def correct_build_location(self):
"""If the build location was a temporary directory, this will move it
to a new more permanent location"""
if self.source_dir is not None:
return
assert self.req is not None
assert self._temp_build_dir
old_location = self._temp_build_dir
new_build_dir = self._ideal_build_dir
del self._ideal_build_dir
if self.editable:
name = self.name.lower()
else:
name = self.name
new_location = os.path.join(new_build_dir, name)
if not os.path.exists(new_build_dir):
logger.debug('Creating directory %s' % new_build_dir)
_make_build_dir(new_build_dir)
if os.path.exists(new_location):
raise InstallationError(
'A package already exists in %s; please remove it to continue'
% display_path(new_location))
logger.debug(
'Moving package %s from %s to new location %s' %
(self, display_path(old_location), display_path(new_location)))
shutil.move(old_location, new_location)
self._temp_build_dir = new_location
self.source_dir = new_location
self._egg_info_path = None
def _get_queued_page(self, req, pending_queue, done, seen):
while 1:
try:
location = pending_queue.get(False)
except QueueEmpty:
return
if location in seen:
continue
seen.add(location)
page = self._get_page(location, req)
if page is None:
continue
done.append(page)
for link in page.rel_links():
normalized = normalize_name(req.name).lower()
if (not normalized in self.allow_external
and not self.allow_all_external):
self.need_warn_external = True
logger.debug("Not searching %s for files because external "
"urls are disallowed." % link)
continue
if (link.trusted is not None and not link.trusted
and not normalized in self.allow_insecure
and not self.allow_all_insecure
): # TODO: Remove after release
logger.debug("Not searching %s for urls, it is an "
"untrusted link and cannot produce safe or "
"verifiable files." % link)
self.need_warn_insecure = True
continue
pending_queue.put(link)
def paths(self):
"""All the entries of sys.path, possibly restricted by --path"""
if not self.select_paths:
return sys.path
result = []
match_any = set()
for path in sys.path:
path = os.path.normcase(os.path.abspath(path))
for match in self.select_paths:
match = os.path.normcase(os.path.abspath(match))
if '*' in match:
if re.search(fnmatch.translate(match+'*'), path):
result.append(path)
match_any.add(match)
break
else:
if path.startswith(match):
result.append(path)
match_any.add(match)
break
else:
logger.debug("Skipping path %s because it doesn't match %s"
% (path, ', '.join(self.select_paths)))
for match in self.select_paths:
if match not in match_any and '*' not in match:
result.append(match)
logger.debug("Adding path %s because it doesn't match anything already on sys.path"
% match)
return result
def paths(self):
"""All the entries of sys.path, possibly restricted by --path"""
if not self.select_paths:
return sys.path
result = []
match_any = set()
for path in sys.path:
path = os.path.normcase(os.path.abspath(path))
for match in self.select_paths:
match = os.path.normcase(os.path.abspath(match))
if '*' in match:
if re.search(fnmatch.translate(match+'*'), path):
result.append(path)
match_any.add(match)
break
else:
if path.startswith(match):
result.append(path)
match_any.add(match)
break
else:
logger.debug("Skipping path %s because it doesn't match %s"
% (path, ', '.join(self.select_paths)))
for match in self.select_paths:
if match not in match_any and '*' not in match:
result.append(match)
logger.debug("Adding path %s because it doesn't match anything already on sys.path"
% match)
return result
def correct_build_location(self):
"""If the build location was a temporary directory, this will move it
to a new more permanent location"""
if self.source_dir is not None:
return
assert self.req is not None
assert self._temp_build_dir
old_location = self._temp_build_dir
new_build_dir = self._ideal_build_dir
del self._ideal_build_dir
if self.editable:
name = self.name.lower()
else:
name = self.name
new_location = os.path.join(new_build_dir, name)
if not os.path.exists(new_build_dir):
logger.debug('Creating directory %s' % new_build_dir)
_make_build_dir(new_build_dir)
if os.path.exists(new_location):
raise InstallationError(
'A package already exists in %s; please remove it to continue'
% display_path(new_location))
logger.debug(
'Moving package %s from %s to new location %s' %
(self, display_path(old_location), display_path(new_location))
)
shutil.move(old_location, new_location)
self._temp_build_dir = new_location
self.source_dir = new_location
self._egg_info_path = None
def _get_queued_page(self, req, pending_queue, done, seen):
while 1:
try:
location = pending_queue.get(False)
except QueueEmpty:
return
if location in seen:
continue
seen.add(location)
page = self._get_page(location, req)
if page is None:
continue
done.append(page)
for link in page.rel_links():
normalized = normalize_name(req.name).lower()
if (not normalized in self.allow_external
and not self.allow_all_external):
self.need_warn_external = True
logger.debug("Not searching %s for files because external "
"urls are disallowed." % link)
continue
if (link.trusted is not None
and not link.trusted
and not normalized in self.allow_insecure
and not self.allow_all_insecure): # TODO: Remove after release
logger.debug("Not searching %s for urls, it is an "
"untrusted link and cannot produce safe or "
"verifiable files." % link)
self.need_warn_insecure = True
continue
pending_queue.put(link)
def get_available_substitute(self, install_req):
"""Find an available substitute for the given package.
Returns a PackageData object.
"""
global __InstallationErrorMessage__
new_candidate_package_data = PackageData.from_dist(install_req)
if new_candidate_package_data.name is None:
# cannot find alternative versions without a name.
return None
existing_package_data = self.find_potential_substitutes(new_candidate_package_data.name)
if existing_package_data is None:
return None
packages_in_conflict = [new_candidate_package_data, existing_package_data]
# Force a hash so that the cmp operator is not used
editables = set([hash(p) for p in packages_in_conflict if p.editable])
if len(editables) == 2:
local_editable_path = os.path.join(sys.prefix, 'src', existing_package_data.name)
if os.path.isdir(local_editable_path):
if self.check_for_uncommited_git_changes(local_editable_path):
raise InstallationError("{message}. In path: {path}".format(
message=__InstallationErrorMessage__,
path=local_editable_path))
# This is an expensive comparison, so let's cache results
competing_version_urls = [str(r.url) for r in packages_in_conflict]
cmp_result = self.get_cached_comparison_result(*competing_version_urls)
if cmp_result is None:
# We're comparing two versions of an editable because we know we're going to use the software in
# the given repo (its just the version that's not decided yet).
# So let's check out the repo into the src directory. Later (when we have the version) update_editable
# will use the correct version anyway.
repo_dir = self.checkout_if_necessary(new_candidate_package_data)
cmp = GitVersionComparator(repo_dir, self.prefer_pinned_revision)
try:
versions = [GitVersionComparator.get_version_string_from_url(r.url) for r in packages_in_conflict]
if len([v for v in versions if v == None]) == 2:
