Timerix/utf8proc
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Minor fixes to work with Unicode 8.0.0 data

Browse Source
This commit is contained in:
Jiahao Chen 2015-06-20 08:03:40 -04:00
parent f7219d516e
commit d18963cc46
1 changed files with 3 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

5
data/charwidths.jl
View File

@ -27,6 +27,7 @@ end
function parsesfd(filename::String, CharWidths::Dict{Int,Int}=Dict{Int,Int}()) function parsesfd(filename::String, CharWidths::Dict{Int,Int}=Dict{Int,Int}())
state=:seekchar state=:seekchar
lineno = 0 lineno = 0
codepoint = width = nothing
for line in readlines(open(filename)) for line in readlines(open(filename))
lineno += 1 lineno += 1
if state==:seekchar #StartChar: nonmarkingreturn if state==:seekchar #StartChar: nonmarkingreturn
@ -91,7 +92,7 @@ function catcode(c)
end end
# use Base.UTF8proc module to get category codes constants, since # use Base.UTF8proc module to get category codes constants, since
# we aren't goint to change these in utf8proc. # we won't change these in utf8proc.
import Base.UTF8proc import Base.UTF8proc
for c in keys(CharWidths) for c in keys(CharWidths)
@ -116,7 +117,7 @@ for c in keys(CharWidths)
if cat==UTF8proc.UTF8PROC_CATEGORY_CO || cat==UTF8proc.UTF8PROC_CATEGORY_CN if cat==UTF8proc.UTF8PROC_CATEGORY_CO || cat==UTF8proc.UTF8PROC_CATEGORY_CN
CharWidths[c]=0 CharWidths[c]=0
end end
# for some reason, Unifont has width-2 glyphs for ASCII control chars # for some reason, Unifont has width-2 glyphs for ASCII control chars
if cat==UTF8proc.UTF8PROC_CATEGORY_CC if cat==UTF8proc.UTF8PROC_CATEGORY_CC
CharWidths[c]=0 CharWidths[c]=0