1
# Copyright (C) 2006, 2008 Canonical Ltd
1
# Bazaar -- distributed version control
3
# Copyright (C) 2006 Canonical Ltd
3
5
# This program is free software; you can redistribute it and/or modify
4
6
# it under the terms of the GNU General Public License as published by
13
15
# You should have received a copy of the GNU General Public License
14
16
# along with this program; if not, write to the Free Software
15
# Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
17
# Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
17
19
"""A collection of function for handling URL operations."""
66
68
relpath = relpath.encode('utf-8')
67
69
# After quoting and encoding, the path should be perfectly
68
70
# safe as a plain ASCII string, str() just enforces this
69
return str(urllib.quote(relpath, safe='/~'))
71
return str(urllib.quote(relpath))
72
74
def file_relpath(base, path):
73
75
"""Compute just the relative sub-portion of a url
75
77
This assumes that both paths are already fully specified file:// URLs.
77
79
if len(base) < MIN_ABS_FILEURL_LENGTH:
78
raise ValueError('Length of base (%r) must equal or'
80
raise ValueError('Length of base must be equal or'
79
81
' exceed the platform minimum url length (which is %d)' %
80
(base, MIN_ABS_FILEURL_LENGTH))
82
MIN_ABS_FILEURL_LENGTH)
81
83
base = local_path_from_url(base)
82
84
path = local_path_from_url(path)
83
85
return escape(osutils.relpath(base, path))
184
186
def joinpath(base, *args):
185
187
"""Join URL path segments to a URL path segment.
187
189
This is somewhat like osutils.joinpath, but intended for URLs.
189
191
XXX: this duplicates some normalisation logic, and also duplicates a lot of
217
219
# jam 20060502 Sorted to 'l' because the final target is 'local_path_from_url'
218
220
def _posix_local_path_from_url(url):
219
221
"""Convert a url like file:///path/to/foo into /path/to/foo"""
220
file_localhost_prefix = 'file://localhost/'
221
if url.startswith(file_localhost_prefix):
222
path = url[len(file_localhost_prefix) - 1:]
223
elif not url.startswith('file:///'):
224
raise errors.InvalidURL(
225
url, 'local urls must start with file:/// or file://localhost/')
227
path = url[len('file://'):]
222
if not url.startswith('file:///'):
223
raise errors.InvalidURL(url, 'local urls must start with file:///')
228
224
# We only strip off 2 slashes
229
return unescape(path)
225
return unescape(url[len('file://'):])
232
228
def _posix_local_path_to_url(path):
235
231
This also handles transforming escaping unicode characters, etc.
237
# importing directly from posixpath allows us to test this
233
# importing directly from posixpath allows us to test this
238
234
# on non-posix platforms
239
235
return 'file://' + escape(_posix_normpath(
240
236
osutils._posix_abspath(path)))
275
271
This also handles transforming escaping unicode characters, etc.
277
# importing directly from ntpath allows us to test this
273
# importing directly from ntpath allows us to test this
278
274
# on non-win32 platform
279
275
# FIXME: It turns out that on nt, ntpath.abspath uses nt._getfullpathname
280
276
# which actually strips trailing space characters.
374
370
dummy, base_first_slash = _find_scheme_and_separator(base)
375
371
if base_first_slash is None:
378
374
dummy, other_first_slash = _find_scheme_and_separator(other)
379
375
if other_first_slash is None:
422
418
# Strip off the drive letter
423
419
# path is currently /C:/foo
424
420
if len(path) < 3 or path[2] not in ':|' or path[3] != '/':
425
raise errors.InvalidURL(url_base + path,
421
raise errors.InvalidURL(url_base + path,
426
422
'win32 file:/// paths need a drive letter')
427
423
url_base += path[0:3] # file:// + /C:
428
424
path = path[3:] # /foo
436
432
:param exclude_trailing_slash: Strip off a final '/' if it is part
437
433
of the path (but not if it is part of the protocol specification)
439
:return: (parent_url, child_dir). child_dir may be the empty string if we're at
435
:return: (parent_url, child_dir). child_dir may be the empty string if we're at
442
438
scheme_loc, first_path_slash = _find_scheme_and_separator(url)
546
542
# These are characters that if escaped, should stay that way
547
543
_no_decode_chars = ';/?:@&=+$,#'
548
544
_no_decode_ords = [ord(c) for c in _no_decode_chars]
549
_no_decode_hex = (['%02x' % o for o in _no_decode_ords]
545
_no_decode_hex = (['%02x' % o for o in _no_decode_ords]
550
546
+ ['%02X' % o for o in _no_decode_ords])
551
547
_hex_display_map = dict(([('%02x' % o, chr(o)) for o in range(256)]
552
548
+ [('%02X' % o, chr(o)) for o in range(256)]))
578
574
This will turn file:// urls into local paths, and try to decode
579
575
any portions of a http:// style url that it can.
581
Any sections of the URL which can't be represented in the encoding or
577
Any sections of the URL which can't be represented in the encoding or
582
578
need to stay as escapes are left alone.
584
580
:param url: A 7-bit ASCII URL
585
581
:param encoding: The final output encoding
587
:return: A unicode string which can be safely encoded into the
583
:return: A unicode string which can be safely encoded into the
588
584
specified encoding.
590
586
if encoding is None:
691
687
if len(segments) == 0:
693
689
return osutils.pathjoin(*segments)
698
"""Extract the server address, the credentials and the path from the url.
700
user, password, host and path should be quoted if they contain reserved
703
:param url: an quoted url
705
:return: (scheme, user, password, host, port, path) tuple, all fields
708
if isinstance(url, unicode):
709
raise errors.InvalidURL('should be ascii:\n%r' % url)
710
url = url.encode('utf-8')
711
(scheme, netloc, path, params,
712
query, fragment) = urlparse.urlparse(url, allow_fragments=False)
713
user = password = host = port = None
715
user, host = netloc.rsplit('@', 1)
717
user, password = user.split(':', 1)
718
password = urllib.unquote(password)
719
user = urllib.unquote(user)
723
if ':' in host and not (host[0] == '[' and host[-1] == ']'): #there *is* port
724
host, port = host.rsplit(':',1)
728
raise errors.InvalidURL('invalid port number %s in url:\n%s' %
730
if host != "" and host[0] == '[' and host[-1] == ']': #IPv6
733
host = urllib.unquote(host)
734
path = urllib.unquote(path)
736
return (scheme, user, password, host, port, path)