# if either the existing requirement or the new candidate has no version info and is editable,
# we better update our clone and re-run setup.
return existing_package_data # OPTIMIZE return with the installed version
cmp_result = cmp.compare_versions(*versions)
self.save_comparison_result(competing_version_urls[0], competing_version_urls[1], cmp_result)
except SeparateBranchException, exc:
raise InstallationError(
"%s: Conflicting versions cannot be compared as they are not direct descendants according to git. Exception: %s, Package data: %s." % (
new_candidate_package_data.name,
str([p.__dict__ for p in packages_in_conflict]),
str(exc.args)))
else:
logger.debug("using cached comparison: %s %s -> %s" % (competing_version_urls[0], competing_version_urls[1], cmp_result))
return None if cmp_result == GitVersionComparator.GT else existing_package_data
def assert_source_matches_version(self):
assert self.source_dir
version = self.installed_version
if version not in self.req:
logger.warn('Requested %s, but installing version %s' %
(self, self.installed_version))
else:
logger.debug(
'Source in %s has version %s, which satisfies requirement %s' %
(display_path(self.source_dir), version, self))
def _egg_info_matches(self, egg_info, search_name, link):
match = self._egg_info_re.search(egg_info)
if not match:
logger.debug('Could not parse version from link: %s' % link)
return None
name = match.group(0).lower()
# To match the "safe" name that pkg_resources creates:
name = name.replace('_', '-')
if name.startswith(search_name.lower()):
return match.group(0)[len(search_name):].lstrip('-')
else:
return None
def _egg_info_matches(self, egg_info, search_name, link):
match = self._egg_info_re.search(egg_info)
if not match:
logger.debug('Could not parse version from link: %s' % link)
return None
name = match.group(0).lower()
# To match the "safe" name that pkg_resources creates:
name = name.replace('_', '-')
if name.startswith(search_name.lower()):
return match.group(0)[len(search_name):].lstrip('-')
else:
return None
def requirements(self, extras=()):
in_extra = None
for line in self.egg_info_lines('requires.txt'):
match = self._requirements_section_re.match(line.lower())
if match:
in_extra = match.group(1)
continue
if in_extra and in_extra not in extras:
logger.debug('skipping extra %s' % in_extra)
# Skip requirement for an extra we aren't requiring
continue
yield line
def requirements(self, extras=()):
in_extra = None
for line in self.egg_info_lines('requires.txt'):
match = self._requirements_section_re.match(line.lower())
if match:
in_extra = match.group(1)
continue
if in_extra and in_extra not in extras:
logger.debug('skipping extra %s' % in_extra)
# Skip requirement for an extra we aren't requiring
continue
yield line
def assert_source_matches_version(self):
assert self.source_dir
version = self.installed_version
if version not in self.req:
logger.warn(
'Requested %s, but installing version %s' %
(self, self.installed_version)
)
else:
logger.debug(
'Source in %s has version %s, which satisfies requirement %s' %
(display_path(self.source_dir), version, self)
)
def _egg_info_matches(self, egg_info, search_name, link):
match = self._egg_info_re.search(egg_info)
if not match:
logger.debug('Could not parse version from link: %s' % link)
return None
name = match.group(0).lower()
# To match the "safe" name that pkg_resources creates:
name = name.replace('_', '-')
# project name and version must be separated by a dash
look_for = search_name.lower() + "-"
if name.startswith(look_for):
return match.group(0)[len(look_for):]
else:
return None
def _egg_info_matches(self, egg_info, search_name, link):
match = self._egg_info_re.search(egg_info)
if not match:
logger.debug('Could not parse version from link: %s' % link)
return None
name = match.group(0).lower()
# To match the "safe" name that pkg_resources creates:
name = name.replace('_', '-')
# project name and version must be separated by a dash
look_for = search_name.lower() + "-"
if name.startswith(look_for):
return match.group(0)[len(look_for):]
else:
return None
def __build_one(self, req, tempd, python_tag=None):
base_args = self._base_setup_args(req)
spin_message = 'Running setup.py bdist_wheel for %s' % (req.name,)
with open_spinner(spin_message) as spinner:
logger.debug('Destination directory: %s', tempd)
wheel_args = base_args + ['bdist_wheel', '-d', tempd] \
+ self.build_options
if python_tag is not None:
wheel_args += ["--python-tag", python_tag]
try:
call_subprocess(wheel_args, cwd=req.source_dir,
show_stdout=False, spinner=spinner)
return True
except:
def _download_url(resp, link, temp_location):
fp = open(temp_location, 'wb')
download_hash = None
if link.md5_hash:
download_hash = md5()
try:
total_length = int(resp.info()['content-length'])
except (ValueError, KeyError):
total_length = 0
downloaded = 0
show_progress = total_length > 40 * 1000 or not total_length
show_url = link.show_url
try:
if show_progress:
## FIXME: the URL can get really long in this message:
if total_length:
logger.start_progress('Downloading %s (%s): ' %
(show_url, format_size(total_length)))
else:
logger.start_progress('Downloading %s (unknown size): ' %
show_url)
else:
logger.notify('Downloading %s' % show_url)
logger.debug('Downloading from URL %s' % link)
while 1:
chunk = resp.read(4096)
if not chunk:
break
downloaded += len(chunk)
if show_progress:
if not total_length:
logger.show_progress('%s' % format_size(downloaded))
else:
logger.show_progress('%3i%% %s' %
(100 * downloaded / total_length,
format_size(downloaded)))
if link.md5_hash:
download_hash.update(chunk)
fp.write(chunk)
fp.close()
finally:
if show_progress:
logger.end_progress('%s downloaded' % format_size(downloaded))
return download_hash
def _download_url(resp, link, temp_location):
fp = open(temp_location, 'wb')
download_hash = None
if link.hash and link.hash_name:
try:
download_hash = hashlib.new(link.hash_name)
except ValueError:
logger.warn("Unsupported hash name %s for package %s" % (link.hash_name, link))
try:
total_length = int(resp.info()['content-length'])
except (ValueError, KeyError, TypeError):
total_length = 0
downloaded = 0
show_progress = total_length > 40*1000 or not total_length
show_url = link.show_url
try:
if show_progress:
## FIXME: the URL can get really long in this message:
if total_length:
logger.start_progress('Downloading %s (%s): ' % (show_url, format_size(total_length)))
else:
logger.start_progress('Downloading %s (unknown size): ' % show_url)
else:
logger.notify('Downloading %s' % show_url)
logger.debug('Downloading from URL %s' % link)
while True:
chunk = resp.read(4096)
if not chunk:
break
downloaded += len(chunk)
if show_progress:
if not total_length:
logger.show_progress('%s' % format_size(downloaded))
else:
logger.show_progress('%3i%% %s' % (100*downloaded/total_length, format_size(downloaded)))
if download_hash is not None:
download_hash.update(chunk)
fp.write(chunk)
fp.close()
finally:
if show_progress:
logger.end_progress('%s downloaded' % format_size(downloaded))
return download_hash
def _link_package_versions(self, link, search_name):
"""
Return an iterable of triples (pkg_resources_version_key,
link, python_version) that can be extracted from the given
link.
Meant to be overridden by subclasses, not called by clients.
"""
if link.egg_fragment:
egg_info = link.egg_fragment
else:
egg_info, ext = link.splitext()
if not ext:
if link not in self.logged_links:
logger.debug('Skipping link %s; not a file' % link)
self.logged_links.add(link)
return []
if egg_info.endswith('.tar'):
# Special double-extension case:
egg_info = egg_info[:-4]
ext = '.tar' + ext
if ext not in ('.tar.gz', '.tar.bz2', '.tar', '.tgz', '.zip'):
if link not in self.logged_links:
logger.debug(
'Skipping link %s; unknown archive format: %s' %
(link, ext))
self.logged_links.add(link)
return []
version = self._egg_info_matches(egg_info, search_name, link)
if version is None:
logger.debug('Skipping link %s; wrong project name (not %s)' %
(link, search_name))
return []
match = self._py_version_re.search(version)
if match:
version = version[:match.start()]
py_version = match.group(1)
if py_version != sys.version[:3]:
logger.debug(
'Skipping %s because Python version is incorrect' % link)
return []
logger.debug('Found link %s, version: %s' % (link, version))
return [(pkg_resources.parse_version(version), link, version)]
def _package_versions(self, links, search_name):
seen_links = {}
for link in self._sort_links(links):
if link.url in seen_links:
continue
seen_links[link.url] = None
if link.egg_fragment:
egg_info = link.egg_fragment
else:
path = link.path
egg_info, ext = link.splitext()
if not ext:
if link not in self.logged_links:
logger.debug('Skipping link %s; not a file' % link)
self.logged_links.add(link)
continue
if egg_info.endswith('.tar'):
# Special double-extension case:
egg_info = egg_info[:-4]
ext = '.tar' + ext
if ext not in ('.tar.gz', '.tar.bz2', '.tar', '.tgz', '.zip'):
if link not in self.logged_links:
logger.debug(
'Skipping link %s; unknown archive format: %s' %
(link, ext))
self.logged_links.add(link)
continue
version = self._egg_info_matches(egg_info, search_name, link)
if version is None:
logger.debug('Skipping link %s; wrong project name (not %s)' %
(link, search_name))
continue
match = self._py_version_re.search(version)
if match:
version = version[:match.start()]
py_version = match.group(1)
if py_version != sys.version[:3]:
logger.debug(
'Skipping %s because Python version is incorrect' %
link)
continue
logger.debug('Found link %s, version: %s' % (link, version))
yield (pkg_resources.parse_version(version), link, version)
def safe_requirements(self):
"""
safe implementation of pip.req.InstallRequirement.requirements() generator, doesn't blow up with OSError
"""
in_extra = None
try:
for line in self.egg_info_lines('requires.txt'):
match = self._requirements_section_re.match(line.lower())
if match:
in_extra = match.group(1)
continue
if in_extra:
logger.debug('skipping extra %s' % in_extra)
# Skip requirement for an extra we aren't requiring
continue
yield line
except OSError:
pass
def safe_requirements(self):
"""
safe implementation of pip.req.InstallRequirement.requirements() generator, doesn't blow up with OSError
"""
in_extra = None
try:
for line in self.egg_info_lines('requires.txt'):
match = self._requirements_section_re.match(line.lower())
if match:
in_extra = match.group(1)
continue
if in_extra:
logger.debug('skipping extra %s' % in_extra)
# Skip requirement for an extra we aren't requiring
continue
yield line
except OSError:
pass
def _get_pages(self, locations, req):
"""
Yields (page, page_url) from the given locations, skipping
locations that have errors, and adding download/homepage links
"""
all_locations = list(locations)
seen = set()
while all_locations:
location = all_locations.pop(0)
if location in seen:
continue
seen.add(location)
page = self._get_page(location, req)
if page is None:
continue
yield page
for link in page.rel_links():
normalized = normalize_name(req.name).lower()
if (normalized not in self.allow_external
and not self.allow_all_external):
self.need_warn_external = True
logger.debug("Not searching %s for files because external "
"urls are disallowed." % link)
continue
if (link.trusted is not None
and not link.trusted
and normalized not in self.allow_unverified):
logger.debug(
"Not searching %s for urls, it is an "
"untrusted link and cannot produce safe or "
"verifiable files." % link
)
self.need_warn_unverified = True
continue
all_locations.append(link)
def _get_pages(self, locations, req):
"""
Yields (page, page_url) from the given locations, skipping
locations that have errors, and adding download/homepage links
"""
all_locations = list(locations)
seen = set()
while all_locations:
location = all_locations.pop(0)
if location in seen:
continue
seen.add(location)
page = self._get_page(location, req)
if page is None:
continue
yield page
for link in page.rel_links():
normalized = normalize_name(req.name).lower()
if (normalized not in self.allow_external
and not self.allow_all_external):
self.need_warn_external = True
logger.debug("Not searching %s for files because external "
"urls are disallowed." % link)
continue
if (link.trusted is not None
and not link.trusted
and normalized not in self.allow_unverified):
logger.debug(
"Not searching %s for urls, it is an "
"untrusted link and cannot produce safe or "
"verifiable files." % link
)
self.need_warn_unverified = True
continue
all_locations.append(link)
def _download_url(resp, link, temp_location):
fp = open(temp_location, "wb")
download_hash = None
if link.md5_hash:
download_hash = md5()
try:
total_length = int(resp.info()["content-length"])
except (ValueError, KeyError):
total_length = 0
downloaded = 0
show_progress = total_length > 40 * 1000 or not total_length
show_url = link.show_url
try:
if show_progress:
## FIXME: the URL can get really long in this message:
if total_length:
logger.start_progress("Downloading %s (%s): " % (show_url, format_size(total_length)))
else:
logger.start_progress("Downloading %s (unknown size): " % show_url)
else:
logger.notify("Downloading %s" % show_url)
logger.debug("Downloading from URL %s" % link)
while True:
chunk = resp.read(4096)
if not chunk:
break
downloaded += len(chunk)
if show_progress:
if not total_length:
logger.show_progress("%s" % format_size(downloaded))
else:
logger.show_progress("%3i%% %s" % (100 * downloaded / total_length, format_size(downloaded)))
if link.md5_hash:
download_hash.update(chunk)
fp.write(chunk)
fp.close()
finally:
if show_progress:
logger.end_progress("%s downloaded" % format_size(downloaded))
return download_hash
def _download_url(resp, link, temp_location):
fp = open(temp_location, 'wb')
download_hash = None
if link.md5_hash:
download_hash = md5()
try:
total_length = int(resp.info()['content-length'])
except (ValueError, KeyError):
total_length = 0
downloaded = 0
show_progress = total_length > 40*1000 or not total_length
show_url = link.show_url
try:
if show_progress:
## FIXME: the URL can get really long in this message:
if total_length:
logger.start_progress('Downloading {0!s} ({1!s}): '.format(show_url, format_size(total_length)))
else:
logger.start_progress('Downloading {0!s} (unknown size): '.format(show_url))
else:
logger.notify('Downloading {0!s}'.format(show_url))
logger.debug('Downloading from URL {0!s}'.format(link))
while True:
chunk = resp.read(4096)
if not chunk:
break
downloaded += len(chunk)
if show_progress:
if not total_length:
logger.show_progress('{0!s}'.format(format_size(downloaded)))
else:
logger.show_progress('{0:3d}% {1!s}'.format(100*downloaded/total_length, format_size(downloaded)))
if link.md5_hash:
download_hash.update(chunk)
fp.write(chunk)
fp.close()
finally:
if show_progress:
logger.end_progress('{0!s} downloaded'.format(format_size(downloaded)))
return download_hash
def _link_package_versions(self, link, search_name):
"""
Return an iterable of triples (pkg_resources_version_key,
link, python_version) that can be extracted from the given
link.
