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
229
231
This also handles transforming escaping unicode characters, etc.
231
# importing directly from posixpath allows us to test this
233
# importing directly from posixpath allows us to test this
232
234
# on non-posix platforms
233
235
return 'file://' + escape(_posix_normpath(
234
236
osutils._posix_abspath(path)))
269
271
This also handles transforming escaping unicode characters, etc.
271
# importing directly from ntpath allows us to test this
273
# importing directly from ntpath allows us to test this
272
274
# on non-win32 platform
273
275
# FIXME: It turns out that on nt, ntpath.abspath uses nt._getfullpathname
274
276
# which actually strips trailing space characters.
368
370
dummy, base_first_slash = _find_scheme_and_separator(base)
369
371
if base_first_slash is None:
372
374
dummy, other_first_slash = _find_scheme_and_separator(other)
373
375
if other_first_slash is None:
416
418
# Strip off the drive letter
417
419
# path is currently /C:/foo
418
420
if len(path) < 3 or path[2] not in ':|' or path[3] != '/':
419
raise errors.InvalidURL(url_base + path,
421
raise errors.InvalidURL(url_base + path,
420
422
'win32 file:/// paths need a drive letter')
421
423
url_base += path[0:3] # file:// + /C:
422
424
path = path[3:] # /foo
430
432
:param exclude_trailing_slash: Strip off a final '/' if it is part
431
433
of the path (but not if it is part of the protocol specification)
433
: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
436
438
scheme_loc, first_path_slash = _find_scheme_and_separator(url)
540
542
# These are characters that if escaped, should stay that way
541
543
_no_decode_chars = ';/?:@&=+$,#'
542
544
_no_decode_ords = [ord(c) for c in _no_decode_chars]
543
_no_decode_hex = (['%02x' % o for o in _no_decode_ords]
545
_no_decode_hex = (['%02x' % o for o in _no_decode_ords]
544
546
+ ['%02X' % o for o in _no_decode_ords])
545
547
_hex_display_map = dict(([('%02x' % o, chr(o)) for o in range(256)]
546
548
+ [('%02X' % o, chr(o)) for o in range(256)]))
572
574
This will turn file:// urls into local paths, and try to decode
573
575
any portions of a http:// style url that it can.
575
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
576
578
need to stay as escapes are left alone.
578
580
:param url: A 7-bit ASCII URL
579
581
:param encoding: The final output encoding
581
:return: A unicode string which can be safely encoded into the
583
:return: A unicode string which can be safely encoded into the
582
584
specified encoding.
584
586
if encoding is None:
685
687
if len(segments) == 0:
687
689
return osutils.pathjoin(*segments)
692
"""Extract the server address, the credentials and the path from the url.
694
user, password, host and path should be quoted if they contain reserved
697
:param url: an quoted url
699
:return: (scheme, user, password, host, port, path) tuple, all fields
702
if isinstance(url, unicode):
703
raise errors.InvalidURL('should be ascii:\n%r' % url)
704
url = url.encode('utf-8')
705
(scheme, netloc, path, params,
706
query, fragment) = urlparse.urlparse(url, allow_fragments=False)
707
user = password = host = port = None
709
user, host = netloc.rsplit('@', 1)
711
user, password = user.split(':', 1)
712
password = urllib.unquote(password)
713
user = urllib.unquote(user)
717
if ':' in host and not (host[0] == '[' and host[-1] == ']'): #there *is* port
718
host, port = host.rsplit(':',1)
722
raise errors.InvalidURL('invalid port number %s in url:\n%s' %
724
if host != "" and host[0] == '[' and host[-1] == ']': #IPv6
727
host = urllib.unquote(host)
728
path = urllib.unquote(path)
730
return (scheme, user, password, host, port, path)