20
20
"""Some functions to enable caching the conversion between unicode to utf8"""
25
_utf8_encode = codecs.getencoder("utf-8")
26
_utf8_decode = codecs.getdecoder("utf-8")
23
28
# Map revisions from and to utf8 encoding
24
29
# Whenever we do an encode/decode operation, we save the result, so that
30
35
def encode(unicode_str,
31
36
_uni_to_utf8=_unicode_to_utf8_map,
32
_utf8_to_uni=_utf8_to_unicode_map):
37
_utf8_to_uni=_utf8_to_unicode_map,
38
_utf8_encode=_utf8_encode):
33
39
"""Take this unicode revision id, and get a unicode version"""
34
40
# If the key is in the cache try/KeyError is 50% faster than
35
41
# val = dict.get(key), if val is None:
43
49
return _uni_to_utf8[unicode_str]
45
_uni_to_utf8[unicode_str] = utf8_str = unicode_str.encode('utf-8')
51
_uni_to_utf8[unicode_str] = utf8_str = _utf8_encode(unicode_str)[0]
46
52
_utf8_to_uni[utf8_str] = unicode_str
50
56
def decode(utf8_str,
51
57
_uni_to_utf8=_unicode_to_utf8_map,
52
_utf8_to_uni=_utf8_to_unicode_map):
58
_utf8_to_uni=_utf8_to_unicode_map,
59
_utf8_decode=_utf8_decode):
53
60
"""Take a utf8 revision id, and decode it, but cache the result"""
55
62
return _utf8_to_uni[utf8_str]
57
_utf8_to_uni[utf8_str] = unicode_str = utf8_str.decode('utf-8')
64
unicode_str = _utf8_decode(utf8_str)[0]
65
_utf8_to_uni[utf8_str] = unicode_str
58
66
_uni_to_utf8[unicode_str] = utf8_str
71
79
return decode(encode(unicode_str))
82
def get_cached_utf8(utf8_str):
83
"""Return a cached version of the utf-8 string.
85
Get a cached version of this string (similar to intern()).
86
At present, this will be decoded to ensure it is a utf-8 string. In the
87
future this might change to simply caching the string.
89
return encode(decode(utf8_str))
92
def get_cached_ascii(ascii_str,
93
_uni_to_utf8=_unicode_to_utf8_map,
94
_utf8_to_uni=_utf8_to_unicode_map):
95
"""This is a string which is identical in utf-8 and unicode."""
96
# We don't need to do any encoding, but we want _utf8_to_uni to return a
97
# real Unicode string. Unicode and plain strings of this type will have the
98
# same hash, so we can just use it as the key in _uni_to_utf8, but we need
99
# the return value to be different in _utf8_to_uni
100
ascii_str = _uni_to_utf8.setdefault(ascii_str, ascii_str)
101
_utf8_to_uni.setdefault(ascii_str, unicode(ascii_str))
74
105
def clear_encoding_cache():
75
106
"""Clear the encoding and decoding caches"""
76
global _unicode_to_utf8_map, _utf8_to_unicode_map
77
107
_unicode_to_utf8_map.clear()
78
108
_utf8_to_unicode_map.clear()
added
removed
bzrlib/cache_utf8.py
