Add migration script to move English texts to language-specific tables

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

diff --git a/pokedex/roomaji.py b/pokedex/roomaji.py
index d0aa3ca..e91c960 100644 (file)
--- a/pokedex/roomaji.py
+++ b/pokedex/roomaji.py
@@ -37,6 +37,7 @@ _roomaji_kana = {
     u'ダ': 'da',    u'ヂ': 'ji',    u'ヅ': 'dzu',   u'デ': 'de',    u'ド': 'do',
     u'バ': 'ba',    u'ビ': 'bi',    u'ブ': 'bu',    u'ベ': 'be',    u'ボ': 'bo',
     u'パ': 'pa',    u'ピ': 'pi',    u'プ': 'pu',    u'ペ': 'pe',    u'ポ': 'po',
     u'ダ': 'da',    u'ヂ': 'ji',    u'ヅ': 'dzu',   u'デ': 'de',    u'ド': 'do',
     u'バ': 'ba',    u'ビ': 'bi',    u'ブ': 'bu',    u'ベ': 'be',    u'ボ': 'bo',
     u'パ': 'pa',    u'ピ': 'pi',    u'プ': 'pu',    u'ペ': 'pe',    u'ポ': 'po',

+                                    u'ヴ': 'vu',

 }
 
 _roomaji_youon = {
 }
 
 _roomaji_youon = {


@@ -55,12 +56,24 @@ _roomaji_small_kana = {
     u'ァ': 'a',     u'ィ': 'i',     u'ゥ': 'u',     u'ェ': 'e',     u'ォ': 'o',
 }
 _roomaji_small_kana_combos = {
     u'ァ': 'a',     u'ィ': 'i',     u'ゥ': 'u',     u'ェ': 'e',     u'ォ': 'o',
 }
 _roomaji_small_kana_combos = {

-                                    u'ウィ': 'wi',
+    # These are, by the way, fairly arbitrary.  "shi xi" to mean "sy" is
+    # particularly weird, but it seems to be what GF intends
+
+    # Simple vowel replacement
+                    u'ウィ': 'wi',  u'ウゥ': 'wu',  u'ウェ': 'we',  u'ウォ': 'wo',
+    u'ヴァ': 'va',  u'ヴィ': 'vi',                  u'ヴェ': 've',  u'ヴォ': 'vo',

                                                     u'チェ': 'che',
                                                     u'シェ': 'she',
                                                     u'チェ': 'che',
                                                     u'シェ': 'she',

+                                                    u'ジェ': 'je',

     u'テァ': 'tha', u'ティ': 'ti',  u'テゥ': 'thu', u'テェ': 'tye', u'テォ': 'tho',
     u'デァ': 'dha', u'ディ': 'di',  u'デゥ': 'dhu', u'デェ': 'dye', u'デォ': 'dho',
     u'ファ': 'fa',  u'フィ': 'fi',  u'ホゥ': 'hu',  u'フェ': 'fe',  u'フォ': 'fo',
     u'テァ': 'tha', u'ティ': 'ti',  u'テゥ': 'thu', u'テェ': 'tye', u'テォ': 'tho',
     u'デァ': 'dha', u'ディ': 'di',  u'デゥ': 'dhu', u'デェ': 'dye', u'デォ': 'dho',
     u'ファ': 'fa',  u'フィ': 'fi',  u'ホゥ': 'hu',  u'フェ': 'fe',  u'フォ': 'fo',

+
+    # Not so much
+    u'シィ': 'sy',
+    u'ミィ': 'my',
+    u'ビィ': 'by',
+    u'ピィ': 'py',

 }
 
 def romanize(string):
 }
 
 def romanize(string):


@@ -73,11 +86,12 @@ def romanize(string):
     last_char = None  # Used for small kana combos
     for char in string:
         # Full-width Latin
     last_char = None  # Used for small kana combos
     for char in string:
         # Full-width Latin

-        if ord(char) >= 0xff11 and ord(char) <= 0xff5e:
+        if 0xff01 <= ord(char) <= 0xff5e:

             if last_kana == 'sokuon':
                 raise ValueError("Sokuon cannot precede Latin characters.")
 
             if last_kana == 'sokuon':
                 raise ValueError("Sokuon cannot precede Latin characters.")
 

-            char = chr(ord(char) - 0xff11 + 0x31)
+            # XXX Real Unicode decomposition would be nicer
+            char = chr(ord(char) - 0xff01 + 0x21)

             characters.append(char)
 
             last_kana = None
             characters.append(char)
 
             last_kana = None


@@ -113,8 +127,7 @@ def romanize(string):
             last_kana = new_char
 
         # Sokuon
             last_kana = new_char
 
         # Sokuon

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