77
75
This assumes that both paths are already fully specified file:// URLs.
79
77
if len(base) < MIN_ABS_FILEURL_LENGTH:
80
raise ValueError('Length of base must be equal or'
78
raise ValueError('Length of base (%r) must equal or'
81
79
' exceed the platform minimum url length (which is %d)' %
82
MIN_ABS_FILEURL_LENGTH)
80
(base, MIN_ABS_FILEURL_LENGTH))
83
81
base = local_path_from_url(base)
84
82
path = local_path_from_url(path)
85
83
return escape(osutils.relpath(base, path))
123
match = _url_scheme_re.match(base)
126
scheme = match.group('scheme')
127
path = match.group('path').split('/')
128
if path[-1:] == ['']:
129
# Strip off a trailing slash
130
# This helps both when we are at the root, and when
131
# 'base' has an extra slash at the end
134
path = base.split('/')
136
if scheme is not None and len(path) >= 1:
138
# the path should be represented as an abs path.
139
# we know this must be absolute because of the presence of a URL scheme.
141
path = [''] + path[1:]
143
# create an empty host, but dont alter the path - this might be a
144
# relative url fragment.
126
scheme_end, path_start = _find_scheme_and_separator(base)
127
if scheme_end is None and path_start is None:
129
elif path_start is None:
130
path_start = len(base)
131
path = base[path_start:]
149
match = _url_scheme_re.match(arg)
152
scheme = match.group('scheme')
153
# this skips .. normalisation, making http://host/../../..
155
path = match.group('path').split('/')
156
# set the host and path according to new absolute URL, discarding
157
# any previous values.
158
# XXX: duplicates mess from earlier in this function. This URL
159
# manipulation code needs some cleaning up.
160
if scheme is not None and len(path) >= 1:
163
# url scheme implies absolute path.
166
# no url scheme we take the path as is.
133
arg_scheme_end, arg_path_start = _find_scheme_and_separator(arg)
134
if arg_scheme_end is None and arg_path_start is None:
136
elif arg_path_start is None:
137
arg_path_start = len(arg)
138
if arg_scheme_end is not None:
140
path = arg[arg_path_start:]
141
scheme_end = arg_scheme_end
142
path_start = arg_path_start
169
path = '/'.join(path)
170
144
path = joinpath(path, arg)
171
path = path.split('/')
172
if remove_root and path[0:1] == ['']:
175
# Remove the leading slash from the path, so long as it isn't also the
176
# trailing slash, which we want to keep if present.
177
if path and path[0] == '' and len(path) > 1:
182
return '/'.join(path)
183
return scheme + '://' + '/'.join(path)
145
return base[:path_start] + path
186
148
def joinpath(base, *args):
219
181
# jam 20060502 Sorted to 'l' because the final target is 'local_path_from_url'
220
182
def _posix_local_path_from_url(url):
221
183
"""Convert a url like file:///path/to/foo into /path/to/foo"""
222
if not url.startswith('file:///'):
223
raise errors.InvalidURL(url, 'local urls must start with file:///')
184
file_localhost_prefix = 'file://localhost/'
185
if url.startswith(file_localhost_prefix):
186
path = url[len(file_localhost_prefix) - 1:]
187
elif not url.startswith('file:///'):
188
raise errors.InvalidURL(
189
url, 'local urls must start with file:/// or file://localhost/')
191
path = url[len('file://'):]
224
192
# We only strip off 2 slashes
225
return unescape(url[len('file://'):])
193
return unescape(path)
228
196
def _posix_local_path_to_url(path):
335
303
:param url: Either a hybrid URL or a local path
336
304
:return: A normalized URL which only includes 7-bit ASCII characters.
