Add pt9 and pt9all options for Partext 9 format

scripts/makeoxt

@@ -121,7 +121,7 @@ parser.add_argument('-d','--dict',help='Wordlist dictionary. For hunspell dictio
 parser.add_argument('-a','--affix',help='Merge the given affix file data into the generated .aff file')
 parser.add_argument('-m','--merge',help='Merge with an existing .oxt')
 parser.add_argument('-v','--version',default='0.1',help='OXT version number')
-parser.add_argument('--dicttype',help='Specifies dictionary type [hunspell, pt, ptall, text]')
+parser.add_argument('--dicttype',help='Specifies dictionary type [hunspell, pt9, pt9all, text, pt, ptall]')
 parser.add_argument('--publisher',help='Name of publisher')
 parser.add_argument('--puburl',default='',help='URL of publisher')
 parser.add_argument('--license',default='',help='Add your own license in a text file (.txt)')
@@ -241,6 +241,21 @@ if args.dict :
  if args.affix is not None :
  hun.mergeaffix(args.affix)
  ziphunspell(ozip, hun, args.langtag)
+ elif args.dicttype == 'pt9' or args.dicttype == 'pt9all' :
+ itemcount = 0
+ wordcount = 0
+ doc = et.parse(args.dict)
+ hun = hs.Hunspell(args.langtag, puncs=args.word)
+ for e in doc.findall('//Status') :
+ itemcount += 1
+ if args.dicttype != 'pt9all' and e.attrib['State'] != 'R' : continue
+ hun.addword(str(e.attrib['Word']))
+ wordcount += 1
+ if wordcount * 4 < itemcount : #warn if less than 25% of the words are valid
+ print("Warning: only {:.0f}% of the words marked as correct and entered into the dictionary. Consider using --dicttype pt9all".format(wordcount / float(itemcount) * 100) )
+ if args.affix is not None :
+ hun.mergeaffix(args.affix)
+ ziphunspell(ozip, hun, args.langtag)
  elif args.dicttype == 'text' :
  hun = hs.Hunspell(args.langtag, puncs=args.word)
  with codecs.open(args.dict, encoding='utf-8') as infile :