2
# -*- Mode: perl; indent-tabs-mode: nil; c-basic-offset: 4 -*-
5
# The Intltool Message Extractor
7
# Copyright (C) 2000-2001, 2003 Free Software Foundation.
9
# Intltool is free software; you can redistribute it and/or
10
# modify it under the terms of the GNU General Public License as
11
# published by the Free Software Foundation; either version 2 of the
12
# License, or (at your option) any later version.
14
# Intltool is distributed in the hope that it will be useful,
15
# but WITHOUT ANY WARRANTY; without even the implied warranty of
16
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
17
# General Public License for more details.
19
# You should have received a copy of the GNU General Public License
20
# along with this program; if not, write to the Free Software
21
# Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
23
# As a special exception to the GNU General Public License, if you
24
# distribute this file as part of a program that contains a
25
# configuration script generated by Autoconf, you may include it under
26
# the same distribution terms that you use for the rest of that program.
28
# Authors: Kenneth Christiansen <kenneth@gnu.org>
29
# Darin Adler <darin@bentspoon.com>
32
## Release information
33
my $PROGRAM = "intltool-extract";
34
my $PACKAGE = "intltool";
35
my $VERSION = "0.31.2";
42
## Scalars used by the option stuff
46
my $VERSION_ARG = "0";
54
my $gettext_type = "";
62
## Use this instead of \w for XML files to handle more possible characters.
63
my $w = "[-A-Za-z0-9._:]";
70
"type=s" => \$TYPE_ARG,
71
"local|l" => \$LOCAL_ARG,
72
"help|h" => \$HELP_ARG,
73
"version|v" => \$VERSION_ARG,
74
"update" => \$UPDATE_ARG,
75
"quiet|q" => \$QUIET_ARG,
76
"srcdir=s" => \$SRCDIR_ARG,
83
## This section will check for the different options.
85
sub split_on_argument {
93
} elsif ($LOCAL_ARG) {
97
} elsif ($UPDATE_ARG) {
101
} elsif (@ARGV > 0) {
114
$OUTFILE = "$FILE.h";
118
$OUTFILE = fileparse($FILE, ());
120
system("mkdir tmp/");
122
$OUTFILE = "./tmp/$OUTFILE.h"
126
if ($TYPE_ARG =~ /^gettext\/(.*)/) {
131
## Sub for printing release information
134
${PROGRAM} (${PACKAGE}) $VERSION
135
Copyright (C) 2000, 2003 Free Software Foundation, Inc.
136
Written by Kenneth Christiansen, 2000.
138
This is free software; see the source for copying conditions. There is NO
139
warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
144
## Sub for printing usage information
147
Usage: ${PROGRAM} [OPTION]... [FILENAME]
148
Generates a header file from an XML source file.
150
It grabs all strings between <_translatable_node> and its end tag in
151
XML files. Read manpage (man ${PROGRAM}) for more info.
153
--type=TYPE Specify the file type of FILENAME. Currently supports:
154
"gettext/glade", "gettext/ini", "gettext/keys"
155
"gettext/rfc822deb", "gettext/schemas",
156
"gettext/scheme", "gettext/xml"
157
-l, --local Writes output into current working directory
158
(conflicts with --update)
159
--update Writes output into the same directory the source file
160
reside (conflicts with --local)
161
--srcdir Root of the source tree
162
-v, --version Output version information and exit
163
-h, --help Display this help and exit
164
-q, --quiet Quiet mode
166
Report bugs to http://bugzilla.gnome.org/ (product name "$PACKAGE")
167
or send email to <xml-i18n-tools\@gnome.org>.
172
## Sub for printing error messages
174
print STDERR "Try `${PROGRAM} --help' for more information.\n";
179
print "Generating C format header file for translation.\n" unless $QUIET_ARG;
187
open OUT, ">$OUTFILE";
191
print "Wrote $OUTFILE\n" unless $QUIET_ARG;
199
local $/; #slurp mode
200
open (IN, "<$SRCDIR_ARG/$FILE") || die "can't open $SRCDIR_ARG/$FILE: $!";
204
&type_ini if $gettext_type eq "ini";
205
&type_keys if $gettext_type eq "keys";
206
&type_xml if $gettext_type eq "xml";
207
&type_glade if $gettext_type eq "glade";
208
&type_scheme if $gettext_type eq "scheme";
209
&type_schemas if $gettext_type eq "schemas";
210
&type_rfc822deb if $gettext_type eq "rfc822deb";
213
sub entity_decode_minimal
239
return '\"' if $_ eq '"';
240
return '\n' if $_ eq "\n";
241
return '\\' if $_ eq '\\';
249
return join "", map &escape_char, split //, $string;
253
### For generic translatable desktop files ###
254
while ($input =~ /^_.*=(.*)$/mg) {
260
### For generic translatable mime/keys files ###
261
while ($input =~ /^\s*_\w+=(.*)$/mg) {
267
### For generic translatable XML files ###
269
while ($input =~ /(?:<!--([^>]*?)-->[^\n]*\n?[^\n]*)?\s_$w+\s*=\s*\"([^"]+)\"/sg) { # "
270
$messages{entity_decode_minimal($2)} = [];
271
$comments{entity_decode_minimal($2)} = $1 if (defined($1));
274
while ($input =~ /(?:<!--([^>]*?)-->\s*)?<_($w+)(?: xml:space="($w+)")?[^>]*>(.+?)<\/_\2>/sg) {
276
if (!defined($3) || $3 ne "preserve") {
282
$comments{$_} = $1 if (defined($1));
287
### For schemas XML files ###
289
# FIXME: We should handle escaped < (less than)
291
<locale\ name="C">\s*
292
(<default>\s*(?:<!--([^>]*?)-->\s*)?(.*?)\s*<\/default>\s*)?
293
(<short>\s*(?:<!--([^>]*?)-->\s*)?(.*?)\s*<\/short>\s*)?
294
(<long>\s*(?:<!--([^>]*?)-->\s*)?(.*?)\s*<\/long>\s*)?
297
my @totranslate = ($3,$6,$9);
298
my @eachcomment = ($2,$5,$8);
299
foreach (@totranslate) {
300
my $currentcomment = shift @eachcomment;
303
$messages{entity_decode_minimal($_)} = [];
304
$comments{entity_decode_minimal($_)} = $currentcomment if (defined($currentcomment));
310
### For rfc822-style Debian configuration files ###
314
while ($input =~ /\G(.*?)(^|\n)(_+)([^:]+):[ \t]*(.*?)(?=\n\S|$)/sg)
316
my ($pre, $newline, $underscore, $tag, $text) = ($1, $2, $3, $4, $5);
317
while ($pre =~ m/\n/g)
321
$lineno += length($newline);
322
my @str_list = rfc822deb_split(length($underscore), $text);
323
for my $str (@str_list)
326
$messages{$str} = [];
327
$loc{$str} = $lineno;
328
$count{$str} = $strcount;
329
my $usercomment = '';
330
while($pre =~ s/(^|\n)#([^\n]*)$//s)
332
$usercomment = "\n" . $2 . $usercomment;
334
$comments{$str} = $tag . $usercomment;
336
$lineno += ($text =~ s/\n//g);
340
sub rfc822deb_split {
341
# Debian defines a special way to deal with rfc822-style files:
342
# when a value contain newlines, it consists of
343
# 1. a short form (first line)
344
# 2. a long description, all lines begin with a space,
345
# and paragraphs are separated by a single dot on a line
346
# This routine returns an array of all paragraphs, and reformat
348
# When first argument is 2, the string is a comma separated list of
352
$text =~ s/^[ \t]//mg;
353
return (split(/, */, $text, 0)) if $type ne 1;
354
return ($text) if $text !~ /\n/;
356
$text =~ s/([^\n]*)\n//;
359
for my $line (split (/\n/, $text))
362
if ($line =~ /^\.\s*$/)
369
elsif ($line =~ /^\s/)
371
# Line which must not be reformatted
372
$str .= "\n" if length ($str) && $str !~ /\n$/;
378
# Continuation line, remove newline
379
$str .= " " if length ($str) && $str !~ /\n$/;
384
push(@list, $str) if length ($str);
389
### For translatable Glade XML files ###
391
my $tags = "label|title|text|format|copyright|comments|preview_text|tooltip|message";
393
while ($input =~ /<($tags)>([^<]+)<\/($tags)>/sg) {
394
# Glade sometimes uses tags that normally mark translatable things for
395
# little bits of non-translatable content. We work around this by not
396
# translating strings that only includes something like label4 or window1.
397
$messages{entity_decode($2)} = [] unless $2 =~ /^(window|label)[0-9]+$/;
400
while ($input =~ /<items>(..[^<]*)<\/items>/sg) {
401
for my $item (split (/\n/, $1)) {
402
$messages{entity_decode($item)} = [];
406
## handle new glade files
407
while ($input =~ /<(property|atkproperty)\s+[^>]*translatable\s*=\s*"yes"[^>]*>([^<]+)<\/\1>/sg) {
408
$messages{entity_decode($2)} = [] unless $2 =~ /^(window|label)[0-9]+$/;
410
while ($input =~ /<atkaction\s+action_name="([^>]*)"\s+description="([^>]+)"\/>/sg) {
411
$messages{entity_decode_minimal($2)} = [];
416
while ($input =~ /_\w*\(?"((?:[^"\\]+|\\.)*)"\)?/sg) {
425
@msgids = sort { $count{$a} <=> $count{$b} } keys %count;
429
@msgids = sort keys %messages;
431
for my $message (@msgids)
434
$offsetlines++ if $message =~ /%/;
435
if (defined ($comments{$message}))
437
while ($comments{$message} =~ m/\n/g)
442
print OUT "# ".($loc{$message} - $offsetlines). " \"$FILE\"\n"
443
if defined $loc{$message};
444
print OUT "/* ".$comments{$message}." */\n"
445
if defined $comments{$message};
446
print OUT "/* xgettext:no-c-format */\n" if $message =~ /%/;
448
my @lines = split (/\n/, $message, -1);
449
for (my $n = 0; $n < @lines; $n++)
453
print OUT "char *s = N_(\"";
460
print OUT escape($lines[$n]);