This is a utility script that loads a database dump from Wikimedia Commons and uses it as a starting point to request metadata and thumbnails from images on Commons. It calls upon and runs a Blockhash algorithm (see http://blockhash.io) on this data and stores in a database file.
This script requires Python 3 to run. It also requires a bzip2'ed copy of a recent Commons database dump (a file such as commonswiki-20140823-pages-articles.xml.bz2). You can optionally specify the name of the local database file, which will otherwise default to commons.db.
Use commonshasher.py -i [commons database file] to run.
Each additional -v will increase the verbosity level. Use -h to get
more detailed help.
The script is developed by Commons Machinery, and you can reach out to the development team by writing [email protected]
- There's no technical need to use a database dump, the script should just run on data from Commons directly to also capture information that comes between dumps.
- The script should output media information consistent with the Data Package Format at https://github.com/commonsmachinery/catalog/blob/master/doc/datapackage.md
Copyright 2014 Commons Machinery http://commonsmachinery.se/
Distributed under the GNU General Public License v3.0 or any later version (see the file LICENSE for details).