cldr2json.py

   1 #!/usr/bin/python3
   2 #
   3 # Copyright 2015  Daiki Ueno <dueno@src.gnome.org>
   4 #           2016  Parag Nemade <pnemade@redhat.com>
   5 #           2017  Alan <alan@boum.org>
   6 #
   7 # This program is free software; you can redistribute it and/or modify
   8 # it under the terms of the GNU Lesser General Public License as
   9 # published by the Free Software Foundation; either version 2 of the
  10 # License, or (at your option) any later version.
  11 #
  12 # This program is distributed in the hope that it will be useful, but
  13 # WITHOUT ANY WARRANTY; without even the implied warranty of
  14 # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  15 # Lesser General Public License for more details.
  16 #
  17 # You should have received a copy of the GNU Lesser General Public
  18 # License along with this program; if not, see
  19 # <http://www.gnu.org/licenses/>.
  20
  21 import glob
  22 import json
  23 import locale
  24 import logging
  25 import os
  26 import re
  27 import sys
  28 import xml.etree.ElementTree
  29
  30 import gi
  31 gi.require_version('GnomeDesktop', '3.0')   # NOQA: E402
  32 from gi.repository import GnomeDesktop
  33
  34 ESCAPE_PATTERN = re.compile(r'\\u\{([0-9A-Fa-f]+?)\}')
  35 ISO_PATTERN = re.compile(r'[A-E]([0-9]+)')
  36
  37 LOCALE_TO_XKB_OVERRIDES = {
  38     'af':    'za',
  39     'en':    'us',
  40     'en-GB': 'uk',
  41     'es-US': 'latam',
  42     'fr-CA': 'ca',
  43     'hi':    'in+bolnagri',
  44     'ky':    'kg',
  45     'nl-BE': 'be',
  46     'zu':    None
  47 }
  48
  49
  50 def parse_single_key(value):
  51     def unescape(m):
  52         return chr(int(m.group(1), 16))
  53     value = ESCAPE_PATTERN.sub(unescape, value)
  54     return value
  55
  56
  57 def parse_rows(keymap):
  58     unsorted_rows = {}
  59     for _map in keymap.iter('map'):
  60         value = _map.get('to')
  61         key = [parse_single_key(value)]
  62         iso = _map.get('iso')
  63         if not ISO_PATTERN.match(iso):
  64             sys.stderr.write('invalid ISO key name: %s\n' % iso)
  65             continue
  66         if not iso[0] in unsorted_rows:
  67             unsorted_rows[iso[0]] = []
  68         unsorted_rows[iso[0]].append((int(iso[1:]), key))
  69         # add subkeys
  70         longPress = _map.get('longPress')
  71         if longPress:
  72             for value in longPress.split(' '):
  73                 subkey = parse_single_key(value)
  74                 key.append(subkey)
  75
  76     rows = []
  77     for k, v in sorted(list(unsorted_rows.items()),
  78                        key=lambda x: x[0],
  79                        reverse=True):
  80         row = []
  81         for key in sorted(v, key=lambda x: x):
  82             row.append(key[1])
  83         rows.append(row)
  84
  85     return rows
  86
  87
  88 def convert_xml(tree):
  89     root = {}
  90     for xml_keyboard in tree.iter("keyboard"):
  91         locale_full = xml_keyboard.get("locale")
  92         locale, sep, end = locale_full.partition("-t-")
  93     root["locale"] = locale
  94     for xml_name in tree.iter("name"):
  95         name = xml_name.get("value")
  96     root["name"] = name
  97     root["levels"] = []
  98     # parse levels
  99     for index, keymap in enumerate(tree.iter('keyMap')):
 100         # FIXME: heuristics here
 101         modifiers = keymap.get('modifiers')
 102         if not modifiers:
 103             mode = 'default'
 104             modifiers = ''
 105         elif 'shift' in modifiers.split(' '):
 106             mode = 'latched'
 107             modifiers = 'shift'
 108         else:
 109             mode = 'locked'
 110         level = {}
 111         level["level"] = modifiers
 112         level["mode"] = mode
 113         level["rows"] = parse_rows(keymap)
 114         root["levels"].append(level)
 115     return root
 116
 117
 118 def locale_to_xkb(locale, name):
 119     if locale in sorted(LOCALE_TO_XKB_OVERRIDES.keys()):
 120         xkb = LOCALE_TO_XKB_OVERRIDES[locale]
 121         logging.debug("override for %s → %s",
 122                       locale, xkb)
 123         if xkb:
 124             return xkb
 125         else:
 126             raise KeyError("layout %s explicitely disabled in overrides"
 127                            % locale)
 128     xkb_names = sorted(name_to_xkb.keys())
 129     if name in xkb_names:
 130         return name_to_xkb[name]
 131     else:
 132         logging.debug("name %s failed" % name)
 133     for sub_name in name.split(' '):
 134         if sub_name in xkb_names:
 135             xkb = name_to_xkb[sub_name]
 136             logging.debug("dumb mapping failed but match with locale word: "
 137                           "%s (%s) → %s (%s)",
 138                           locale, name, xkb, sub_name)
 139             return xkb
 140         else:
 141             logging.debug("sub_name failed")
 142     for xkb_name in xkb_names:
 143         for xkb_sub_name in xkb_name.split(' '):
 144             if xkb_sub_name.strip('()') == name:
 145                 xkb = name_to_xkb[xkb_name]
 146                 logging.debug("dumb mapping failed but match with xkb word: "
 147                               "%s (%s) → %s (%s)",
 148                               locale, name, xkb, xkb_name)
 149                 return xkb
 150     raise KeyError("failed to find XKB mapping for %s" % locale)
 151
 152
 153 def convert_file(source_file, destination_path):
 154     logging.info("Parsing %s", source_file)
 155
 156     itree = xml.etree.ElementTree.ElementTree()
 157     itree.parse(source_file)
 158
 159     root = convert_xml(itree)
 160
 161     try:
 162         xkb_name = locale_to_xkb(root["locale"], root["name"])
 163     except KeyError as e:
 164         logging.warn(e)
 165         return False
 166     destination_file = os.path.join(destination_path, xkb_name + ".json")
 167
 168     with open(destination_file, 'w', encoding="utf-8") as dest_fd:
 169         json.dump(root, dest_fd, ensure_ascii=False, indent=2, sort_keys=True)
 170
 171     logging.debug("written %s", destination_file)
 172
 173
 174 def load_xkb_mappings():
 175     xkb = GnomeDesktop.XkbInfo()
 176     layouts = xkb.get_all_layouts()
 177     name_to_xkb = {}
 178
 179     for layout in layouts:
 180         name = xkb.get_layout_info(layout).display_name
 181         name_to_xkb[name] = layout
 182
 183     return name_to_xkb
 184
 185
 186 locale.setlocale(locale.LC_ALL, "C")
 187 name_to_xkb = load_xkb_mappings()
 188
 189
 190 if __name__ == "__main__":
 191     if "DEBUG" in os.environ:
 192         logging.basicConfig(level=logging.DEBUG)
 193
 194     if len(sys.argv) < 2:
 195         print("supply a CLDR keyboard file")
 196         sys.exit(1)
 197
 198     if len(sys.argv) < 3:
 199         print("supply an output directory")
 200         sys.exit(1)
 201
 202     source = sys.argv[1]
 203     destination = sys.argv[2]
 204     if os.path.isfile(source):
 205         convert_file(source, destination)
 206     elif os.path.isdir(source):
 207         for path in glob.glob(source + "/*-t-k0-android.xml"):
 208             convert_file(path, destination)