338
m = _url_scheme_re.match(url)
306
scheme_end, path_start = _find_scheme_and_separator(url)
307
if scheme_end is None:
340
308
return local_path_to_url(url)
341
scheme = m.group('scheme')
342
path = m.group('path')
309
prefix = url[:path_start]
310
path = url[path_start:]
343
311
if not isinstance(url, unicode):
345
313
if c not in _url_safe_characters:
346
314
raise errors.InvalidURL(url, 'URLs can only contain specific'
347
315
' safe characters (not %r)' % c)
348
316
path = _url_hex_escapes_re.sub(_unescape_safe_chars, path)
349
return str(scheme + '://' + ''.join(path))
317
return str(prefix + ''.join(path))
351
319
# We have a unicode (hybrid) url
352
320
path_chars = list(path)
465
433
return url_base + head, tail
436
def split_segment_parameters_raw(url):
437
"""Split the subsegment of the last segment of a URL.
439
:param url: A relative or absolute URL
440
:return: (url, subsegments)
442
(parent_url, child_dir) = split(url)
443
subsegments = child_dir.split(",")
444
if len(subsegments) == 1:
446
return (join(parent_url, subsegments[0]), subsegments[1:])
449
def split_segment_parameters(url):
450
"""Split the segment parameters of the last segment of a URL.
452
:param url: A relative or absolute URL
453
:return: (url, segment_parameters)
455
(base_url, subsegments) = split_segment_parameters_raw(url)
457
for subsegment in subsegments:
458
(key, value) = subsegment.split("=", 1)
459
parameters[key] = value
460
return (base_url, parameters)
463
def join_segment_parameters_raw(base, *subsegments):
464
"""Create a new URL by adding subsegments to an existing one.
466
This adds the specified subsegments to the last path in the specified
467
base URL. The subsegments should be bytestrings.
469
:note: You probably want to use join_segment_parameters instead.
473
for subsegment in subsegments:
474
if type(subsegment) is not str:
475
raise TypeError("Subsegment %r is not a bytestring" % subsegment)
476
if "," in subsegment:
477
raise errors.InvalidURLJoin(", exists in subsegments",
479
return ",".join((base,) + subsegments)
482
def join_segment_parameters(url, parameters):
483
"""Create a new URL by adding segment parameters to an existing one.
485
The parameters of the last segment in the URL will be updated; if a
486
parameter with the same key already exists it will be overwritten.
488
:param url: A URL, as string
489
:param parameters: Dictionary of parameters, keys and values as bytestrings
491
(base, existing_parameters) = split_segment_parameters(url)
493
new_parameters.update(existing_parameters)
494
for key, value in parameters.iteritems():
495
if type(key) is not str:
496
raise TypeError("parameter key %r is not a bytestring" % key)
497
if type(value) is not str:
498
raise TypeError("parameter value %r for %s is not a bytestring" %
501
raise errors.InvalidURLJoin("= exists in parameter key", url,
503
new_parameters[key] = value
504
return join_segment_parameters_raw(base,
505
*["%s=%s" % item for item in sorted(new_parameters.items())])
468
508
def _win32_strip_local_trailing_slash(url):
469
509
"""Strip slashes after the drive letter"""
470
510
if len(url) > WIN32_MIN_ABS_FILEURL_LENGTH:
687
727
if len(segments) == 0:
689
729
return osutils.pathjoin(*segments)
734
"""Extract the server address, the credentials and the path from the url.
736
user, password, host and path should be quoted if they contain reserved
739
:param url: an quoted url
741
:return: (scheme, user, password, host, port, path) tuple, all fields
744
if isinstance(url, unicode):
745
raise errors.InvalidURL('should be ascii:\n%r' % url)
746
url = url.encode('utf-8')
747
(scheme, netloc, path, params,
748
query, fragment) = urlparse.urlparse(url, allow_fragments=False)
749
user = password = host = port = None
751
user, host = netloc.rsplit('@', 1)
753
user, password = user.split(':', 1)
754
password = urllib.unquote(password)
755
user = urllib.unquote(user)
759
if ':' in host and not (host[0] == '[' and host[-1] == ']'): #there *is* port
760
host, port = host.rsplit(':',1)
764
raise errors.InvalidURL('invalid port number %s in url:\n%s' %
766
if host != "" and host[0] == '[' and host[-1] == ']': #IPv6
769
host = urllib.unquote(host)
770
path = urllib.unquote(path)
772
return (scheme, user, password, host, port, path)