« back to all changes in this revision
Viewing changes to bzrlib/urlutils.py
- browse files at revision 1685.1.45
- compare with another revision
- compare with revision 5200.3.3
- download diff
- download tarball
- view history from revision 1685.1.45
- Committer: John Arbash Meinel
- Date: 2006-05-02 20:46:11 UTC
- mto: This revision was merged to the branch mainline in revision 1752.
- Revision ID: john@arbash-meinel.com-20060502204611-02caa5c20fb84ef8
Moved url functions into bzrlib.urlutils
- files added:
- bzrlib/tests/test_urlutils.py
- files modified:
- bzrlib/builtins.py
added
removed
bzrlib/urlutils.py
1
# Bazaar-NG -- distributed version control
2
#
3
# Copyright (C) 2006 by Canonical Ltd
4
#
5
# This program is free software; you can redistribute it and/or modify
6
# it under the terms of the GNU General Public License as published by
7
# the Free Software Foundation; either version 2 of the License, or
8
# (at your option) any later version.
9
#
10
# This program is distributed in the hope that it will be useful,
11
# but WITHOUT ANY WARRANTY; without even the implied warranty of
12
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
13
# GNU General Public License for more details.
14
#
15
# You should have received a copy of the GNU General Public License
16
# along with this program; if not, write to the Free Software
17
# Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
18
19
"""A collection of function for handling URL operations."""
20
21
import urllib
22
import sys
23
24
import bzrlib.errors as errors
25
import bzrlib.osutils
26
27
28
def escape(relpath):
29
"""Escape relpath to be a valid url."""
30
if isinstance(relpath, unicode):
31
relpath = relpath.encode('utf-8')
32
# After quoting and encoding, the path should be perfectly
33
# safe as a plain ASCII string, str() just enforces this
34
return str(urllib.quote(relpath))
35
36
37
def unescape(url):
38
"""Unescape relpath from url format.
39
40
This returns a Unicode path from a URL
41
"""
42
# jam 20060427 URLs are supposed to be ASCII only strings
43
# If they are passed in as unicode, urllib.unquote
44
# will return a UNICODE string, which actually contains
45
# utf-8 bytes. So we have to ensure that they are
46
# plain ASCII strings, or the final .decode will
47
# try to encode the UNICODE => ASCII, and then decode
48
# it into utf-8.
49
try:
50
url = str(url)
51
except UnicodeError, e:
52
raise errors.InvalidURL(url, 'URL was not a plain ASCII url: %s' % (e,))
53
unquoted = urllib.unquote(url)
54
try:
55
unicode_path = unquoted.decode('utf-8')
56
except UnicodeError, e:
57
raise errors.InvalidURL(url, 'Unable to encode the URL as utf-8: %s' % (e,))
58
return unicode_path
59
60
61
def file_relpath(base, path):
62
"""Compute just the relative sub-portion of a url
63
64
This assumes that both paths are already fully specified file:// URLs.
65
"""
66
assert len(base) >= MIN_ABS_URLPATHLENGTH, ('Length of base must be equal or'
67
' exceed the platform minimum url length (which is %d)' %
68
MIN_ABS_URLPATHLENGTH)
69
70
base = local_path_from_url(base)
71
path = local_path_from_url(path)
72
return escape(bzrlib.osutils.relpath(base, path))
73
74
75
def strip_trailing_slash(url):
76
"""Strip trailing slash, except for root paths.
77
78
The definition of 'root path' is platform-dependent.
79
But the passed in URL must be a file:/// url.
80
"""
81
assert url.startswith('file:///'), \
82
'strip_trailing_slash expects file:// urls (%s)' % url
83
if len(url) != MIN_ABS_URLPATHLENGTH and url[-1] == '/':
84
return url[:-1]
85
else:
86
return url
87
88
89
# These are characters that if escaped, should stay that way
90
_no_decode_chars = ';/?:@&=+$,#'
91
_no_decode_ords = [ord(c) for c in _no_decode_chars]
92
_no_decode_hex = (['%02x' % o for o in _no_decode_ords]
93
+ ['%02X' % o for o in _no_decode_ords])
94
_hex_display_map = urllib._hextochr.copy()
95
_hex_display_map.update((hex,'%'+hex) for hex in _no_decode_hex)
96
#These entries get mapped to themselves
97
98
99
def unescape_for_display(url):
100
"""Decode what you can for a URL, so that we get a nice looking path.
