- Fix hachoir-subfile: hachoir.regex only supports Unicode?
- Remove old files? See MANIFEST.in for the list of "IGNORED files"
- Write more tests:
- use coverage to check which parsers are never tested
- write tests for hachoir-subfile
- Rename Dict.iteritems()
- wxPython doesn't support Python3 on Fedora
- convert all methods names to PEP8!!!
- test hachoir-gtk
- test hachoir-http
- convert remaining wiki pages
- MPEG audio is disabled
- PCX: PhotoRec regex: "x0a[02345]][[01][14x08x18]" (magic, version, compression, bits/pixel)
- TGA
- MPEG video, proposition: regex "x00x00x01[xB0xB3xB5xBAxBB" (from PhotoRec) at offset 0 (0xBA is the most common value)
- gzip: need to decompress flow (deflate using zlib)
- bzip2: need to decompress flow
Use http://www.gnu.org/software/gperf/ (GNU perfect hash function generator) for faster pattern recognition?