856775" format="1.1" version="1.12"}%
*Developers:* MarkusDoering

*Availability/Location:* source code hosted by sourceforge http://digir.svn.sourceforge.net/viewvc/digir/pywrapper/

*Language:* Python

*Documentation:* http://digir.svn.sourceforge.net/viewvc/digir/pywrapper/docs/PyWrapper%20Guide.pdf

*Description:* PyWrapper for TAPIR is not actively maintained anymore and the project site pywrapper.org has been closed, migrating all source code to the DiGIR sourceforge site. The previous BioCASe version2 branch of PyWrapper is still maintained and available from the BioCASE website: http://www.biocase.org/products/provider_software/index.shtml

The third version of the PyWrapper is an interface to databases using the TAPIR protocol. The entire TAPIR specification is supported including the processing of dynamic output models. The TAPIR protocol is mainly a XML based query language that abstracts local data(base) schemas from the semantics. By utilizing one or more agreed data standards (mostly defined with the help of XML schemas within a community) a query does not need to know the local data(base) structure, as each PyWrapper installation is configured for these standards by mapping them to the underlying datasource structure.

PyWrapper comes with a graphical web based configuration tool, some testing and querying utilities and it can read alias.txt files from a ConceptNameServer? to manage concept, models and templates in a community network. The software runs on a small python application server (CherryPy) that can be run as a standalone service (ideal to give it a fast try) or behind an Apache webserver (which is recommended for production). A LSID resolver is being integrated into the server at the moment as part of the final 3.1 release. 

The connection to the database is done by different database modules that use SQL templates. This way it is easy to build new DB modules for different DB vendors. Non SQL capable databases will not be supported. Currently only MySQL and PostgreSQL are implemented, but please get in touch with us if you require a different system.