Meant to be overridden by subclasses, not called by clients.
"""
if link.egg_fragment:
egg_info = link.egg_fragment
else:
egg_info, ext = link.splitext()
if not ext:
if link not in self.logged_links:
logger.debug('Skipping link {0!s}; not a file'.format(link))
self.logged_links.add(link)
return []
if egg_info.endswith('.tar'):
# Special double-extension case:
egg_info = egg_info[:-4]
ext = '.tar' + ext
if ext not in ('.tar.gz', '.tar.bz2', '.tar', '.tgz', '.zip'):
if link not in self.logged_links:
logger.debug('Skipping link {0!s}; unknown archive format: {1!s}'.format(link, ext))
self.logged_links.add(link)
return []
version = self._egg_info_matches(egg_info, search_name, link)
if version is None:
logger.debug('Skipping link {0!s}; wrong project name (not {1!s})'.format(link, search_name))
return []
match = self._py_version_re.search(version)
if match:
version = version[:match.start()]
py_version = match.group(1)
if py_version != sys.version[:3]:
logger.debug('Skipping {0!s} because Python version is incorrect'.format(link))
return []
logger.debug('Found link {0!s}, version: {1!s}'.format(link, version))
return [(pkg_resources.parse_version(version),
link,
version)]
def _package_versions(self, links, search_name):
seen_links = {}
for link in self._sort_links(links):
if link.url in seen_links:
continue
seen_links[link.url] = None
if link.egg_fragment:
egg_info = link.egg_fragment
else:
path = link.path
egg_info, ext = link.splitext()
if not ext:
if link not in self.logged_links:
logger.debug('Skipping link %s; not a file' % link)
self.logged_links.add(link)
continue
if egg_info.endswith('.tar'):
# Special double-extension case:
egg_info = egg_info[:-4]
ext = '.tar' + ext
if ext not in ('.tar.gz', '.tar.bz2', '.tar', '.tgz', '.zip'):
if link not in self.logged_links:
logger.debug('Skipping link %s; unknown archive format: %s' % (link, ext))
self.logged_links.add(link)
continue
version = self._egg_info_matches(egg_info, search_name, link)
if version is None:
logger.debug('Skipping link %s; wrong project name (not %s)' % (link, search_name))
continue
match = self._py_version_re.search(version)
if match:
version = version[:match.start()]
py_version = match.group(1)
if py_version != sys.version[:3]:
logger.debug('Skipping %s because Python version is incorrect' % link)
continue
logger.debug('Found link %s, version: %s' % (link, version))
yield (pkg_resources.parse_version(version),
link,
version)
def get_page(cls, link, req, cache=None, skip_archives=True):
url = link.url
url = url.split('#', 1)[0]
if cache.too_many_failures(url):
return None
if url.lower().startswith('svn'):
logger.debug('Cannot look at svn URL %s' % link)
return None
if cache is not None:
inst = cache.get_page(url)
if inst is not None:
return inst
try:
if skip_archives:
if cache is not None:
if cache.is_archive(url):
return None
filename = link.filename
for bad_ext in ['.tar', '.tar.gz', '.tar.bz2', '.tgz', '.zip']:
if filename.endswith(bad_ext):
content_type = cls._get_content_type(url)
if content_type.lower().startswith('text/html'):
break
else:
logger.debug('Skipping page %s because of Content-Type: %s' % (link, content_type))
if cache is not None:
cache.set_is_archive(url)
return None
logger.debug('Getting page %s' % url)
resp = urllib2.urlopen(url)
real_url = resp.geturl()
headers = resp.info()
inst = cls(resp.read(), real_url, headers)
except (urllib2.HTTPError, urllib2.URLError, socket.timeout, socket.error), e:
desc = str(e)
if isinstance(e, socket.timeout):
log_meth = logger.info
level =1
desc = 'timed out'
elif isinstance(e, urllib2.URLError):
log_meth = logger.info
if hasattr(e, 'reason') and isinstance(e.reason, socket.timeout):
desc = 'timed out'
level = 1
else:
level = 2
elif isinstance(e, urllib2.HTTPError) and e.code == 404:
## FIXME: notify?
log_meth = logger.info
level = 2
else:
log_meth = logger.info
level = 1
log_meth('Could not fetch URL %s: %s' % (link, desc))
log_meth('Will skip URL %s when looking for download links for %s' % (link.url, req))
if cache is not None:
cache.add_page_failure(url, level)
return None
def find_requirement(self, req, upgrade):
def mkurl_pypi_url(url):
loc = posixpath.join(url, url_name)
# For maximum compatibility with easy_install, ensure the path
# ends in a trailing slash. Although this isn't in the spec
# (and PyPI can handle it without the slash) some other index
# implementations might break if they relied on easy_install's behavior.
if not loc.endswith('/'):
loc = loc + '/'
return loc
url_name = req.url_name
# Only check main index if index URL is given:
main_index_url = None
if self.index_urls:
# Check that we have the url_name correctly spelled:
main_index_url = Link(mkurl_pypi_url(self.index_urls[0]),
trusted=True)
# This will also cache the page, so it's okay that we get it again later:
page = self._get_page(main_index_url, req)
if page is None:
url_name = self._find_url_name(
Link(self.index_urls[0], trusted=True), url_name,
req) or req.url_name
if url_name is not None:
locations = [mkurl_pypi_url(url)
for url in self.index_urls] + self.find_links
else:
locations = list(self.find_links)
for version in req.absolute_versions:
if url_name is not None and main_index_url is not None:
locations = [posixpath.join(main_index_url.url, version)
] + locations
file_locations, url_locations = self._sort_locations(locations)
_flocations, _ulocations = self._sort_locations(self.dependency_links)
file_locations.extend(_flocations)
# We trust every url that the user has given us whether it was given
# via --index-url or --find-links
locations = [Link(url, trusted=True) for url in url_locations]
# We explicitly do not trust links that came from dependency_links
locations.extend([Link(url) for url in _ulocations])
logger.debug('URLs to search for versions for %s:' % req)
for location in locations:
logger.debug('* %s' % location)
# Determine if this url used a secure transport mechanism
parsed = urlparse.urlparse(str(location))
if parsed.scheme in INSECURE_SCHEMES:
secure_schemes = INSECURE_SCHEMES[parsed.scheme]
if len(secure_schemes) == 1:
ctx = (location, parsed.scheme, secure_schemes[0],
parsed.netloc)
logger.warn("%s uses an insecure transport scheme (%s). "
"Consider using %s if %s has it available" %
ctx)
elif len(secure_schemes) > 1:
ctx = (location, parsed.scheme, ", ".join(secure_schemes),
parsed.netloc)
logger.warn("%s uses an insecure transport scheme (%s). "
"Consider using one of %s if %s has any of "
"them available" % ctx)
else:
ctx = (location, parsed.scheme)
logger.warn("%s uses an insecure transport scheme (%s)." %
ctx)
found_versions = []
found_versions.extend(
self._package_versions(
# We trust every directly linked archive in find_links
[Link(url, '-f', trusted=True) for url in self.find_links],
req.name.lower()))
page_versions = []
for page in self._get_pages(locations, req):
logger.debug('Analyzing links from page %s' % page.url)
logger.indent += 2
try:
page_versions.extend(
self._package_versions(page.links, req.name.lower()))
finally:
logger.indent -= 2
dependency_versions = list(
self._package_versions(
[Link(url) for url in self.dependency_links],
req.name.lower()))
if dependency_versions:
logger.info('dependency_links found: %s' % ', '.join(
[link.url for parsed, link, version in dependency_versions]))
file_versions = list(
self._package_versions([Link(url) for url in file_locations],
req.name.lower()))
if not found_versions and not page_versions and not dependency_versions and not file_versions:
logger.fatal(
'Could not find any downloads that satisfy the requirement %s'
% req)
if self.need_warn_external:
logger.warn("Some externally hosted files were ignored (use "
"--allow-external %s to allow)." % req.name)
if self.need_warn_unverified:
logger.warn("Some insecure and unverifiable files were ignored"
" (use --allow-unverified %s to allow)." %
req.name)
raise DistributionNotFound('No distributions at all found for %s' %
req)
installed_version = []
if req.satisfied_by is not None:
installed_version = [(req.satisfied_by.parsed_version,
INSTALLED_VERSION, req.satisfied_by.version)]
if file_versions:
file_versions.sort(reverse=True)
logger.info('Local files found: %s' % ', '.join([
url_to_path(link.url)
for parsed, link, version in file_versions
]))
#this is an intentional priority ordering
all_versions = installed_version + file_versions + found_versions + page_versions + dependency_versions
applicable_versions = []
for (parsed_version, link, version) in all_versions:
if version not in req.req:
logger.info("Ignoring link %s, version %s doesn't match %s" %
(link, version, ','.join(
[''.join(s) for s in req.req.specs])))
continue
elif is_prerelease(version) and not (self.allow_all_prereleases
or req.prereleases):
