diff options
author | duerst <duerst@b2dd03c8-39d4-4d8f-98ff-823fe69b080e> | 2016-02-07 13:10:20 +0000 |
---|---|---|
committer | duerst <duerst@b2dd03c8-39d4-4d8f-98ff-823fe69b080e> | 2016-02-07 13:10:20 +0000 |
commit | f78e8fcef2b0759986a00b0f360ea80f8f30ce76 (patch) | |
tree | 85636539b1cd2efbe18f905b9602ce5e9416e15b /enc | |
parent | 3ef23e087251f9d25341ebdc8b6bbfb29b1dedeb (diff) | |
download | ruby-f78e8fcef2b0759986a00b0f360ea80f8f30ce76.tar.gz |
* common.mk: Added two more precondition files for enc/unicode/casefold.h
* enc/unicode.c: Added shortening macros for enc/unicode/casefold.h
* enc/unicode/case-folding.rb: Fixed file encoding for CaseFolding.txt
to ASCII-8BIT (should fix some ci errors). Clarified usage. Created
class MapItem. Partially implemented class CaseMapping.
(with Kimihito Matsui)
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@53767 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
Diffstat (limited to 'enc')
-rw-r--r-- | enc/unicode.c | 10 | ||||
-rwxr-xr-x | enc/unicode/case-folding.rb | 35 |
2 files changed, 41 insertions, 4 deletions
diff --git a/enc/unicode.c b/enc/unicode.c index 99dc6dc0d3..b0ecc1bc50 100644 --- a/enc/unicode.c +++ b/enc/unicode.c @@ -140,8 +140,18 @@ code3_equal(const OnigCodePoint *x, const OnigCodePoint *y) return 1; } +#define UP ONIGENC_CASE_UPCASE +#define DOWN ONIGENC_CASE_DOWNCASE +#define TITLE ONIGENC_CASE_TITLECASE +#define FOLD ONIGENC_CASE_FOLD + #include "enc/unicode/casefold.h" +#undef UP +#undef DOWN +#undef TITLE +#undef FOLD + #include "enc/unicode/name2ctype.h" #define CODE_RANGES_NUM numberof(CodeRanges) diff --git a/enc/unicode/case-folding.rb b/enc/unicode/case-folding.rb index 33cbee5add..e1293a74db 100755 --- a/enc/unicode/case-folding.rb +++ b/enc/unicode/case-folding.rb @@ -1,10 +1,12 @@ #!/usr/bin/ruby -# Usage: +# Usage (for case folding only): # $ wget http://www.unicode.org/Public/UNIDATA/CaseFolding.txt # $ ruby case-folding.rb CaseFolding.txt -o casefold.h -# or: +# or (for case folding and case mapping): # $ wget http://www.unicode.org/Public/UNIDATA/CaseFolding.txt +# $ wget http://www.unicode.org/Public/UNIDATA/UnicodeData.txt +# $ wget http://www.unicode.org/Public/UNIDATA/SpecialCasing.txt # $ ruby case-folding.rb -m . -o casefold.h class CaseFolding @@ -175,12 +177,37 @@ class CaseFolding end end +class MapItem + def initialize(code, upper, lower, title) + @code = code + @upper = upper unless upper == '' + @lower = lower unless lower == '' + @title = title unless title == '' + end + + def flags + "" # preliminary implementation + end +end + class CaseMapping def initialize (mapping_directory) + @mappings = {} + IO.readlines(File.expand_path('UnicodeData.txt', mapping_directory), encoding: Encoding::ASCII_8BIT).each do |line| + next if line =~ /</ + code, _1,_2,_3,_4,_5,_6,_7,_8,_9,_10,_11, upper, lower, title = line.chomp.split ';' + unless upper and lower and title and (upper+lower+title)=='' + @mappings[code] = MapItem.new(code, upper, lower, title) + end + + end + + # IO.readlines(File.expand_path('SpecialCasing.txt', mapping_directory)) end def flags(from) - "" # preliminary implementation + to = @mappings[from] + to ? to.flags : "" end def self.load(*args) @@ -216,7 +243,7 @@ if $0 == __FILE__ warn "Either specify directory or individual file, but not both." exit end - filename = File.expand_path("CaseFolding.txt", mapping_directory) + filename = File.expand_path('CaseFolding.txt', mapping_directory) mapping_data = CaseMapping.load(mapping_directory) end filename ||= ARGV[0] || 'CaseFolding.txt' |