##// END OF EJS Templates
win32mbcs: use extsetup() to wrap functions only once....
Shun-ichi GOTO -
r13067:96969544 stable
parent child Browse files
Show More
@@ -1,154 +1,159 b''
1 1 # win32mbcs.py -- MBCS filename support for Mercurial
2 2 #
3 3 # Copyright (c) 2008 Shun-ichi Goto <shunichi.goto@gmail.com>
4 4 #
5 5 # Version: 0.3
6 6 # Author: Shun-ichi Goto <shunichi.goto@gmail.com>
7 7 #
8 8 # This software may be used and distributed according to the terms of the
9 9 # GNU General Public License version 2 or any later version.
10 10 #
11 11
12 12 '''allow the use of MBCS paths with problematic encodings
13 13
14 14 Some MBCS encodings are not good for some path operations (i.e.
15 15 splitting path, case conversion, etc.) with its encoded bytes. We call
16 16 such a encoding (i.e. shift_jis and big5) as "problematic encoding".
17 17 This extension can be used to fix the issue with those encodings by
18 18 wrapping some functions to convert to Unicode string before path
19 19 operation.
20 20
21 21 This extension is useful for:
22 22
23 23 - Japanese Windows users using shift_jis encoding.
24 24 - Chinese Windows users using big5 encoding.
25 25 - All users who use a repository with one of problematic encodings on
26 26 case-insensitive file system.
27 27
28 28 This extension is not needed for:
29 29
30 30 - Any user who use only ASCII chars in path.
31 31 - Any user who do not use any of problematic encodings.
32 32
33 33 Note that there are some limitations on using this extension:
34 34
35 35 - You should use single encoding in one repository.
36
36 - If the repository path ends with 0x5c, .hg/hgrc cannot be read.
37 - win32mbcs is not compatible with fixutf8 extention.
37 38
38 39 By default, win32mbcs uses encoding.encoding decided by Mercurial.
39 40 You can specify the encoding by config option::
40 41
41 42 [win32mbcs]
42 43 encoding = sjis
43 44
44 45 It is useful for the users who want to commit with UTF-8 log message.
45 46 '''
46 47
47 48 import os, sys
48 49 from mercurial.i18n import _
49 50 from mercurial import util, encoding
50 51
51 _encoding = None # see reposetup()
52 _encoding = None # see extsetup
52 53
53 54 def decode(arg):
54 55 if isinstance(arg, str):
55 56 uarg = arg.decode(_encoding)
56 57 if arg == uarg.encode(_encoding):
57 58 return uarg
58 59 raise UnicodeError("Not local encoding")
59 60 elif isinstance(arg, tuple):
60 61 return tuple(map(decode, arg))
61 62 elif isinstance(arg, list):
62 63 return map(decode, arg)
63 64 elif isinstance(arg, dict):
64 65 for k, v in arg.items():
65 66 arg[k] = decode(v)
66 67 return arg
67 68
68 69 def encode(arg):
69 70 if isinstance(arg, unicode):
70 71 return arg.encode(_encoding)
71 72 elif isinstance(arg, tuple):
72 73 return tuple(map(encode, arg))
73 74 elif isinstance(arg, list):
74 75 return map(encode, arg)
75 76 elif isinstance(arg, dict):
76 77 for k, v in arg.items():
77 78 arg[k] = encode(v)
78 79 return arg
79 80
80 81 def appendsep(s):
81 82 # ensure the path ends with os.sep, appending it if necessary.
82 83 try:
83 84 us = decode(s)
84 85 except UnicodeError:
85 86 us = s
86 87 if us and us[-1] not in ':/\\':
87 88 s += os.sep
88 89 return s
89 90
90 91 def wrapper(func, args, kwds):
91 92 # check argument is unicode, then call original
92 93 for arg in args:
93 94 if isinstance(arg, unicode):
94 95 return func(*args, **kwds)
95 96
96 97 try:
97 98 # convert arguments to unicode, call func, then convert back
98 99 return encode(func(*decode(args), **decode(kwds)))
99 100 except UnicodeError:
100 101 raise util.Abort(_("[win32mbcs] filename conversion failed with"
101 102 " %s encoding\n") % (_encoding))
102 103
103 104 def wrapperforlistdir(func, args, kwds):
104 105 # Ensure 'path' argument ends with os.sep to avoids
105 106 # misinterpreting last 0x5c of MBCS 2nd byte as path separator.
106 107 if args:
107 108 args = list(args)
108 109 args[0] = appendsep(args[0])
109 110 if 'path' in kwds:
110 111 kwds['path'] = appendsep(kwds['path'])
111 112 return func(*args, **kwds)
112 113
113 114 def wrapname(name, wrapper):
114 115 module, name = name.rsplit('.', 1)
115 116 module = sys.modules[module]
116 117 func = getattr(module, name)
117 118 def f(*args, **kwds):
118 119 return wrapper(func, args, kwds)
119 120 try:
120 121 f.__name__ = func.__name__ # fail with python23
121 122 except Exception:
122 123 pass
123 124 setattr(module, name, f)
124 125
125 126 # List of functions to be wrapped.
126 127 # NOTE: os.path.dirname() and os.path.basename() are safe because
127 128 # they use result of os.path.split()
128 129 funcs = '''os.path.join os.path.split os.path.splitext
129 130 os.path.splitunc os.path.normpath os.path.normcase os.makedirs
130 131 mercurial.util.endswithsep mercurial.util.splitpath mercurial.util.checkcase
131 132 mercurial.util.fspath mercurial.util.pconvert mercurial.util.normpath'''
132 133
133 134 # codec and alias names of sjis and big5 to be faked.
134 135 problematic_encodings = '''big5 big5-tw csbig5 big5hkscs big5-hkscs
135 136 hkscs cp932 932 ms932 mskanji ms-kanji shift_jis csshiftjis shiftjis
136 137 sjis s_jis shift_jis_2004 shiftjis2004 sjis_2004 sjis2004
137 138 shift_jisx0213 shiftjisx0213 sjisx0213 s_jisx0213 950 cp950 ms950 '''
138 139
139 def reposetup(ui, repo):
140 def extsetup(ui):
140 141 # TODO: decide use of config section for this extension
141 142 if not os.path.supports_unicode_filenames:
142 143 ui.warn(_("[win32mbcs] cannot activate on this platform.\n"))
143 144 return
144 145 # determine encoding for filename
145 146 global _encoding
146 147 _encoding = ui.config('win32mbcs', 'encoding', encoding.encoding)
147 148 # fake is only for relevant environment.
148 149 if _encoding.lower() in problematic_encodings.split():
149 150 for f in funcs.split():
150 151 wrapname(f, wrapper)
151 152 wrapname("mercurial.osutil.listdir", wrapperforlistdir)
152 ui.debug("[win32mbcs] activated with encoding: %s\n"
153 % _encoding)
153 # Check sys.args manually instead of using ui.debug() because
154 # command line options is not yet applied when
155 # extensions.loadall() is called.
156 if '--debug' in sys.argv:
157 ui.write("[win32mbcs] activated with encoding: %s\n"
158 % _encoding)
154 159
General Comments 0
You need to be logged in to leave comments. Login now