# If this version isn't the already installed one, then
# ignore it if it's a pre-release.
if link is not INSTALLED_VERSION:
logger.info(
"Ignoring link %s, version %s is a pre-release (use --pre to allow)."
% (link, version))
continue
applicable_versions.append((parsed_version, link, version))
applicable_versions = self._sort_versions(applicable_versions)
existing_applicable = bool([
link for parsed_version, link, version in applicable_versions
if link is INSTALLED_VERSION
])
if not upgrade and existing_applicable:
if applicable_versions[0][1] is INSTALLED_VERSION:
logger.info(
'Existing installed version (%s) is most up-to-date and satisfies requirement'
% req.satisfied_by.version)
else:
logger.info(
'Existing installed version (%s) satisfies requirement (most up-to-date version is %s)'
% (req.satisfied_by.version, applicable_versions[0][2]))
return None
if not applicable_versions:
logger.fatal(
'Could not find a version that satisfies the requirement %s (from versions: %s)'
% (req, ', '.join([
version for parsed_version, link, version in all_versions
])))
if self.need_warn_external:
logger.warn("Some externally hosted files were ignored (use "
"--allow-external to allow).")
if self.need_warn_unverified:
logger.warn("Some insecure and unverifiable files were ignored"
" (use --allow-unverified %s to allow)." %
req.name)
raise DistributionNotFound(
'No distributions matching the version for %s' % req)
if applicable_versions[0][1] is INSTALLED_VERSION:
# We have an existing version, and its the best version
logger.info(
'Installed version (%s) is most up-to-date (past versions: %s)'
% (req.satisfied_by.version, ', '.join([
version for parsed_version, link, version in
applicable_versions[1:]
]) or 'none'))
raise BestVersionAlreadyInstalled
if len(applicable_versions) > 1:
logger.info(
'Using version %s (newest of versions: %s)' %
(applicable_versions[0][2], ', '.join([
version
for parsed_version, link, version in applicable_versions
])))
selected_version = applicable_versions[0][1]
# TODO: Remove after 1.4 has been released
# if (selected_version.internal is not None
# and not selected_version.internal):
# logger.warn("You are installing an externally hosted file. Future "
# "versions of pip will default to disallowing "
# "externally hosted files.")
# if (selected_version.verifiable is not None
# and not selected_version.verifiable):
# logger.warn("You are installing a potentially insecure and "
# "unverifiable file. Future versions of pip will "
# "default to disallowing insecure files.")
if selected_version._deprecated_regex:
logger.deprecated(
"1.7", "%s discovered using a deprecated method of parsing, "
"in the future it will no longer be discovered" % req.name)
return selected_version
def get_page(cls, link, req, skip_archives=True, session=None):
if session is None:
raise TypeError(
"get_page() missing 1 required keyword argument: 'session'"
)
url = link.url
url = url.split('#', 1)[0]
# Check for VCS schemes that do not support lookup as web pages.
from pip.vcs import VcsSupport
for scheme in VcsSupport.schemes:
if url.lower().startswith(scheme) and url[len(scheme)] in '+:':
logger.debug(
'Cannot look at %(scheme)s URL %(link)s' % locals()
)
return None
try:
if skip_archives:
filename = link.filename
for bad_ext in ['.tar', '.tar.gz', '.tar.bz2', '.tgz', '.zip']:
if filename.endswith(bad_ext):
content_type = cls._get_content_type(
url, session=session,
)
if content_type.lower().startswith('text/html'):
break
else:
logger.debug(
'Skipping page %s because of Content-Type: '
'%s' % (link, content_type)
)
return
logger.debug('Getting page %s' % url)
# Tack index.html onto file:// URLs that point to directories
(scheme, netloc, path, params, query, fragment) = \
urlparse.urlparse(url)
if scheme == 'file' and os.path.isdir(url2pathname(path)):
# add trailing slash if not present so urljoin doesn't trim
# final segment
if not url.endswith('/'):
url += '/'
url = urlparse.urljoin(url, 'index.html')
logger.debug(' file: URL is directory, getting %s' % url)
resp = session.get(
url,
headers={
"Accept": "text/html",
"Cache-Control": "max-age=600",
},
)
resp.raise_for_status()
# The check for archives above only works if the url ends with
# something that looks like an archive. However that is not a
# requirement of an url. Unless we issue a HEAD request on every
# url we cannot know ahead of time for sure if something is HTML
# or not. However we can check after we've downloaded it.
content_type = resp.headers.get('Content-Type', 'unknown')
if not content_type.lower().startswith("text/html"):
logger.debug(
'Skipping page %s because of Content-Type: %s' %
(link, content_type)
)
return
inst = cls(resp.text, resp.url, resp.headers, trusted=link.trusted)
except requests.HTTPError as exc:
level = 2 if exc.response.status_code == 404 else 1
cls._handle_fail(req, link, exc, url, level=level)
except requests.ConnectionError as exc:
cls._handle_fail(
req, link, "connection error: %s" % exc, url,
)
except requests.Timeout:
cls._handle_fail(req, link, "timed out", url)
except SSLError as exc:
reason = ("There was a problem confirming the ssl certificate: "
"%s" % exc)
cls._handle_fail(
req, link, reason, url,
level=2,
meth=logger.notify,
)
else:
return inst
def get_page(cls, link, req, cache=None, skip_archives=True):
url = link.url
url = url.split('#', 1)[0]
if cache.too_many_failures(url):
return None
# Check for VCS schemes that do not support lookup as web pages.
from pip.vcs import VcsSupport
for scheme in VcsSupport.schemes:
if url.lower().startswith(scheme) and url[len(scheme)] in '+:':
logger.debug('Cannot look at %(scheme)s URL %(link)s' %
locals())
return None
if cache is not None:
inst = cache.get_page(url)
if inst is not None:
return inst
try:
if skip_archives:
if cache is not None:
if cache.is_archive(url):
return None
filename = link.filename
for bad_ext in ['.tar', '.tar.gz', '.tar.bz2', '.tgz', '.zip']:
if filename.endswith(bad_ext):
content_type = cls._get_content_type(url)
if content_type.lower().startswith('text/html'):
break
else:
logger.debug(
'Skipping page %s because of Content-Type: %s'
% (link, content_type))
if cache is not None:
cache.set_is_archive(url)
return None
logger.debug('Getting page %s' % url)
# Tack list.html onto file:// URLs that point to directories
(scheme, netloc, path, params, query,
fragment) = urlparse.urlparse(url)
if scheme == 'file' and os.path.isdir(url2pathname(path)):
# add trailing slash if not present so urljoin doesn't trim final segment
if not url.endswith('/'):
url += '/'
url = urlparse.urljoin(url, 'list.html')
logger.debug(' file: URL is directory, getting %s' % url)
resp = urlopen(url)
real_url = geturl(resp)
headers = resp.info()
contents = resp.read()
encoding = headers.get('Content-Encoding', None)
#XXX need to handle exceptions and add testing for this
if encoding is not None:
if encoding == 'gzip':
contents = gzip.GzipFile(fileobj=BytesIO(contents)).read()
if encoding == 'deflate':
contents = zlib.decompress(contents)
