Wikipedia Dump Reader

This simple programs display the text-only wikipedia compressed dumps, currently available at http://download.wikimedia.org/backup-index.html, generally named something like pages-articles.xml.bz2.

It's fairly useable now although lots of rendering issues occurs

Features includes a Qt viewer with basic text markup, following links, ability to read directly on the .bz2 compressed file (altough some index creations step is needed on first run), tab-like list of articles with load-in-the-background by default, a simple but useful keyword search, very light source-code, optional latex rendering.

The code requires PyQt4

Older versions has been tested on Fedora Core 4 and Kubuntu with PyQt4.1 (Python 2.4, Qt 4.2), and Ubuntu Gutsy.

See included README

Note that the development tree is now hosted on launchpad. See https://launchpad.net/wikipediadumpreader/

Any comment is welcome.

Updated to 0.2.10:
- Use a new indexing scheme for the entrylist - articles load faster now
- Upgrade path for old indexing scheme
- Utf8 fixes for non-ascii pathnames
- experimental RPM package - feedback welcome at the project website : https://launchpad.net/wikipediadumpreader

(jul 09: updated the ubuntu package for Jaunty's Python2.6 compatibility)

Updated to 0.2.9:
- make it able to load Wiktionary non-uppercased words
- Ability to load a 64-bits module - Thanks to Michael Heide
- added a small UI layout - Thanks to GreenReaper
- Better corrupted files handling

Updated to 0.2.8:
- Sorry : no program changes, but a much more friendly opening dialog
Built a rough Ubuntu package, to ease installation for unexperienced users running Ubuntu Gutsy or Hardy


Updated to 0.2.7:
- minor rendering fixes
- a few more macros

Updated to 0.2.6:
- better wikisyntax parsing
- minor bugfixes

Updated to 0.2.5:
- Bugfixes and improvement in rendering.
- Moved the development tree to lp
- optional fontsize

Updated to 0.2.4:
- Optional Latex/texvc call to render math. thanks to Mathieu Beliveau

Updated to 0.2.3:
- Fixed an obvious overflow bug in the index creation code.
Rebuilding the index is necessary, sorry. To force it, delete the two *idx files before running the program, and be patient (English dumps index creation takes several dozen minutes)
- basic table and footnotes support

Updated to 0.2.1 : fix a bug when reading articles on blocks boundaries
Updated to 0.2.2 : improved wiki rendering for lists and definitions