Kovid Goyal
ff2e5b3966
Avoid unnecessary calls to mark_for_codepoint
2018-02-06 11:23:39 +05:30
Kovid Goyal
80301d465b
Handle non-BMP combining characters
...
Use a level of indirection to store combining characters. This allows
combining characters to be stored using only two bytes, even if they are
after USHORT_MAX
2018-01-18 16:25:42 +05:30
Kovid Goyal
5faa649452
Drop the dependency on libunistring
2018-01-18 00:09:40 +05:30
Kovid Goyal
ed700ff830
...
2018-01-17 21:59:10 +05:30
Kovid Goyal
33ed873997
Remove unnecessary extra test for combinig characters
...
There are not combining characters with a non-zero combining class that
are not in the marks category.
2018-01-17 21:56:30 +05:30
Kovid Goyal
804c4fbe19
Recognize characters from the unicode Mark categories as combining characters, even if they do not have a combining class (i.e. are not re-ordered). Fixes #286
2018-01-15 11:24:11 +05:30
Kovid Goyal
0fcce6ec58
Remove trailing whitespace from native code files
2017-12-20 08:44:47 +05:30
Kovid Goyal
464291bbb1
Port click on URL code to C
2017-09-15 10:45:27 +05:30
Kovid Goyal
ed3427f349
Dont use the python unicodedata module as we use libunistring
...
No sense in loading two hude unicode datasets into memory
2017-09-15 10:45:27 +05:30
Kovid Goyal
1c1d0a4e91
Port mouse cursor change over hyperlinks to C
2017-09-15 10:45:25 +05:30
Kovid Goyal
9bea1001f9
Speedup unicode character property lookup
...
Use libunistring instead of building predicates from the unicode
database
2017-09-15 10:45:19 +05:30
Kovid Goyal
3989413ff9
...
2016-11-19 14:41:40 +05:30
Kovid Goyal
5dc0b9af13
Fix unicode data generation
2016-11-13 13:57:02 +05:30
Kovid Goyal
fab2213c25
More work on native streams
2016-11-13 10:24:00 +05:30
Kovid Goyal