# The check for archives above only works if the url ends with
# something that looks like an archive. However that is not a
# requirement. For instance http://sourceforge.net/projects/docutils/files/docutils/0.8.1/docutils-0.8.1.tar.gz/download
# redirects to http://superb-dca3.dl.sourceforge.net/project/docutils/docutils/0.8.1/docutils-0.8.1.tar.gz
# Unless we issue a HEAD request on every url we cannot know
# ahead of time for sure if something is HTML or not. However we
# can check after we've downloaded it.
content_type = headers.get('Content-Type', 'unknown')
if not content_type.lower().startswith("text/html"):
logger.debug('Skipping page %s because of Content-Type: %s' %
(link, content_type))
if cache is not None:
cache.set_is_archive(url)
return None
inst = cls(u(contents), real_url, headers, trusted=link.trusted)
except (HTTPError, URLError, socket.timeout, socket.error, OSError,
WindowsError):
e = sys.exc_info()[1]
desc = str(e)
if isinstance(e, socket.timeout):
log_meth = logger.info
level = 1
desc = 'timed out'
elif isinstance(e, URLError):
#ssl/certificate error
if hasattr(e, 'reason') and (
isinstance(e.reason, ssl.SSLError)
or isinstance(e.reason, CertificateError)):
desc = 'There was a problem confirming the ssl certificate: %s' % e
log_meth = logger.notify
else:
log_meth = logger.info
if hasattr(e, 'reason') and isinstance(e.reason,
socket.timeout):
desc = 'timed out'
level = 1
else:
level = 2
elif isinstance(e, HTTPError) and e.code == 404:
## FIXME: notify?
log_meth = logger.info
level = 2
else:
log_meth = logger.info
level = 1
log_meth('Could not fetch URL %s: %s' % (link, desc))
log_meth(
'Will skip URL %s when looking for download links for %s' %
(link.url, req))
if cache is not None:
cache.add_page_failure(url, level)
return None
if cache is not None:
cache.add_page([url, real_url], inst)
return inst
def find_requirement(self, req, upgrade):
def mkurl_pypi_url(url):
loc = posixpath.join(url, url_name)
# For maximum compatibility with easy_install, ensure the path
# ends in a trailing slash. Although this isn't in the spec
# (and PyPI can handle it without the slash) some other index
# implementations might break if they relied on easy_install's behavior.
if not loc.endswith('/'):
loc = loc + '/'
return loc
url_name = req.url_name
# Only check main index if index URL is given:
main_index_url = None
if self.index_urls:
# Check that we have the url_name correctly spelled:
main_index_url = Link(mkurl_pypi_url(self.index_urls[0]), trusted=True)
# This will also cache the page, so it's okay that we get it again later:
page = self._get_page(main_index_url, req)
if page is None:
url_name = self._find_url_name(Link(self.index_urls[0], trusted=True), url_name, req) or req.url_name
# Combine index URLs with mirror URLs here to allow
# adding more index URLs from requirements files
all_index_urls = self.index_urls + self.mirror_urls
if url_name is not None:
locations = [
mkurl_pypi_url(url)
for url in all_index_urls] + self.find_links
else:
locations = list(self.find_links)
for version in req.absolute_versions:
if url_name is not None and main_index_url is not None:
locations = [
posixpath.join(main_index_url.url, version)] + locations
file_locations, url_locations = self._sort_locations(locations)
_flocations, _ulocations = self._sort_locations(self.dependency_links)
file_locations.extend(_flocations)
# We trust every url that the user has given us whether it was given
# via --index-url, --user-mirrors/--mirror, or --find-links or a
# default option thereof
locations = [Link(url, trusted=True) for url in url_locations]
# We explicitly do not trust links that came from dependency_links
locations.extend([Link(url) for url in _ulocations])
logger.debug('URLs to search for versions for %s:' % req)
for location in locations:
logger.debug('* %s' % location)
found_versions = []
found_versions.extend(
self._package_versions(
# We trust every directly linked archive in find_links
[Link(url, '-f', trusted=True) for url in self.find_links], req.name.lower()))
page_versions = []
for page in self._get_pages(locations, req):
logger.debug('Analyzing links from page %s' % page.url)
logger.indent += 2
try:
page_versions.extend(self._package_versions(page.links, req.name.lower()))
finally:
logger.indent -= 2
dependency_versions = list(self._package_versions(
[Link(url) for url in self.dependency_links], req.name.lower()))
if dependency_versions:
logger.info('dependency_links found: %s' % ', '.join([link.url for parsed, link, version in dependency_versions]))
file_versions = list(self._package_versions(
[Link(url) for url in file_locations], req.name.lower()))
if not found_versions and not page_versions and not dependency_versions and not file_versions:
logger.fatal('Could not find any downloads that satisfy the requirement %s' % req)
if self.need_warn_external:
logger.warn("Some externally hosted files were ignored (use "
"--allow-external %s to allow)." % req.name)
if self.need_warn_insecure:
logger.warn("Some insecure and unverifiable files were ignored"
" (use --allow-insecure %s to allow)." % req.name)
raise DistributionNotFound('No distributions at all found for %s' % req)
installed_version = []
if req.satisfied_by is not None:
installed_version = [(req.satisfied_by.parsed_version, InfLink, req.satisfied_by.version)]
if file_versions:
file_versions.sort(reverse=True)
logger.info('Local files found: %s' % ', '.join([url_to_path(link.url) for parsed, link, version in file_versions]))
#this is an intentional priority ordering
all_versions = installed_version + file_versions + found_versions + page_versions + dependency_versions
applicable_versions = []
for (parsed_version, link, version) in all_versions:
if version not in req.req:
logger.info("Ignoring link %s, version %s doesn't match %s"
% (link, version, ','.join([''.join(s) for s in req.req.specs])))
continue
elif is_prerelease(version) and not (self.allow_all_prereleases or req.prereleases):
# If this version isn't the already installed one, then
# ignore it if it's a pre-release.
if link is not InfLink:
logger.info("Ignoring link %s, version %s is a pre-release (use --pre to allow)." % (link, version))
continue
applicable_versions.append((parsed_version, link, version))
applicable_versions = self._sort_versions(applicable_versions)
existing_applicable = bool([link for parsed_version, link, version in applicable_versions if link is InfLink])
if not upgrade and existing_applicable:
if applicable_versions[0][1] is InfLink:
logger.info('Existing installed version (%s) is most up-to-date and satisfies requirement'
% req.satisfied_by.version)
else:
logger.info('Existing installed version (%s) satisfies requirement (most up-to-date version is %s)'
% (req.satisfied_by.version, applicable_versions[0][2]))
return None
if not applicable_versions:
logger.fatal('Could not find a version that satisfies the requirement %s (from versions: %s)'
% (req, ', '.join([version for parsed_version, link, version in all_versions])))
if self.need_warn_external:
logger.warn("Some externally hosted files were ignored (use "
"--allow-external to allow).")
if self.need_warn_insecure:
logger.warn("Some insecure and unverifiable files were ignored"
" (use --allow-insecure %s to allow)." % req.name)
raise DistributionNotFound('No distributions matching the version for %s' % req)
if applicable_versions[0][1] is InfLink:
# We have an existing version, and its the best version
logger.info('Installed version (%s) is most up-to-date (past versions: %s)'
% (req.satisfied_by.version, ', '.join([version for parsed_version, link, version in applicable_versions[1:]]) or 'none'))
raise BestVersionAlreadyInstalled
if len(applicable_versions) > 1:
logger.info('Using version %s (newest of versions: %s)' %
(applicable_versions[0][2], ', '.join([version for parsed_version, link, version in applicable_versions])))
selected_version = applicable_versions[0][1]
# TODO: Remove after 1.4 has been released
if (selected_version.internal is not None
and not selected_version.internal):
logger.warn("You are installing an externally hosted file. Future "
"versions of pip will default to disallowing "
"externally hosted files.")
if (selected_version.verifiable is not None
and not selected_version.verifiable):
logger.warn("You are installing a potentially insecure and "
"unverifiable file. Future versions of pip will "
"default to disallowing insecure files.")
return selected_version
def get_page(cls, link, req, cache=None, skip_archives=True, session=None):
if session is None:
session = PipSession()
url = link.url
url = url.split('#', 1)[0]
if cache.too_many_failures(url):
return None
# Check for VCS schemes that do not support lookup as web pages.
from pip.vcs import VcsSupport
for scheme in VcsSupport.schemes:
if url.lower().startswith(scheme) and url[len(scheme)] in '+:':
logger.debug(
'Cannot look at %(scheme)s URL %(link)s' % locals()
)
return None
if cache is not None:
inst = cache.get_page(url)
if inst is not None:
return inst
try:
if skip_archives:
if cache is not None:
if cache.is_archive(url):
return None
filename = link.filename
for bad_ext in ['.tar', '.tar.gz', '.tar.bz2', '.tgz', '.zip']:
if filename.endswith(bad_ext):
content_type = cls._get_content_type(
url, session=session,
)
if content_type.lower().startswith('text/html'):
break
else:
logger.debug(
'Skipping page %s because of Content-Type: '
'%s' % (link, content_type)
)
if cache is not None:
cache.set_is_archive(url)
return None
logger.debug('Getting page %s' % url)
# Tack index.html onto file:// URLs that point to directories
(scheme, netloc, path, params, query, fragment) = \
urlparse.urlparse(url)
if scheme == 'file' and os.path.isdir(url2pathname(path)):
# add trailing slash if not present so urljoin doesn't trim
# final segment
if not url.endswith('/'):
url += '/'
url = urlparse.urljoin(url, 'index.html')
logger.debug(' file: URL is directory, getting %s' % url)
resp = session.get(url, headers={"Accept": "text/html"})
resp.raise_for_status()
