def normalize_url(self, url):
# type: (str) -> str
"""
Normalize a URL for comparison by unquoting it and removing any
trailing slash.
"""
return urllib_parse.unquote(url).rstrip('/')
def filename(self):
# type: () -> str
_, netloc, path, _, _ = urllib_parse.urlsplit(self.url)
name = posixpath.basename(path.rstrip('/')) or netloc
name = urllib_parse.unquote(name)
assert name, ('URL %r produced no filename' % self.url)
return name
def filename(self):
# type: () -> str
path = self.path.rstrip('/')
name = posixpath.basename(path)
if not name:
# Make sure we don't leak auth information if the netloc
# includes a username and password.
netloc, user_pass = split_auth_from_netloc(self.netloc)
return netloc
name = urllib_parse.unquote(name)
assert name, ('URL %r produced no filename' % self._url)
return name
def get_file_content(url, comes_from=None, session=None):
# type: (str, Optional[str], Optional[PipSession]) -> Tuple[str, Text]
"""Gets the content of a file; it may be a filename, file: URL, or
http: URL. Returns (location, content). Content is unicode.
:param url: File path or url.
:param comes_from: Origin description of requirements.
:param session: Instance of pip.download.PipSession.
"""
if session is None:
raise TypeError(
"get_file_content() missing 1 required keyword argument: 'session'"
)
scheme = get_url_scheme(url)
if scheme in ['http', 'https']:
# FIXME: catch some errors
resp = session.get(url)
resp.raise_for_status()
return resp.url, resp.text
elif scheme == 'file':
if comes_from and comes_from.startswith('http'):
raise InstallationError(
'Requirements file %s references URL %s, which is local' %
(comes_from, url))
path = url.split(':', 1)[1]
path = path.replace('\\', '/')
match = _url_slash_drive_re.match(path)
if match:
path = match.group(1) + ':' + path.split('|', 1)[1]
path = urllib_parse.unquote(path)
if path.startswith('/'):
path = '/' + path.lstrip('/')
url = path
try:
with open(url, 'rb') as f:
content = auto_decode(f.read())
except IOError as exc:
raise InstallationError('Could not open requirements file: %s' %
str(exc))
return url, content
def get_file_content(url, comes_from=None, session=None):
"""Gets the content of a file; it may be a filename, file: URL, or
http: URL. Returns (location, content). Content is unicode.
:param url: File path or url.
:param comes_from: Origin description of requirements.
:param session: Instance of pip.download.PipSession.
"""
if session is None:
raise TypeError(
"get_file_content() missing 1 required keyword argument: 'session'"
)
match = _scheme_re.search(url)
if match:
scheme = match.group(1).lower()
if (scheme == 'file' and comes_from and
comes_from.startswith('http')):
raise InstallationError(
'Requirements file %s references URL %s, which is local'
% (comes_from, url))
if scheme == 'file':
path = url.split(':', 1)[1]
path = path.replace('\\', '/')
match = _url_slash_drive_re.match(path)
if match:
path = match.group(1) + ':' + path.split('|', 1)[1]
path = urllib_parse.unquote(path)
if path.startswith('/'):
path = '/' + path.lstrip('/')
url = path
else:
# FIXME: catch some errors
resp = session.get(url)
resp.raise_for_status()
return resp.url, resp.text
try:
with open(url, 'rb') as f:
content = auto_decode(f.read())
except IOError as exc:
raise InstallationError(
'Could not open requirements file: %s' % str(exc)
)
return url, content
def _clean_link(url):
# type: (str) -> str
"""Makes sure a link is fully encoded. That is, if a ' ' shows up in
the link, it will be rewritten to %20 (while not over-quoting
% or other characters)."""
# Split the URL into parts according to the general structure
# `scheme://netloc/path;parameters?query#fragment`. Note that the
# `netloc` can be empty and the URI will then refer to a local
# filesystem path.
result = urllib_parse.urlparse(url)
# In both cases below we unquote prior to quoting to make sure
# nothing is double quoted.
if result.netloc == "":
# On Windows the path part might contain a drive letter which
# should not be quoted. On Linux where drive letters do not
# exist, the colon should be quoted. We rely on urllib.request
# to do the right thing here.
path = urllib_request.pathname2url(
urllib_request.url2pathname(result.path))
else:
# In addition to the `/` character we protect `@` so that
# revision strings in VCS URLs are properly parsed.
path = urllib_parse.quote(urllib_parse.unquote(result.path), safe="/@")
return urllib_parse.urlunparse(result._replace(path=path))
def path(self):
return urllib_parse.unquote(urllib_parse.urlsplit(self.url)[2])
def filename(self):
_, netloc, path, _, _ = urllib_parse.urlsplit(self.url)
name = posixpath.basename(path.rstrip('/')) or netloc
name = urllib_parse.unquote(name)
assert name, ('URL %r produced no filename' % self.url)
return name
def path(self):
# type: () -> str
return urllib_parse.unquote(urllib_parse.urlsplit(self.url)[2])
def path(self):
return urllib_parse.unquote(urllib_parse.urlsplit(self.url)[2])
def normalize_url(self, url):
"""
Normalize a URL for comparison by unquoting it and removing any
trailing slash.
"""
return urllib_parse.unquote(url).rstrip('/')
def path(self):
# type: () -> str
return urllib_parse.unquote(self._parsed_url.path)