# -*- mode: ruby; coding: utf-8 -*- # # poparser.ry - ruby version of msgfmt # # Copyright (C) 2002-2008 Masao Mutoh # Copyright (C) 2012 Kouhei Sutou # Copyright (C) 2012 Haruka Yoshihara # # You may redistribute it and/or modify it under the same # license terms as Ruby or LGPL. class GetText::PoParser token COMMENT MSGID MSGCTXT MSGID_PLURAL MSGSTR STRING PLURAL_NUM rule msgfmt : /* empty */ | msgfmt comment | msgfmt msgctxt | msgfmt message ; msgctxt : MSGCTXT string_list { @msgctxt = unescape(val[1]) + "\004" } ; message : single_message | plural_message ; single_message : MSGID string_list MSGSTR string_list { msgid_raw = val[1] msgid = unescape(msgid_raw) msgstr = unescape(val[3]) use_message_p = true if @fuzzy and not msgid.empty? use_message_p = (not ignore_fuzzy?) if report_warning? if ignore_fuzzy? $stderr.print _("Warning: fuzzy message was ignored.\n") else $stderr.print _("Warning: fuzzy message was used.\n") end $stderr.print " #{@po_file}: msgid '#{msgid_raw}'\n" end end @fuzzy = false on_message(@msgctxt + msgid, msgstr) if use_message_p result = "" } plural_message : MSGID string_list MSGID_PLURAL string_list msgstr_plural { if @fuzzy and ignore_fuzzy? if val[1] != "" if report_warning? $stderr.print _("Warning: fuzzy message was ignored.\n") $stderr.print "msgid = '#{val[1]}\n" end else on_message('', unescape(val[3])) end @fuzzy = false else on_message(@msgctxt + unescape(val[1]) + "\000" + unescape(val[3]), unescape(val[4])) end result = "" } ; msgstr_plural : msgstr_plural msgstr_plural_line { if val[0].size > 0 result = val[0] + "\000" + val[1] else result = "" end } | msgstr_plural_line ; msgstr_plural_line : MSGSTR PLURAL_NUM string_list { result = val[2] } ; comment : COMMENT { on_comment(val[0]) } #| COMMENT #; string_list : string_list STRING { result = val.delete_if{|item| item == ""}.join } | STRING { result = val[0] } ; end ---- inner if GetText.respond_to?(:bindtextdomain) include GetText GetText.bindtextdomain("gettext") else def _(message_id) message_id end private :_ end attr_writer :ignore_fuzzy, :report_warning def initialize @ignore_fuzzy = true @report_warning = true end def ignore_fuzzy? @ignore_fuzzy end def report_warning? @report_warning end def unescape(orig) ret = orig.gsub(/\\n/, "\n") ret.gsub!(/\\t/, "\t") ret.gsub!(/\\r/, "\r") ret.gsub!(/\\"/, "\"") ret end private :unescape def unescape_string(string) string.gsub(/\\\\/, "\\") end private :unescape_string def parse(str, data) @comments = [] @data = data @fuzzy = false @msgctxt = "" str.strip! @q = [] until str.empty? do case str when /\A\s+/ str = $' when /\Amsgctxt/ @q.push [:MSGCTXT, $&] str = $' when /\Amsgid_plural/ @q.push [:MSGID_PLURAL, $&] str = $' when /\Amsgid/ @q.push [:MSGID, $&] str = $' when /\Amsgstr/ @q.push [:MSGSTR, $&] str = $' when /\A\[(\d+)\]/ @q.push [:PLURAL_NUM, $1] str = $' when /\A\#~(.*)/ if report_warning? $stderr.print _("Warning: obsolete msgid exists.\n") $stderr.print " #{$&}\n" end @q.push [:COMMENT, $&] str = $' when /\A\#(.*)/ @q.push [:COMMENT, $&] str = $' when /\A\"(.*)\"/ @q.push [:STRING, unescape_string($1)] str = $' else #c = str[0,1] #@q.push [:STRING, c] str = str[1..-1] end end @q.push [false, '$end'] if $DEBUG @q.each do |a,b| puts "[#{a}, #{b}]" end end @yydebug = true if $DEBUG do_parse if @comments.size > 0 @data.set_comment(:last, @comments.join("\n")) end @data end def next_token @q.shift end def on_message(msgid, msgstr) msgstr = nil if msgstr.empty? @data[msgid] = msgstr @data.set_comment(msgid, @comments.join("\n")) @comments.clear @msgctxt = "" end def on_comment(comment) @fuzzy = true if (/fuzzy/ =~ comment) @comments << comment end def parse_file(po_file, data) args = [ po_file ] # In Ruby 1.9, we must detect proper encoding of a PO file. if String.instance_methods.include?(:encode) encoding = detect_file_encoding(po_file) args << "r:#{encoding}" end @po_file = po_file parse(File.open(*args) {|io| io.read }, data) end def detect_file_encoding(po_file) open(po_file, :encoding => 'ASCII-8BIT') do |input| input.lines.each do |line| return Encoding.find($1) if %r["Content-Type:.*\scharset=(.*)\\n"] =~ line end end Encoding.default_external end private :detect_file_encoding ---- footer