101
102
This will turn file:// urls into local paths, and try to decode
103
any portions of a http:// style url that it can.
104
"""
105
if url.startswith('file://'):
106
return local_path_from_url(url)
107
108
# Split into sections to try to decode utf-8
109
res = url.split('/')
110
for i in xrange(1, len(res)):
111
escaped_chunks = res[i].split('%')
112
for j in xrange(1, len(escaped_chunks)):
113
item = escaped_chunks[j]
114
try:
115
escaped_chunks[j] = _hex_display_map[item[:2]] + item[2:]
116
except KeyError:
117
# Put back the percent symbol
118
escaped_chunks[j] = '%' + item
119
except UnicodeDecodeError:
120
escaped_chunks[j] = unichr(int(item[:2], 16)) + item[2:]
121
unescaped = ''.join(escaped_chunks)
122
try:
123
res[i] = unescaped.decode('utf-8')
124
except UnicodeDecodeError:
125
# If this path segment cannot be properly utf-8 decoded
126
# after doing unescaping we will just leave it alone
127
pass
128
return '/'.join(res)
129
130
def _posix_local_path_to_url(path):
131
"""Convert a local path like ./foo into a URL like file:///path/to/foo
132
133
This also handles transforming escaping unicode characters, etc.
134
"""
135
# importing directly from posixpath allows us to test this
136
# on non-posix platforms
137
from posixpath import normpath
138
return 'file://' + escape(normpath(bzrlib.osutils._posix_abspath(path)))
139
140
141
def _posix_local_path_from_url(url):
142
"""Convert a url like file:///path/to/foo into /path/to/foo"""
143
if not url.startswith('file:///'):
144
raise errors.InvalidURL(url, 'local urls must start with file:///')
145
# We only strip off 2 slashes
146
return unescape(url[len('file://'):])
147
148
149
def _win32_local_path_to_url(path):
150
"""Convert a local path like ./foo into a URL like file:///C|/path/to/foo
151
152
This also handles transforming escaping unicode characters, etc.
153
"""
154
# importing directly from ntpath allows us to test this
155
# on non-win32 platforms
156
# TODO: jam 20060426 consider moving this import outside of the function
157
win32_path = bzrlib.osutils._nt_normpath(
158
bzrlib.osutils._win32_abspath(path)).replace('\\', '/')
159
return 'file:///' + win32_path[0].upper() + '|' + escape(win32_path[2:])
160
161
162
def _win32_local_path_from_url(url):
163
"""Convert a url like file:///C|/path/to/foo into C:/path/to/foo"""
164
if not url.startswith('file:///'):
165
raise errors.InvalidURL(url, 'local urls must start with file:///')
166
# We strip off all 3 slashes
167
win32_url = url[len('file:///'):]
168
if (win32_url[0] not in 'abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ'
169
or win32_url[1] not in '|:'
170
or win32_url[2] != '/'):
171
raise errors.InvalidURL(url, 'Win32 file urls start with file:///X|/, where X is a valid drive letter')
172
# TODO: jam 20060426, we could .upper() or .lower() the drive letter
173
# for better consistency.
174
return win32_url[0].upper() + u':' + unescape(win32_url[2:])
175
176
177
local_path_to_url = _posix_local_path_to_url
178
local_path_from_url = _posix_local_path_from_url
179
MIN_ABS_URLPATHLENGTH = len('file:///')
180
181
if sys.platform == 'win32':
182
local_path_to_url = _win32_local_path_to_url
183
local_path_from_url = _win32_local_path_from_url
184
185
MIN_ABS_URLPATHLENGTH = len('file:///C|/')
Loggerhead is a web-based interface for Breezy
Version: 2.0.1