Commit graph

8 commits

Author SHA1 Message Date
Thomas Goyne
96cf5ea7ed Kill the #ifndef AGI_PRE guards
They don't actually improve compilation performance and make it more
annoying to modify what things are in the precompiled header.
2012-12-02 06:33:29 -08:00
Thomas Goyne
47c678bd63 Make UCDetect a little less insane 2012-11-13 08:46:31 -08:00
Thomas Goyne
e44c47c863 Remove $Id$ markers since git doesn't support them 2012-10-25 17:39:49 -07:00
Amar Takhar
944eaa535e Remove universalchardet headers from the precompiled header. For some reason this was causing issues with compliation on FreeBSD, rather than investigate why I removed it since it's used by exactly 1 file (charset_ucd.cpp). If it's going to cause one problem now it may cause another later that we may not notice.
Originally committed to SVN as r4919.
2010-12-08 14:32:30 +00:00
Thomas Goyne
75bd20a6dd Fix the rest of the build errors when not using precompiled headers
Originally committed to SVN as r4530.
2010-06-17 00:23:44 +00:00
Amar Takhar
60b9e5dee6 Define _X86_ so univeralchardet builds properly, this eventually needs to be fixed to use the autodetection code already in configure.in
Originally committed to SVN as r4486.
2010-06-12 07:04:46 +00:00
Amar Takhar
9d854b69f3 Don't bother storing storing a single charset into an std::string, instead insert it into the std::map and fix Single() to return the first element. This keeps things simple and also ensures that DetectAll() will always return atleast one element which wasn't being done before.
Originally committed to SVN as r4369.
2010-05-28 13:08:00 +00:00
Amar Takhar
e408fe49c9 Add support for detecting character sets using UniversalChardet -- this superceeds character_detect.cpp.
Originally committed to SVN as r4367.
2010-05-28 07:40:21 +00:00