Mercurial > public > mercurial-scm > hg-stable

comparison mercurial/util.py @ 3835:d1ce5461beed

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
Allow the user to specify the fallback encoding for the changelog Example: use EUC-JP instead of ISO-8859-1: [ui] fallbackencoding = EUC-JP
author Alexis S. L. Carvalho <alexis@cecm.usp.br>
date Fri, 08 Dec 2006 22:01:05 -0200
parents fc5ba0ab7f45
children abaa2cd00d2b
comparison
equal deleted inserted replaced
3834:a7b61c3b0f93 3835:d1ce5461beed
17 demandload(globals(), "cStringIO errno getpass popen2 re shutil sys tempfile") 17 demandload(globals(), "cStringIO errno getpass popen2 re shutil sys tempfile")
18 demandload(globals(), "os threading time calendar ConfigParser locale") 18 demandload(globals(), "os threading time calendar ConfigParser locale")
19 19
20 _encoding = os.environ.get("HGENCODING") or locale.getpreferredencoding() 20 _encoding = os.environ.get("HGENCODING") or locale.getpreferredencoding()
21 _encodingmode = os.environ.get("HGENCODINGMODE", "strict") 21 _encodingmode = os.environ.get("HGENCODINGMODE", "strict")
22 _fallbackencoding = 'ISO-8859-1'
22 23
23 def tolocal(s): 24 def tolocal(s):
24 """ 25 """
25 Convert a string from internal UTF-8 to local encoding 26 Convert a string from internal UTF-8 to local encoding
26 27
28 implementation of locale support may contain latin1 or possibly 29 implementation of locale support may contain latin1 or possibly
29 other character sets. We attempt to decode everything strictly 30 other character sets. We attempt to decode everything strictly
30 using UTF-8, then Latin-1, and failing that, we use UTF-8 and 31 using UTF-8, then Latin-1, and failing that, we use UTF-8 and
31 replace unknown characters. 32 replace unknown characters.
32 """ 33 """
33 for e in "utf-8 latin1".split(): 34 for e in ('UTF-8', _fallbackencoding):
34 try: 35 try:
35 u = s.decode(e) # attempt strict decoding 36 u = s.decode(e) # attempt strict decoding
36 return u.encode(_encoding, "replace") 37 return u.encode(_encoding, "replace")
37 except UnicodeDecodeError: 38 except UnicodeDecodeError:
38 pass 39 pass