Commit graph

9 commits

Author SHA1 Message Date
wangqr
fbca222295 Add back original travis tests 2019-09-02 22:06:41 -04:00
wangqr
d6ddea0f65 Detect EBML magic number to skip encoding detection for MKV
MKV loads slow after f733297499
2019-06-16 19:16:31 -04:00
wangqr
f733297499 Rewrite encoding detection
Now feeds all data to uchardet, when uchardet is available. The file
size limit is removed.

When uchardet is not available, we look for UTF-8 BOM.
This should make loading UTF-8-BOM files faster.
Because Aegisub always save file in UTF-8-BOM, this should also
guarentee Aegisub will load large (>100MB) file saved by itself.

See Aegisub/Aegisub#110
2019-05-18 22:13:26 -04:00
Thomas Goyne
45315476bd Switch to a maintained fork of universalchardet 2016-02-09 20:29:29 -08:00
Thomas Goyne
879788fe83 Eliminate some gratuitous exceptions 2015-02-15 08:19:10 -08:00
Thomas Goyne
09e325a1c3 Clean up unused includes 2014-05-23 07:28:24 -07:00
Thomas Goyne
3d21f00356 Use read_file_mapping for charset detection 2014-03-21 08:06:41 -07:00
Thomas Goyne
ea96c6e2ad Make everything final that can be
Apparently gcc does use final for devirtualization.
2014-03-12 19:07:30 -07:00
Thomas Goyne
33a4a056a4 Move everything up a level since the root dir no longer has stuff 2014-03-11 12:14:57 -07:00
Renamed from aegisub/libaegisub/common/charset.cpp (Browse further)