kotovalexarian-likes-github/ruby--ruby
1
0
Fork 0
mirror of https://github.com/ruby/ruby.git synced 2022-11-09 12:17:21 -05:00
Code Releases Activity

enc-unicode.rb: check Unicode version

Browse source 
* tool/enc-unicode.rb (data_foreach): check Unicode version in
  data files, and yield each lines.

git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@55685 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
This commit is contained in:
nobu 2016-07-14 16:32:13 +00:00
parent 230a91c133
commit 0fd7666d57
1 changed files with 25 additions and 7 deletions
Download patch file Download diff file Expand all files Collapse all files

32
tool/enc-unicode.rb
View file

@ -15,10 +15,11 @@ if ARGV[0] == "--header"
ARGV.shift ARGV.shift
end end
unless ARGV.size == 1 unless ARGV.size == 1
$stderr.puts "Usage: #{$0} data_directory" abort "Usage: #{$0} data_directory"
exit(1)
end end
$unicode_version = File.basename(ARGV[0])[/\A[.\d]+\z/]
POSIX_NAMES = %w[NEWLINE Alpha Blank Cntrl Digit Graph Lower Print Punct Space Upper XDigit Word Alnum ASCII] POSIX_NAMES = %w[NEWLINE Alpha Blank Cntrl Digit Graph Lower Print Punct Space Upper XDigit Word Alnum ASCII]
def pair_codepoints(codepoints) def pair_codepoints(codepoints)
@ -137,7 +138,7 @@ def parse_scripts(data, categories)
cps = [] cps = []
names = {} names = {}
files.each do |file| files.each do |file|
IO.foreach(get_file(file[:fn])) do |line| data_foreach(file[:fn]) do |line|
if /^# Total code points: / =~ line if /^# Total code points: / =~ line
data[current] = cps data[current] = cps
categories[current] = file[:title] categories[current] = file[:title]
@ -158,12 +159,12 @@ end
def parse_aliases(data) def parse_aliases(data)
kv = {} kv = {}
IO.foreach(get_file('PropertyAliases.txt')) do |line| data_foreach('PropertyAliases.txt') do |line|
next unless /^(\w+)\s*; (\w+)/ =~ line next unless /^(\w+)\s*; (\w+)/ =~ line
data[$1] = data[$2] data[$1] = data[$2]
kv[normalize_propname($1)] = normalize_propname($2) kv[normalize_propname($1)] = normalize_propname($2)
end end
IO.foreach(get_file('PropertyValueAliases.txt')) do |line| data_foreach('PropertyValueAliases.txt') do |line|
next unless /^(sc|gc)\s*; (\w+)\s*; (\w+)(?:\s*; (\w+))?/ =~ line next unless /^(sc|gc)\s*; (\w+)\s*; (\w+)(?:\s*; (\w+))?/ =~ line
if $1 == 'gc' if $1 == 'gc'
data[$3] = data[$2] data[$3] = data[$2]
@ -188,7 +189,7 @@ def parse_age(data)
last_constname = nil last_constname = nil
cps = [] cps = []
ages = [] ages = []
IO.foreach(get_file('DerivedAge.txt')) do |line| data_foreach('DerivedAge.txt') do |line|
if /^# Total code points: / =~ line if /^# Total code points: / =~ line
constname = constantize_agename(current) constname = constantize_agename(current)
# each version matches all previous versions # each version matches all previous versions
@ -211,7 +212,7 @@ def parse_block(data)
last_constname = nil last_constname = nil
cps = [] cps = []
blocks = [] blocks = []
IO.foreach(get_file('Blocks.txt')) do |line| data_foreach('Blocks.txt') do |line|
if /^([0-9a-fA-F]+)\.\.([0-9a-fA-F]+);\s*(.*)/ =~ line if /^([0-9a-fA-F]+)\.\.([0-9a-fA-F]+);\s*(.*)/ =~ line
cps = ($1.to_i(16)..$2.to_i(16)).to_a cps = ($1.to_i(16)..$2.to_i(16)).to_a
constname = constantize_blockname($3) constname = constantize_blockname($3)
@ -276,6 +277,23 @@ def get_file(name)
File.join(ARGV[0], name) File.join(ARGV[0], name)
end end
def data_foreach(name, &block)
fn = get_file(name)
warn "Reading #{name}"
pat = /^# #{name.sub(/\./, '-([\\d.]+)\\.')}/
File.open(fn) do |f|
line = f.gets
unless pat =~ line
raise ArgumentError, "#{name}: no Unicode version"
end
if !$unicode_version
$unicode_version = $1
elsif $unicode_version != $1
raise ArgumentError, "#{name}: Unicode version mismatch: #$1"
end
f.each(&block)
end
end
# Write Data # Write Data
class Unifdef class Unifdef