# The check for archives above only works if the url ends with
# something that looks like an archive. However that is not a
# requirement of an url. Unless we issue a HEAD request on every
# url we cannot know ahead of time for sure if something is HTML
# or not. However we can check after we've downloaded it.
content_type = resp.headers.get('Content-Type', 'unknown')
if not content_type.lower().startswith("text/html"):
logger.debug(
'Skipping page %s because of Content-Type: %s' %
(link, content_type)
)
if cache is not None:
cache.set_is_archive(url)
return None
inst = cls(resp.text, resp.url, resp.headers, trusted=link.trusted)
except requests.HTTPError as exc:
level = 2 if exc.response.status_code == 404 else 1
cls._handle_fail(req, link, exc, url, cache=cache, level=level)
except requests.ConnectionError as exc:
cls._handle_fail(
req, link, "connection error: %s" % exc, url,
cache=cache,
)
except requests.Timeout:
cls._handle_fail(req, link, "timed out", url, cache=cache)
except SSLError as exc:
reason = ("There was a problem confirming the ssl certificate: "
"%s" % exc)
cls._handle_fail(
req, link, reason, url,
cache=cache,
level=2,
meth=logger.notify,
)
else:
if cache is not None:
cache.add_page([url, resp.url], inst)
return inst
schemes.extend(backend.schemes)
return schemes
def register(self, cls):
if not hasattr(cls, 'name'):
<<<<<<< HEAD
logger.warn('Cannot register VCS %s' % cls.__name__)
return
if cls.name not in self._registry:
self._registry[cls.name] = cls
=======
logger.warning('Cannot register VCS %s', cls.__name__)
return
if cls.name not in self._registry:
self._registry[cls.name] = cls
logger.debug('Registered VCS backend: %s', cls.name)
>>>>>>> 54eef0be98b1b67c8507db91f4cfa90b64991027
def unregister(self, cls=None, name=None):
if name in self._registry:
del self._registry[name]
elif cls in self._registry.values():
del self._registry[cls.name]
else:
<<<<<<< HEAD
logger.warn('Cannot unregister because no class or name given')
=======
logger.warning('Cannot unregister because no class or name given')
>>>>>>> 54eef0be98b1b67c8507db91f4cfa90b64991027
def get_backend_name(self, location):
def _link_package_versions(self, link, search_name):
"""
Return an iterable of triples (pkg_resources_version_key,
link, python_version) that can be extracted from the given
link.
Meant to be overridden by subclasses, not called by clients.
"""
platform = get_platform()
version = None
if link.egg_fragment:
egg_info = link.egg_fragment
else:
egg_info, ext = link.splitext()
if not ext:
if link not in self.logged_links:
logger.debug('Skipping link %s; not a file' % link)
self.logged_links.add(link)
return []
if egg_info.endswith('.tar'):
# Special double-extension case:
egg_info = egg_info[:-4]
ext = '.tar' + ext
if ext not in self._known_extensions():
if link not in self.logged_links:
logger.debug(
'Skipping link %s; unknown archive format: %s' %
(link, ext)
)
self.logged_links.add(link)
return []
if "macosx10" in link.path and ext == '.zip':
if link not in self.logged_links:
logger.debug('Skipping link %s; macosx10 one' % (link))
self.logged_links.add(link)
return []
if ext == wheel_ext:
try:
wheel = Wheel(link.filename)
except InvalidWheelFilename:
logger.debug(
'Skipping %s because the wheel filename is invalid' %
link
)
return []
if wheel.name.lower() != search_name.lower():
logger.debug(
'Skipping link %s; wrong project name (not %s)' %
(link, search_name)
)
return []
if not wheel.supported():
logger.debug(
'Skipping %s because it is not compatible with this '
'Python' % link
)
return []
# This is a dirty hack to prevent installing Binary Wheels from
# PyPI unless it is a Windows or Mac Binary Wheel. This is
# paired with a change to PyPI disabling uploads for the
# same. Once we have a mechanism for enabling support for
# binary wheels on linux that deals with the inherent problems
# of binary distribution this can be removed.
comes_from = getattr(link, "comes_from", None)
if (
(
not platform.startswith('win')
and not platform.startswith('macosx')
)
and comes_from is not None
and urlparse.urlparse(
comes_from.url
).netloc.endswith("pypi.python.org")):
if not wheel.supported(tags=supported_tags_noarch):
logger.debug(
"Skipping %s because it is a pypi-hosted binary "
"Wheel on an unsupported platform" % link
)
return []
version = wheel.version
if not version:
version = self._egg_info_matches(egg_info, search_name, link)
if version is None:
logger.debug(
'Skipping link %s; wrong project name (not %s)' %
(link, search_name)
)
return []
if (link.internal is not None
and not link.internal
and not normalize_name(search_name).lower()
in self.allow_external
and not self.allow_all_external):
# We have a link that we are sure is external, so we should skip
# it unless we are allowing externals
logger.debug("Skipping %s because it is externally hosted." % link)
self.need_warn_external = True
return []
if (link.verifiable is not None
and not link.verifiable
and not (normalize_name(search_name).lower()
in self.allow_unverified)):
# We have a link that we are sure we cannot verify its integrity,
# so we should skip it unless we are allowing unsafe installs
# for this requirement.
logger.debug("Skipping %s because it is an insecure and "
"unverifiable file." % link)
self.need_warn_unverified = True
return []
match = self._py_version_re.search(version)
if match:
version = version[:match.start()]
py_version = match.group(1)
if py_version != sys.version[:3]:
logger.debug(
'Skipping %s because Python version is incorrect' % link
)
return []
logger.debug('Found link %s, version: %s' % (link, version))
return [(
pkg_resources.parse_version(version),
link,
version,
)]
def get_page(cls, link, req, cache=None, skip_archives=True):
url = link.url
url = url.split('#', 1)[0]
if cache.too_many_failures(url):
return None
# Check for VCS schemes that do not support lookup as web pages.
from pip.vcs import VcsSupport
for scheme in VcsSupport.schemes:
if url.lower().startswith(scheme) and url[len(scheme)] in '+:':
logger.debug('Cannot look at %(scheme)s URL %(link)s' %
locals())
return None
if cache is not None:
inst = cache.get_page(url)
if inst is not None:
return inst
try:
if skip_archives:
if cache is not None:
if cache.is_archive(url):
return None
filename = link.filename
for bad_ext in ['.tar', '.tar.gz', '.tar.bz2', '.tgz', '.zip']:
if filename.endswith(bad_ext):
content_type = cls._get_content_type(url)
if content_type.lower().startswith('text/html'):
break
else:
logger.debug(
'Skipping page %s because of Content-Type: %s'
% (link, content_type))
if cache is not None:
cache.set_is_archive(url)
return None
logger.debug('Getting page %s' % url)
# Tack index.html onto file:// URLs that point to directories
(scheme, netloc, path, params, query,
fragment) = urlparse.urlparse(url)
if scheme == 'file' and os.path.isdir(url2pathname(path)):
# add trailing slash if not present so urljoin doesn't trim final segment
if not url.endswith('/'):
url += '/'
url = urlparse.urljoin(url, 'index.html')
logger.debug(' file: URL is directory, getting %s' % url)
resp = urlopen(url)
real_url = geturl(resp)
headers = resp.info()
contents = resp.read()
encoding = headers.get('Content-Encoding', None)
#XXX need to handle exceptions and add testing for this
if encoding is not None:
if encoding == 'gzip':
contents = gzip.GzipFile(fileobj=BytesIO(contents)).read()
if encoding == 'deflate':
contents = zlib.decompress(contents)
inst = cls(u(contents), real_url, headers)
except (HTTPError, URLError, socket.timeout, socket.error, OSError,
WindowsError):
e = sys.exc_info()[1]
desc = str(e)
if isinstance(e, socket.timeout):
log_meth = logger.info
level = 1
desc = 'timed out'
elif isinstance(e, URLError):
log_meth = logger.info
if hasattr(e, 'reason') and isinstance(e.reason,
socket.timeout):
desc = 'timed out'
level = 1
else:
level = 2
elif isinstance(e, HTTPError) and e.code == 404:
## FIXME: notify?
log_meth = logger.info
level = 2
else:
log_meth = logger.info
level = 1
log_meth('Could not fetch URL %s: %s' % (link, desc))
log_meth(
'Will skip URL %s when looking for download links for %s' %
(link.url, req))
if cache is not None:
cache.add_page_failure(url, level)
return None
if cache is not None:
cache.add_page([url, real_url], inst)
return inst
changed = set()
generated = []
# Compile all of the pyc files that we're going to be installing
if pycompile:
<<<<<<< HEAD
compileall.compile_dir(source, force=True, quiet=True)
def normpath(src, p):
return make_path_relative(src, p).replace(os.path.sep, '/')
=======
with captured_stdout() as stdout:
with warnings.catch_warnings():
warnings.filterwarnings('ignore')
compileall.compile_dir(source, force=True, quiet=True)
logger.debug(stdout.getvalue())
def normpath(src, p):
return os.path.relpath(src, p).replace(os.path.sep, '/')
>>>>>>> 54eef0be98b1b67c8507db91f4cfa90b64991027
def record_installed(srcfile, destfile, modified=False):
"""Map archive RECORD paths to installation RECORD paths."""
oldpath = normpath(srcfile, wheeldir)
newpath = normpath(destfile, lib_dir)
installed[oldpath] = newpath
if modified:
changed.add(destfile)
def clobber(source, dest, is_base, fixer=None, filter=None):
<<<<<<< HEAD
def _link_package_versions(self, link, search_name):
"""
Return an iterable of triples (pkg_resources_version_key,
link, python_version) that can be extracted from the given
link.
Meant to be overridden by subclasses, not called by clients.
