- cosmetic

[packages/rpm-build-tools.git] / spec_utf8

#!/usr/bin/python

debug = False

import os, os.path, re, sys, locale, StringIO

langs={
    'bg':'windows-1251',
    'br':'iso8859-1',
    'ca':'iso8859-1',
    'cs':'iso8859-2',
    'da':'iso8859-1',
    'de':'iso8859-1',
    'en':'iso8859-1',
    'eo':'iso8859-3',
    'es':'iso8859-1',
    'fi':'iso8859-1',
    'fo':'iso8859-1',
    'fr':'iso8859-1',
    'gl':'iso8859-1',
    'he':'iso8859-8',
    'id':'iso8859-1',
    'is':'iso8859-1',
    'it':'iso8859-1',
    'ja':'euc-jp',
    'ko':'euc-kr',
    'nb':'iso8859-1',
    'nl':'iso8859-1', 
    'pl':'iso8859-2',
    'pt':'iso8859-1',
    'pt_BR':'iso8859-1',
    'ro':'iso8859-2',
    'ru':'KOI8-R',
    'se':'UTF-8',
    'sk':'iso8859-2',
    'sl':'iso8859-2',
    'sv':'iso8859-1',
    'tr':'iso8859-9',
    'uk':'KOI8-U',
    'wa':'iso8859-1',
    'zh_CN':'GB2312',
    'zh_HK':'BIG5-HKSCS',
    'zh_TW':'BIG5',
    0:0}

def find_encoding(lang):
  r = re.match("^([^.]+)(\.[^@]+)?$", lang)
  pure_lang = r.group(1)
  if r.group(2) == None:
    try:
      enc = langs[lang]
    except KeyError:
      enc = None
  else:
    # strip dot
    enc = r.group(2)[1:]
  return (enc, pure_lang)

def parse_spec(infile, outfile):
  success = True
  re_summary = re.compile("^Summary\(([^\)]+)\):\t+(.*)$")
  re_utf = re.compile("^utf-8$", re.I)
  re_desc = re.compile("^(%description.*\s)-l\s+([\S]+)($|\s.*$)")
  re_proc = re.compile("^%")
  re_changelog = re.compile("^%changelog")
  in_desc = False
  in_changelog = False

  for l in infile:
    outline = l
    if debug: outfile.write("%s, %s, %s" % (in_desc, in_changelog, l))

    # %description start
    r = re_desc.match(l)
    if r:
      lang = r.group(2)
      (enc, pure_lang) = find_encoding(lang)
      if enc == None:
        outfile.write("#spec_utf8: unknown lang code in %%description -l %s\n" % (lang))
	success = False
      elif not re_utf.search(enc):
        in_desc = True
        outline = "%s-l %s.UTF-8%s\n" % (r.group(1), pure_lang, r.group(3))
    elif in_desc:
      if re_proc.search(l):
        in_desc = False
      else:
        # %description continues
        if not re_utf.search(enc):
          try:
            outline = unicode(l, enc).encode("UTF-8")
          except UnicodeDecodeError:
            outfile.write("#spec_utf8: transcoding error %%description -l %s\n" % (pure_lang))
	    success = False
    elif in_changelog:
      try:
        outline = unicode(l, "UTF-8").encode("UTF-8")
      except UnicodeDecodeError:
        try:
          outline = unicode(l, "ISO-8859-2").encode("UTF-8")
        except UnicodeDecodeError:
          outfile.write("#spec_utf8: transcoding next line from Latin2 failed\n")
	  success = False
    else: 
      # Summary
      r = re_summary.match(l)
      if r:
        lang = r.group(1)
        (enc, pure_lang) = find_encoding(lang)
        if enc == None:
          outfile.write("#spec_utf8: unknow lang code Summary(%s)\n" % (lang))
	  success = False
        elif not re_utf.search(enc):
          try:
            desc = unicode(r.group(2), enc).encode("UTF-8")
            outline = "Summary(%s.UTF-8):   %s\n" % (pure_lang, desc)
          except UnicodeDecodeError:
            outfile.write("#spec_utf8: transcoding error Summary(%s)\n" % (lang))
	    success = False
      elif re_changelog.match(l):
        # %changelog start
        in_changelog = True
    
    
    if debug: outfile.write("%s, %s\n"% (in_desc, in_changelog))
    outfile.write("%s"% (outline, ))
  return success

def main(argv):
  print "Converting %s ..." % argv[1]
  f = open(argv[1], 'r')
  sio = StringIO.StringIO()
  sio.write(f.read())
  f.close()
  sio.seek(0)
  f = open(argv[1] + '.tmp', 'w')
  if not parse_spec(sio, f):
	  sys.stderr.write("   Problems while converting %s.\n" % argv[1])
  f.close()
  os.rename(argv[1] + '.tmp', argv[1])

if __name__ == "__main__":
  if len(sys.argv) != 2:
	  sys.stderr.write("Usage: %s <spec file>\n" % sys.argv[0])
	  sys.exit(1)
  main(sys.argv)