src/util/xmlpool/gen_xmlpool.py

   1 # encoding=utf-8
   2 #
   3 # Usage:
   4 #     gen_xmlpool.py /path/to/t_option.h localedir lang lang lang ...
   5 #
   6 # For each given language, this script expects to find a .mo file at
   7 # `{localedir}/{language}/LC_MESSAGES/options.mo`.
   8 #
   9
  10 from __future__ import print_function, unicode_literals
  11 import argparse
  12 import gettext
  13 import io
  14 import os
  15 import re
  16 import sys
  17
  18 if sys.version_info < (3, 0):
  19     gettext_method = 'ugettext'
  20 else:
  21     gettext_method = 'gettext'
  22
  23 # Escape special characters in C strings
  24 def escapeCString(s):
  25     escapeSeqs = {'\a' : '\\a', '\b' : '\\b', '\f' : '\\f', '\n' : '\\n',
  26                   '\r' : '\\r', '\t' : '\\t', '\v' : '\\v', '\\' : '\\\\'}
  27     # " -> '' is a hack. Quotes (") aren't possible in XML attributes.
  28     # Better use Unicode characters for typographic quotes in option
  29     # descriptions and translations.
  30     last_quote = '”'
  31     i = 0
  32     r = ''
  33     for c in s:
  34         # Special case: escape double quote with “ or ”, depending
  35         # on whether it's an open or close quote. This is needed because plain
  36         # double quotes are not possible in XML attributes.
  37         if c == '"':
  38             if last_quote == '”':
  39                 q = '“'
  40             else:
  41                 q = '”'
  42             last_quote = q
  43             r = r + q
  44         elif c in escapeSeqs:
  45             r = r + escapeSeqs[c]
  46         else:
  47             r = r + c
  48     return r
  49
  50 # Expand escape sequences in C strings (needed for gettext lookup)
  51 def expandCString(s):
  52     escapeSeqs = {'a' : '\a', 'b' : '\b', 'f' : '\f', 'n' : '\n',
  53                   'r' : '\r', 't' : '\t', 'v' : '\v',
  54                   '"' : '"', '\\' : '\\'}
  55     escape = False
  56     hexa = False
  57     octa = False
  58     num = 0
  59     digits = 0
  60     r = u''
  61     for c in s:
  62         if not escape:
  63             if c == '\\':
  64                 escape = True
  65             else:
  66                 r = r + c
  67         elif hexa:
  68             if (c >= '0' and c <= '9') or \
  69                (c >= 'a' and c <= 'f') or \
  70                (c >= 'A' and c <= 'F'):
  71                 num = num * 16 + int(c, 16)
  72                 digits = digits + 1
  73             else:
  74                 digits = 2
  75             if digits >= 2:
  76                 hexa = False
  77                 escape = False
  78                 r = r + chr(num)
  79         elif octa:
  80             if c >= '0' and c <= '7':
  81                 num = num * 8 + int(c, 8)
  82                 digits = digits + 1
  83             else:
  84                 digits = 3
  85             if digits >= 3:
  86                 octa = False
  87                 escape = False
  88                 r = r + chr(num)
  89         else:
  90             if c in escapeSeqs:
  91                 r = r + escapeSeqs[c]
  92                 escape = False
  93             elif c >= '0' and c <= '7':
  94                 octa = True
  95                 num = int(c, 8)
  96                 if num <= 3:
  97                     digits = 1
  98                 else:
  99                     digits = 2
 100             elif c == 'x' or c == 'X':
 101                 hexa = True
 102                 num = 0
 103                 digits = 0
 104             else:
 105                 r = r + c
 106                 escape = False
 107     return r
 108
 109 # Expand matches. The first match is always a DESC or DESC_BEGIN match.
 110 # Subsequent matches are ENUM matches.
 111 #
 112 # DESC, DESC_BEGIN format: \1 \2=<lang> \3 \4=gettext(" \5=<text> \6=") \7
 113 # ENUM format:             \1 \2=gettext(" \3=<text> \4=") \5
 114 def expandMatches(matches, translations, outfile, end=None):
 115     assert len(matches) > 0
 116     nTranslations = len(translations)
 117     i = 0
 118     # Expand the description+enums for all translations
 119     for lang,trans in translations:
 120         i = i + 1
 121         # Make sure that all but the last line of a simple description
 122         # are extended with a backslash.