"""
platform = get_platform()
version = None
if link.egg_fragment:
egg_info = link.egg_fragment
else:
egg_info, ext = link.splitext()
if not ext:
if link not in self.logged_links:
logger.debug('Skipping link %s; not a file' % link)
self.logged_links.add(link)
return []
if egg_info.endswith('.tar'):
# Special double-extension case:
egg_info = egg_info[:-4]
ext = '.tar' + ext
if ext not in self._known_extensions():
if link not in self.logged_links:
logger.debug(
'Skipping link %s; unknown archive format: %s' %
(link, ext))
self.logged_links.add(link)
return []
if "macosx10" in link.path and ext == '.zip':
if link not in self.logged_links:
logger.debug('Skipping link %s; macosx10 one' % (link))
self.logged_links.add(link)
return []
if ext == wheel_ext:
try:
wheel = Wheel(link.filename)
except InvalidWheelFilename:
logger.debug(
'Skipping %s because the wheel filename is invalid' %
link)
return []
if wheel.name.lower() != search_name.lower():
logger.debug(
'Skipping link %s; wrong project name (not %s)' %
(link, search_name))
return []
if not wheel.supported():
logger.debug(
'Skipping %s because it is not compatible with this Python'
% link)
return []
# This is a dirty hack to prevent installing Binary Wheels from
# PyPI unless it is a Windows or Mac Binary Wheel. This is
# paired with a change to PyPI disabling uploads for the
# same. Once we have a mechanism for enabling support for binary
# wheels on linux that deals with the inherent problems of
# binary distribution this can be removed.
comes_from = getattr(link, "comes_from", None)
if ((not platform.startswith('win')
and not platform.startswith('macosx'))
and comes_from is not None and urlparse.urlparse(
comes_from.url).netloc.endswith("pypi.python.org")
):
if not wheel.supported(tags=supported_tags_noarch):
logger.debug(
"Skipping %s because it is a pypi-hosted binary "
"Wheel on an unsupported platform" % link)
return []
version = wheel.version
if not version:
version = self._egg_info_matches(egg_info, search_name, link)
if version is None:
logger.debug('Skipping link %s; wrong project name (not %s)' %
(link, search_name))
return []
if (link.internal is not None and not link.internal and
not normalize_name(search_name).lower() in self.allow_external
and not self.allow_all_external):
# We have a link that we are sure is external, so we should skip
# it unless we are allowing externals
logger.debug("Skipping %s because it is externally hosted." % link)
self.need_warn_external = True
return []
if (link.verifiable is not None and not link.verifiable
and not (normalize_name(search_name).lower()
in self.allow_unverified)
and not self.allow_all_unverified):
# We have a link that we are sure we cannot verify it's integrity,
# so we should skip it unless we are allowing unsafe installs
# for this requirement.
logger.debug("Skipping %s because it is an insecure and "
"unverifiable file." % link)
self.need_warn_unverified = True
return []
match = self._py_version_re.search(version)
if match:
version = version[:match.start()]
py_version = match.group(1)
if py_version != sys.version[:3]:
logger.debug(
'Skipping %s because Python version is incorrect' % link)
return []
logger.debug('Found link %s, version: %s' % (link, version))
return [(pkg_resources.parse_version(version), link, version)]
def get_page(cls, link, req, cache=None, skip_archives=True):
url = link.url
url = url.split('#', 1)[0]
if cache.too_many_failures(url):
return None
# Check for VCS schemes that do not support lookup as web pages.
from pip.vcs import VcsSupport
for scheme in VcsSupport.schemes:
if url.lower().startswith(scheme) and url[len(scheme)] in '+:':
logger.debug('Cannot look at %(scheme)s URL %(link)s' % locals())
return None
if cache is not None:
inst = cache.get_page(url)
if inst is not None:
return inst
try:
if skip_archives:
if cache is not None:
if cache.is_archive(url):
return None
filename = link.filename
for bad_ext in ['.tar', '.tar.gz', '.tar.bz2', '.tgz', '.zip']:
if filename.endswith(bad_ext):
content_type = cls._get_content_type(url)
if content_type.lower().startswith('text/html'):
break
else:
logger.debug('Skipping page %s because of Content-Type: %s' % (link, content_type))
if cache is not None:
cache.set_is_archive(url)
return None
logger.debug('Getting page %s' % url)
# Tack index.html onto file:// URLs that point to directories
(scheme, netloc, path, params, query, fragment) = urlparse.urlparse(url)
if scheme == 'file' and os.path.isdir(url2pathname(path)):
# add trailing slash if not present so urljoin doesn't trim final segment
if not url.endswith('/'):
url += '/'
url = urlparse.urljoin(url, 'index.html')
logger.debug(' file: URL is directory, getting %s' % url)
resp = urlopen(url)
real_url = geturl(resp)
headers = resp.info()
contents = resp.read()
encoding = headers.get('Content-Encoding', None)
#XXX need to handle exceptions and add testing for this
if encoding is not None:
if encoding == 'gzip':
contents = gzip.GzipFile(fileobj=BytesIO(contents)).read()
if encoding == 'deflate':
contents = zlib.decompress(contents)
inst = cls(u(contents), real_url, headers)
except (HTTPError, URLError, socket.timeout, socket.error, OSError, WindowsError):
e = sys.exc_info()[1]
desc = str(e)
if isinstance(e, socket.timeout):
log_meth = logger.info
level =1
desc = 'timed out'
elif isinstance(e, URLError):
log_meth = logger.info
if hasattr(e, 'reason') and isinstance(e.reason, socket.timeout):
desc = 'timed out'
level = 1
else:
level = 2
elif isinstance(e, HTTPError) and e.code == 404:
## FIXME: notify?
log_meth = logger.info
level = 2
else:
log_meth = logger.info
level = 1
log_meth('Could not fetch URL %s: %s' % (link, desc))
log_meth('Will skip URL %s when looking for download links for %s' % (link.url, req))
if cache is not None:
cache.add_page_failure(url, level)
return None
if cache is not None:
cache.add_page([url, real_url], inst)
return inst
def get_page(cls, link, req, cache=None, skip_archives=True, session=None):
if session is None:
session = PipSession()
url = link.url
url = url.split('#', 1)[0]
if cache.too_many_failures(url):
return None
# Check for VCS schemes that do not support lookup as web pages.
from pip.vcs import VcsSupport
for scheme in VcsSupport.schemes:
if url.lower().startswith(scheme) and url[len(scheme)] in '+:':
logger.debug('Cannot look at %(scheme)s URL %(link)s' %
locals())
return None
if cache is not None:
inst = cache.get_page(url)
if inst is not None:
return inst
try:
if skip_archives:
if cache is not None:
if cache.is_archive(url):
return None
filename = link.filename
for bad_ext in ['.tar', '.tar.gz', '.tar.bz2', '.tgz', '.zip']:
if filename.endswith(bad_ext):
content_type = cls._get_content_type(
url,
session=session,
)
if content_type.lower().startswith('text/html'):
break
else:
logger.debug(
'Skipping page %s because of Content-Type: %s'
% (link, content_type))
if cache is not None:
cache.set_is_archive(url)
return None
logger.debug('Getting page %s' % url)
# Tack index.html onto file:// URLs that point to directories
(scheme, netloc, path, params, query,
fragment) = urlparse.urlparse(url)
if scheme == 'file' and os.path.isdir(url2pathname(path)):
# add trailing slash if not present so urljoin doesn't trim final segment
if not url.endswith('/'):
url += '/'
url = urlparse.urljoin(url, 'index.html')
logger.debug(' file: URL is directory, getting %s' % url)
resp = session.get(url, headers={"Accept": "text/html"})
resp.raise_for_status()
# The check for archives above only works if the url ends with
# something that looks like an archive. However that is not a
# requirement. For instance http://sourceforge.net/projects/docutils/files/docutils/0.8.1/docutils-0.8.1.tar.gz/download
# redirects to http://superb-dca3.dl.sourceforge.net/project/docutils/docutils/0.8.1/docutils-0.8.1.tar.gz
# Unless we issue a HEAD request on every url we cannot know
# ahead of time for sure if something is HTML or not. However we
# can check after we've downloaded it.
content_type = resp.headers.get('Content-Type', 'unknown')
if not content_type.lower().startswith("text/html"):
logger.debug('Skipping page %s because of Content-Type: %s' %
(link, content_type))
if cache is not None:
cache.set_is_archive(url)
return None
inst = cls(resp.text, resp.url, resp.headers, trusted=link.trusted)
except requests.HTTPError as exc:
level = 2 if exc.response.status_code == 404 else 1
cls._handle_fail(req, link, exc, url, cache=cache, level=level)
except requests.ConnectionError as exc:
cls._handle_fail(
req,
link,
"connection error: %s" % exc,
url,
cache=cache,
)
except requests.Timeout:
cls._handle_fail(req, link, "timed out", url, cache=cache)
except SSLError as exc:
reason = ("There was a problem confirming the ssl certificate: "
"%s" % exc)
cls._handle_fail(
req,
link,
reason,
url,
cache=cache,
level=2,
meth=logger.notify,
)
except requests.TooManyRedirects as exc:
cls._handle_fail(
req,
link,
"Error: %s" % exc,
url,
cache=cache,
)
else:
if cache is not None:
cache.add_page([url, resp.url], inst)
return inst
def find_requirement(self, req, upgrade):
url_name = req.url_name
# Only check main index if index URL is given:
main_index_url = None
if self.index_urls:
# Check that we have the url_name correctly spelled:
main_index_url = Link(posixpath.join(self.index_urls[0], url_name))
# This will also cache the page, so it's okay that we get it again later:
page = self._get_page(main_index_url, req)
if page is None:
url_name = self._find_url_name(Link(self.index_urls[0]), url_name, req) or req.url_name
# Combine index URLs with mirror URLs here to allow
# adding more index URLs from requirements files
all_index_urls = self.index_urls + self.mirror_urls
def mkurl_pypi_url(url):
loc = posixpath.join(url, url_name)
