Created
January 12, 2012 15:00
-
-
Save reclosedev/1600988 to your computer and use it in GitHub Desktop.
Implementation of transliteration class, as a simple state machine, with posibility to call it from one-key-prees events. E.g for implementing QTextEdit with translit
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
#!/usr/bin/env python | |
# -*- coding: utf-8 -*- | |
# http://stackoverflow.com/a/8785689/1052325 | |
class SMTransliterator(object): | |
two_letters = {u'Ch': u'Ч', u'yo': u'ё', u'Yo': u'Ё', u'Ya': u'Я', u'yu': u'ю', | |
u'sh': u'щ', u'ch': u'ч', u'ya': u'я', u'Sh': u'Щ', u'zh': u'ж', u'Yu': u'Ю', } | |
two_letters_set = set([s[0] for s in two_letters]) | |
single_letters = {u'A': u'А', u'C': u'Ц', u'B': u'Б', u'E': u'Е', u'D': u'Д', u'G': u'Г', | |
u'F': u'Ф', u'I': u'И', u'H': u'Х', u'K': u'К', u'M': u'М', u'L': u'Л', u'O': u'О', u'N': u'Н', | |
u'P': u'П', u'S': u'С', u'R': u'Р', u'U': u'У', u'T': u'Т', u'V': u'В', u'Y': u'Й', u'Z': u'З', | |
u'a': u'а', u'c': u'ц', u'b': u'б', u'e': u'е', u'd': u'д', u'g': u'г', u'f': u'ф', u'i': u'и', | |
u'h': u'х', u'k': u'к', u'm': u'м', u'l': u'л', u'o': u'о', u'n': u'н', u'p': u'п', u's': u'с', | |
u'r': u'р', u'u': u'у', u't': u'т', u'v': u'в', u'y': u'й', u'z': u'з', u'J': u'Ж', | |
u'j': u'ж'} | |
def __init__(self): | |
self._buffer = u'' | |
self._result = u'' | |
def add_letter(self, letter): | |
if letter not in self.two_letters_set and letter not in self.single_letters: | |
self._result = self.single_letters.get(self._buffer, u'') | |
self._buffer = u'' | |
return False | |
if self.need_next: | |
s = self._buffer + letter | |
if s in self.two_letters: | |
self._result = self.two_letters[s] | |
else: | |
self._result = self.single_letters[self._buffer] + self.single_letters[letter] | |
self._buffer = u'' | |
return True | |
if letter in self.two_letters_set: | |
self._buffer += letter | |
return True | |
if letter in self.single_letters: | |
self._buffer = u'' | |
self._result = self.single_letters[letter] | |
return True | |
return False | |
@property | |
def need_next(self): | |
return len(self._buffer) > 0 | |
@property | |
def result(self): | |
return self._result | |
def demo(): | |
tr = SMTransliterator() | |
def on_key_press(text): | |
if not tr.add_letter(text): | |
print 'Pass:', text | |
return | |
if tr.need_next: | |
print 'Next:', text | |
return | |
print 'Add:', tr.result | |
# import trans | |
# text = u'Тестирование работы транслитераторатора с шипящими звуками и алфавитом АБВГДЕЁЖЗИЙКЛМНОПРСТУФХЦЧЩ'.encode('trans') + u' А это без замены' | |
# print text | |
text = u'Testirovanie raboty transliteratoratora s shipyashimi zvukami i alfavitom ABVGDEYoZhZIYKLMNOPRSTUFHCChSh А это без замены' | |
for letter in text: | |
on_key_press(letter) | |
if __name__ == '__main__': | |
demo() | |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment