Added a prefix_lookup() function, for suggestions.

[zzz-pokedex.git] / pokedex / roomaji.py

diff --git a/pokedex/roomaji.py b/pokedex/roomaji.py
index 0a6b269..46f39b5 100644 (file)
--- a/pokedex/roomaji.py
+++ b/pokedex/roomaji.py
@@ -22,7 +22,7 @@ _roomaji_kana = {
 
 _roomaji_youon = {
     u'ャ': 'ya',                    u'ュ': 'yu',                    u'ョ': 'yo',
 
 _roomaji_youon = {
     u'ャ': 'ya',                    u'ュ': 'yu',                    u'ョ': 'yo',

-    u'ゃ': 'ya',                    u'ゅ': 'yu',                    u'ょ': 'yo',
+    #u'ゃ': 'ya',                    u'ゅ': 'yu',                    u'ょ': 'yo',

 }
 
 # XXX If romanize() ever handles hiragana, it will need to make sure that the
 }
 
 # XXX If romanize() ever handles hiragana, it will need to make sure that the


@@ -73,12 +73,12 @@ def romanize(string):
 
         # Youon
         elif char in _roomaji_youon:
 
         # Youon
         elif char in _roomaji_youon:

-            if last_kana[-1] != 'i' or last_kana == 'i':
+            if not last_kana or last_kana[-1] != 'i' or last_kana == 'i':

                 raise ValueError("Youon must follow an -i sound.")
 
             # Drop the -i and append the ya/yu/yo sound
             new_sound = _roomaji_youon[char]
                 raise ValueError("Youon must follow an -i sound.")
 
             # Drop the -i and append the ya/yu/yo sound
             new_sound = _roomaji_youon[char]

-            if last_kana in ['shi', 'ji']:
+            if last_kana in ['chi', 'shi', 'ji']:

                 # Strip the y-
                 new_char = last_kana[:-1] + new_sound[1:]
             else:
                 # Strip the y-
                 new_char = last_kana[:-1] + new_sound[1:]
             else:


@@ -88,7 +88,8 @@ def romanize(string):
             last_kana = new_char
 
         # Sokuon
             last_kana = new_char
 
         # Sokuon

-        elif char in (u'っ', u'ッ'):
+        #elif char in (u'っ', u'ッ'):
+        elif char in (u'ッ',):

             # Remember it and double the consonant next time around
             last_kana = 'sokuon'
 
             # Remember it and double the consonant next time around
             last_kana = 'sokuon'
 


@@ -131,4 +132,4 @@ def romanize(string):
     if last_kana == 'sokuon':
         raise ValueError("Sokuon cannot be the last character.")
 
     if last_kana == 'sokuon':
         raise ValueError("Sokuon cannot be the last character.")
 

-    return ''.join(characters)
+    return unicode(''.join(characters))