solenv/bin/desktop-translate.py

   1 #
   2 # This file is part of the LibreOffice project.
   3 #
   4 # This Source Code Form is subject to the terms of the Mozilla Public
   5 # License, v. 2.0. If a copy of the MPL was not distributed with this
   6 # file, You can obtain one at http://mozilla.org/MPL/2.0/.
   7 #
   8 # This file incorporates work covered by the following license notice:
   9 #
  10 #   Licensed to the Apache Software Foundation (ASF) under one or more
  11 #   contributor license agreements. See the NOTICE file distributed
  12 #   with this work for additional information regarding copyright
  13 #   ownership. The ASF licenses this file to you under the Apache
  14 #   License, Version 2.0 (the "License"); you may not use this file
  15 #   except in compliance with the License. You may obtain a copy of
  16 #   the License at http://www.apache.org/licenses/LICENSE-2.0 .
  17 #
  18
  19 #
  20 # Translates multiple .desktop files at once with strings from .ulf
  21 # files; if you add new translatable .ulf files please add them to
  22 # l10ntools/source/localize.cxx
  23 #
  24
  25 import os, sys, argparse, io
  26
  27 def encodeDesktopString(s):
  28     # <https://specifications.freedesktop.org/desktop-entry-spec/desktop-entry-spec-1.1.html#
  29     # value-types> says "The escape sequences \s, \n, \t, \r, and \\ are supported for values of
  30     # type string and localestring, meaning ASCII space, newline, tab, carriage return, and
  31     # backslash, respectively."  <https://specifications.freedesktop.org/desktop-entry-spec/
  32     # desktop-entry-spec-1.1.html#basic-format> says "A file is interpreted as a series of lines
  33     # that are separated by linefeed characters", so it is apparently necessary to escape at least
  34     # linefeed and backslash characters.  It is unclear why that spec talks about "linefeed" in
  35     # one place and about "newline" ("\n") and "carriage return" ("\r") in another, and how they are
  36     # supposed to relate, so just escape any U+000A LINE FEED as "\n" and any U+000D CARRIAGE RETURN
  37     # as "\r"; it is unclear exactly which occurrences of U+0020 SPACE and U+0009 CHARACTER
  38     # TABULATION would need to be escaped, so they are mostly left unescaped, for readability:
  39     s = s.replace('\\', '\\\\').replace('\n', '\\n').replace('\r', '\\r');
  40     if s.startswith(' '):
  41         # <https://specifications.freedesktop.org/desktop-entry-spec/desktop-entry-spec-1.1.html#
  42         # entries> says "Space before and after the equals sign should be ignored", so escape a
  43         # leading U+0020 SPACE as "\s" (while it is not clear whether "space" there means just
  44         # U+0020 SPACE or any kind of white space, in which case at least a leading U+0009 CHARACTER
  45         # TABULATION should similarly be escaped as "\t"; also, it is unclear whether such
  46         # characters should also be escaped at the end):
  47         s = '\\s' + s[1:]
  48     return s
  49
  50 parser = argparse.ArgumentParser()
  51 parser.add_argument('-p', dest='productname', default='LibreOffice')
  52 parser.add_argument('-d', dest='workdir', default='.')
  53 parser.add_argument('--key', dest='key')
  54 parser.add_argument('--prefix', dest='prefix', default='')
  55 parser.add_argument('--ext', dest='ext')
  56 parser.add_argument('--template-dir', dest='template_dir', default=None)
  57 parser.add_argument('ifile')
  58
  59 o = parser.parse_args()
  60
  61 if o.template_dir is None:
  62     template_dir = '{}/{}'.format(o.workdir, o.prefix)
  63 else:
  64     template_dir = o.template_dir
  65
  66 # hack for unity section
  67 if o.key == "UnityQuickList":
  68     outkey = "Name"
  69 else:
  70     outkey = o.key
  71
  72
  73 templates = {}
  74
  75 # open input file
  76 source = io.open(o.ifile, encoding='utf-8')
  77
  78 template = None
  79
  80 # read ulf file
  81 for line in source:
  82     if line.strip() == '':
  83         continue
  84     if line[0] == "[":
  85         template = line.split(']', 1)[0][1:]
  86         entry = {}
  87         # For every section in the specified ulf file there should exist
  88         # a template file in $workdir ..
  89         entry['outfile'] = "{}{}.{}".format(template_dir, template, o.ext)
  90         entry['translations'] = {}
  91         templates[template] = entry
  92     else:
  93         # split locale = "value" into 2 strings
  94         if ' = ' not in line:
  95             continue
  96         locale, value = line.split(' = ')
  97
  98         if locale != line:
  99             # replace en-US with en
 100             locale = locale.replace('en-US', 'en')
 101
 102             # use just anything inside the ""
 103             assert(value[0] == '"')
 104             # Some entries span multiple lines.
 105             # An entry will always end on a double quote.
 106             while not value.endswith('"\n'):
 107                 value += source.readline()
 108             value = value[1:-2]
 109
 110             # replace resource placeholder
 111             value = value.replace('%PRODUCTNAME', o.productname)
 112
 113             locale = locale.replace('-', '_')
 114
 115             templates[template]['translations'][locale] = value
 116
 117 source.close()
 118
 119 processed = 0
 120 # process templates
 121 for template in templates:
 122     outfilename = templates[template]['outfile']
 123
 124     # open the template file - ignore sections for which no
 125     # templates exist
 126     try:
 127         template_file = io.open(outfilename, encoding='utf-8')
 128     except Exception:
 129         # string files processed one by one
 130         if o.ext == 'str':
 131             continue
 132         sys.exit("Warning: No template found for item '{}' : '{}' : '{}': $!\n".format(template, outfile, line))
 133     processed += 1
 134
 135     # open output file
 136     tmpfilename = '{}.tmp'.format(outfilename)
 137     outfile = io.open(tmpfilename, 'w', encoding='utf-8')
 138
 139     # emit the template to the output file
 140     for line in template_file:
 141         keyline = line
 142         if keyline.startswith(o.key):
 143             keyline = outkey + keyline[len(o.key):]
 144         outfile.write(keyline)
 145         if o.key in line:
 146             translations = templates[template]['translations']
 147             for locale in sorted (translations.keys()):
 148                 value = translations.get(locale, None)
 149                 # print "locale is $locale\n";
 150                 # print "value is $value\n";
 151                 if value:
 152                     if o.ext == "desktop" or o.ext == "str":
 153                         if o.ext == "desktop":
 154                             value = encodeDesktopString(value)
 155                         outfile.write(u"""{}[{}]={}\n""".format(outkey, locale, value))
 156                     else:
 157                         outfile.write(u"""\t[{}]{}={}\n""".format(locale, outkey, value))
 158
 159     template_file.close()
 160
 161     outfile.close()
 162     if os.path.exists(outfilename):
 163         os.unlink(outfilename)
 164     os.rename(tmpfilename, outfilename)
 165
 166 if o.ext == 'str' and processed == 0:
 167     sys.exit("Warning: No matching templates processed")