1berry/ruby
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

Support for Indic_Conjunct_Break in Unicode 15.1

Browse Source
This commit is contained in:
Mari Imaizumi 2025-02-24 11:43:23 +09:00
parent 75844889eb
commit 4f82a6f3e8
Notes: git 2025-03-18 12:18:36 +00:00 
Merged: https://github.com/ruby/ruby/pull/12798
1 changed files with 17 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

20
tool/enc-unicode.rb
View File

@ -161,14 +161,24 @@ def parse_scripts(data, categories)
names = {} names = {}
files.each do |file| files.each do |file|
data_foreach(file[:fn]) do |line| data_foreach(file[:fn]) do |line|
# Parse Unicode data files and store code points and properties.
if /^# Total (?:code points|elements): / =~ line if /^# Total (?:code points|elements): / =~ line
data[current] = cps data[current] = cps
categories[current] = file[:title] categories[current] = file[:title]
(names[file[:title]] ||= []) << current (names[file[:title]] ||= []) << current
cps = [] cps = []
elsif /^(\h+)(?:\.\.(\h+))?\s*;\s*(\w+)/ =~ line elsif /^(\h+)(?:\.\.(\h+))?\s*;\s*(\w(?:[\w\s;]*\w)?)/ =~ line
current = $3 # $1: The first hexadecimal code point or the start of a range.
# $2: The end code point of the range, if present.
# If there's no range (just a single code point), $2 is nil.
# $3: The property or other info.
# Example:
# line = "0915..0939 ; InCB; Consonant # Lo [37] DEVANAGARI LETTER KA..DEVANAGARI LETTER HA"
# $1 = "0915"
# $2 = "0939"
# $3 = "InCB; Consonant"
$2 ? cps.concat(($1.to_i(16)..$2.to_i(16)).to_a) : cps.push($1.to_i(16)) $2 ? cps.concat(($1.to_i(16)..$2.to_i(16)).to_a) : cps.push($1.to_i(16))
current = $3.gsub(/\W+/, '_')
end end
end end
end end
@ -486,7 +496,11 @@ end
output.ifdef :USE_UNICODE_PROPERTIES do output.ifdef :USE_UNICODE_PROPERTIES do
props.each do |name| props.each do |name|
i += 1 i += 1
name = normalize_propname(name) name = if name.start_with?('InCB')
name.downcase.gsub(/_/, '=')
else
normalize_propname(name)
end
name_to_index[name] = i name_to_index[name] = i
puts "%-40s %3d" % [name + ',', i] puts "%-40s %3d" % [name + ',', i]
end end