 123         suffix = ''
 124         if len(matches) == 1 and i < len(translations) and \
 125                not matches[0].expand(r'\7').endswith('\\'):
 126             suffix = ' \\'
 127         text = escapeCString(getattr(trans, gettext_method)(expandCString(
 128             matches[0].expand (r'\5'))))
 129         text = (matches[0].expand(r'\1' + lang + r'\3"' + text + r'"\7') + suffix)
 130
 131         outfile.write(text + '\n')
 132
 133         # Expand any subsequent enum lines
 134         for match in matches[1:]:
 135             text = escapeCString(getattr(trans, gettext_method)(expandCString(
 136                 match.expand(r'\3'))))
 137             text = match.expand(r'\1"' + text + r'"\5')
 138
 139             outfile.write(text + '\n')
 140
 141         # Expand description end
 142         if end:
 143             outfile.write(end)
 144
 145 # Regular expressions:
 146 reLibintl_h = re.compile(r'#\s*include\s*<libintl.h>')
 147 reDESC = re.compile(r'(\s*DRI_CONF_DESC\s*\(\s*)([a-z]+)(\s*,\s*)(gettext\s*\(\s*")(.*)("\s*\))(\s*\)[ \t]*\\?)$')
 148 reDESC_BEGIN = re.compile(r'(\s*DRI_CONF_DESC_BEGIN\s*\(\s*)([a-z]+)(\s*,\s*)(gettext\s*\(\s*")(.*)("\s*\))(\s*\)[ \t]*\\?)$')
 149 reENUM = re.compile(r'(\s*DRI_CONF_ENUM\s*\([^,]+,\s*)(gettext\s*\(\s*")(.*)("\s*\))(\s*\)[ \t]*\\?)$')
 150 reDESC_END = re.compile(r'\s*DRI_CONF_DESC_END')
 151
 152
 153 def main():
 154     parser = argparse.ArgumentParser()
 155     parser.add_argument('--template', required=True)
 156     parser.add_argument('--output', required=True)
 157     parser.add_argument('--localedir', required=True)
 158     parser.add_argument('--languages', nargs='*', default=[])
 159     args = parser.parse_args()
 160
 161     # Compile a list of translation classes to all supported languages.
 162     # The first translation is always a NullTranslations.
 163     translations = [("en", gettext.NullTranslations())]
 164     for lang in args.languages:
 165         try:
 166             filename = os.path.join(args.localedir, '{}.gmo'.format(lang))
 167             with io.open(filename, 'rb') as f:
 168                 trans = gettext.GNUTranslations(f)
 169         except (IOError, OSError):
 170             print("Warning: language '%s' not found." % lang, file=sys.stderr)
 171             continue
 172         translations.append((lang, trans))
 173
 174     with io.open(args.output, mode='wt', encoding='utf-8') as output:
 175         output.write("/* This is file is generated automatically. Don't edit!  */\n")
 176
 177         # Process the options template and generate options.h with all
 178         # translations.
 179         with io.open(args.template, mode="rt", encoding='utf-8') as template:
 180             descMatches = []
 181             for line in template:
 182                 if descMatches:
 183                     matchENUM = reENUM.match(line)
 184                     matchDESC_END = reDESC_END.match(line)
 185                     if matchENUM:
 186                         descMatches.append(matchENUM)
 187                     elif matchDESC_END:
 188                         expandMatches(descMatches, translations, output, line)
 189                         descMatches = []
 190                     else:
 191                         print("Warning: unexpected line inside description dropped:\n",
 192                               line, file=sys.stderr)
 193                     continue
 194                 if reLibintl_h.search(line):
 195                     # Ignore (comment out) #include <libintl.h>
 196                     output.write("/* %s * commented out by gen_xmlpool.py */\n" % line)
 197                     continue
 198                 matchDESC = reDESC.match(line)
 199                 matchDESC_BEGIN = reDESC_BEGIN.match(line)
 200                 if matchDESC:
 201                     assert not descMatches
 202                     expandMatches([matchDESC], translations, output)
 203                 elif matchDESC_BEGIN:
 204                     assert not descMatches
 205                     descMatches = [matchDESC_BEGIN]
 206                 else:
 207
 208                     output.write(line)
 209
 210         if descMatches:
 211             print("Warning: unterminated description at end of file.", file=sys.stderr)
 212             expandMatches(descMatches, translations, output)
 213
 214
 215 if __name__ == '__main__':
 216     main()