Debian Codegraph @

src/calibre/ebooks/unihandecode/jadecoder.py

raw ◀ ▶
1 Debian patch touch this file
1 # License: GPLv3 Copyright: 2010, Hiroshi Miura <[email protected]>
2
3 """
4 Decode unicode text to an ASCII representation of the text for Japanese.
5 Translate unicode string to ASCII roman string.
6
7 API is based on the python unidecode,
8 which is based on Ruby gem (http://rubyforge.org/projects/unidecode/)
9 and perl module Text::Unidecode
10 (http://search.cpan.org/~sburke/Text-Unidecode-0.04/).
11
12 This functionality is owned by Kakasi Japanese processing engine.
13
14 Copyright (c) 2010 Hiroshi Miura
15 """
16
17 import re
18 import warnings
19 from importlib.resources import files
20
21 from pykakasi import kakasi
22 from pykakasi.kanji import Itaiji, Kanwa
23 from pykakasi.properties import Configurations
24 from pykakasi.scripts import Jisyo
25
26 from calibre.ebooks.unihandecode.jacodepoints import CODEPOINTS as JACODES
27 from calibre.ebooks.unihandecode.unicodepoints import CODEPOINTS
28 from calibre.ebooks.unihandecode.unidecoder import Unidecoder
29
30
31 class Jadecoder(Unidecoder):
32 def __init__(self):
33 self.codepoints = CODEPOINTS.copy()
34 self.codepoints.update(JACODES)
35
36 # We have to use the deprecated API as the new API does not capitalize
37 # words. Sigh.
38 # https://codeberg.org/miurahr/pykakasi/issues/172
39 with warnings.catch_warnings():
40 warnings.simplefilter('ignore')
41 self.kakasi = kakasi()
42 self.kakasi.setMode('H', 'a') # Hiragana to ascii, default: no conversion
43 self.kakasi.setMode('K', 'a') # Katakana to ascii, default: no conversion
44 self.kakasi.setMode('J', 'a') # Japanese to ascii, default: no conversion
45 self.kakasi.setMode('r', 'Hepburn') # default: use Hepburn Roman table
46 self.kakasi.setMode('s', True) # add space, default: no separator
47 self.kakasi.setMode('C', True) # capitalize, default: no capitalize
48 self.conv = self.kakasi.getConverter()
49
50 def decode(self, text):
51 try:
52 with warnings.catch_warnings():
53 warnings.simplefilter('ignore')
54 text = self.conv.do(text)
55 except Exception:
56 pass
57 return re.sub(r'[^\x00-\x7f]', lambda x: self.replace_point(x.group()), text)