# For maximum compatibility with easy_install, ensure the path
# ends in a trailing slash. Although this isn't in the spec
# (and PyPI can handle it without the slash) some other index
# implementations might break if they relied on easy_install's behavior.
if not loc.endswith('/'):
loc = loc + '/'
return loc
if url_name is not None:
locations = [
mkurl_pypi_url(url)
for url in all_index_urls] + self.find_links
else:
locations = list(self.find_links)
locations.extend(self.dependency_links)
for version in req.absolute_versions:
if url_name is not None and main_index_url is not None:
locations = [
posixpath.join(main_index_url.url, version)] + locations
file_locations, url_locations = self._sort_locations(locations)
locations = [Link(url) for url in url_locations]
logger.debug('URLs to search for versions for %s:' % req)
for location in locations:
logger.debug('* %s' % location)
found_versions = []
found_versions.extend(
self._package_versions(
[Link(url, '-f') for url in self.find_links], req.name.lower()))
page_versions = []
for page in self._get_pages(locations, req):
logger.debug('Analyzing links from page %s' % page.url)
logger.indent += 2
try:
page_versions.extend(self._package_versions(page.links, req.name.lower()))
finally:
logger.indent -= 2
dependency_versions = list(self._package_versions(
[Link(url) for url in self.dependency_links], req.name.lower()))
if dependency_versions:
logger.info('dependency_links found: %s' % ', '.join([link.url for parsed, link, version in dependency_versions]))
file_versions = list(self._package_versions(
[Link(url) for url in file_locations], req.name.lower()))
if not found_versions and not page_versions and not dependency_versions and not file_versions:
logger.fatal('Could not find any downloads that satisfy the requirement %s' % req)
raise DistributionNotFound('No distributions at all found for %s' % req)
if req.satisfied_by is not None:
found_versions.append((req.satisfied_by.parsed_version, Inf, req.satisfied_by.version))
if file_versions:
file_versions.sort(reverse=True)
logger.info('Local files found: %s' % ', '.join([url_to_path(link.url) for parsed, link, version in file_versions]))
found_versions = file_versions + found_versions
all_versions = found_versions + page_versions + dependency_versions
applicable_versions = []
for (parsed_version, link, version) in all_versions:
if version not in req.req:
logger.info("Ignoring link %s, version %s doesn't match %s"
% (link, version, ','.join([''.join(s) for s in req.req.specs])))
continue
applicable_versions.append((link, version))
applicable_versions = sorted(applicable_versions, key=lambda v: pkg_resources.parse_version(v[1]), reverse=True)
existing_applicable = bool([link for link, version in applicable_versions if link is Inf])
if not upgrade and existing_applicable:
if applicable_versions[0][1] is Inf:
logger.info('Existing installed version (%s) is most up-to-date and satisfies requirement'
% req.satisfied_by.version)
raise BestVersionAlreadyInstalled
else:
logger.info('Existing installed version (%s) satisfies requirement (most up-to-date version is %s)'
% (req.satisfied_by.version, applicable_versions[0][1]))
return None
if not applicable_versions:
logger.fatal('Could not find a version that satisfies the requirement %s (from versions: %s)'
% (req, ', '.join([version for parsed_version, link, version in found_versions])))
raise DistributionNotFound('No distributions matching the version for %s' % req)
if applicable_versions[0][0] is Inf:
# We have an existing version, and its the best version
logger.info('Installed version (%s) is most up-to-date (past versions: %s)'
% (req.satisfied_by.version, ', '.join([version for link, version in applicable_versions[1:]]) or 'none'))
raise BestVersionAlreadyInstalled
if len(applicable_versions) > 1:
logger.info('Using version %s (newest of versions: %s)' %
(applicable_versions[0][1], ', '.join([version for link, version in applicable_versions])))
return applicable_versions[0][0]
def find_requirement(self, req, upgrade):
def mkurl_pypi_url(url):
loc = posixpath.join(url, url_name)
# For maximum compatibility with easy_install, ensure the path
# ends in a trailing slash. Although this isn't in the spec
# (and PyPI can handle it without the slash) some other index
# implementations might break if they relied on easy_install's
# behavior.
if not loc.endswith('/'):
loc = loc + '/'
return loc
url_name = req.url_name
# Only check main index if index URL is given:
main_index_url = None
if self.index_urls:
# Check that we have the url_name correctly spelled:
main_index_url = Link(
mkurl_pypi_url(self.index_urls[0]),
trusted=True,
)
# This will also cache the page, so it's okay that we get it again
# later:
page = self._get_page(main_index_url, req)
if page is None:
url_name = self._find_url_name(
Link(self.index_urls[0], trusted=True),
url_name, req
) or req.url_name
if url_name is not None:
locations = [
mkurl_pypi_url(url)
for url in self.index_urls] + self.find_links
else:
locations = list(self.find_links)
for version in req.absolute_versions:
if url_name is not None and main_index_url is not None:
locations = [
posixpath.join(main_index_url.url, version)] + locations
file_locations, url_locations = self._sort_locations(locations)
# We trust every url that the user has given us whether it was given
# via --index-url or --find-links
locations = [Link(url, trusted=True) for url in url_locations]
logger.debug('URLs to search for versions for %s:' % req)
for location in locations:
logger.debug('* %s' % location)
# Determine if this url used a secure transport mechanism
parsed = urlparse.urlparse(str(location))
if parsed.scheme in INSECURE_SCHEMES:
secure_schemes = INSECURE_SCHEMES[parsed.scheme]
if len(secure_schemes) == 1:
ctx = (location, parsed.scheme, secure_schemes[0],
parsed.netloc)
logger.warn("%s uses an insecure transport scheme (%s). "
"Consider using %s if %s has it available" %
ctx)
elif len(secure_schemes) > 1:
ctx = (
location,
parsed.scheme,
", ".join(secure_schemes),
parsed.netloc,
)
logger.warn("%s uses an insecure transport scheme (%s). "
"Consider using one of %s if %s has any of "
"them available" % ctx)
else:
ctx = (location, parsed.scheme)
logger.warn("%s uses an insecure transport scheme (%s)." %
ctx)
found_versions = []
found_versions.extend(
self._package_versions(
# We trust every directly linked archive in find_links
[Link(url, '-f', trusted=True) for url in self.find_links],
req.name.lower()
)
)
page_versions = []
for page in self._get_pages(locations, req):
logger.debug('Analyzing links from page %s' % page.url)
logger.indent += 2
try:
page_versions.extend(
self._package_versions(page.links, req.name.lower())
)
finally:
logger.indent -= 2
file_versions = list(
self._package_versions(
[Link(url) for url in file_locations],
req.name.lower()
)
)
if (not found_versions
and not page_versions
and not file_versions):
logger.fatal(
'Could not find any downloads that satisfy the requirement'
' %s' % req
)
if self.need_warn_external:
logger.warn("Some externally hosted files were ignored (use "
"--allow-external %s to allow)." % req.name)
if self.need_warn_unverified:
logger.warn("Some insecure and unverifiable files were ignored"
" (use --allow-unverified %s to allow)." %
req.name)
raise DistributionNotFound(
'No distributions at all found for %s' % req
)
installed_version = []
if req.satisfied_by is not None:
installed_version = [(
req.satisfied_by.parsed_version,
INSTALLED_VERSION,
req.satisfied_by.version,
)]
if file_versions:
file_versions.sort(reverse=True)
logger.info(
'Local files found: %s' %
', '.join([
url_to_path(link.url)
for parsed, link, version in file_versions
])
)
# this is an intentional priority ordering
all_versions = installed_version + file_versions + found_versions \
+ page_versions
applicable_versions = []
for (parsed_version, link, version) in all_versions:
if version not in req.req:
logger.info(
"Ignoring link %s, version %s doesn't match %s" %
(
link,
version,
','.join([''.join(s) for s in req.req.specs])
)
)
continue
elif (is_prerelease(version)
and not (self.allow_all_prereleases or req.prereleases)):
# If this version isn't the already installed one, then
# ignore it if it's a pre-release.
if link is not INSTALLED_VERSION:
logger.info(
"Ignoring link %s, version %s is a pre-release (use "
"--pre to allow)." % (link, version)
)
continue
applicable_versions.append((parsed_version, link, version))
applicable_versions = self._sort_versions(applicable_versions)
existing_applicable = bool([
link
for parsed_version, link, version in applicable_versions
if link is INSTALLED_VERSION
])
if not upgrade and existing_applicable:
if applicable_versions[0][1] is INSTALLED_VERSION:
logger.info(
'Existing installed version (%s) is most up-to-date and '
'satisfies requirement' % req.satisfied_by.version
)
else:
logger.info(
'Existing installed version (%s) satisfies requirement '
'(most up-to-date version is %s)' %
(req.satisfied_by.version, applicable_versions[0][2])
)
return None
if not applicable_versions:
logger.fatal(
'Could not find a version that satisfies the requirement %s '
'(from versions: %s)' %
(
req,
', '.join([
version
for parsed_version, link, version in all_versions
])
)
)
if self.need_warn_external:
logger.warn("Some externally hosted files were ignored (use "
"--allow-external to allow).")
if self.need_warn_unverified:
logger.warn("Some insecure and unverifiable files were ignored"
" (use --allow-unverified %s to allow)." %
req.name)
raise DistributionNotFound(
'No distributions matching the version for %s' % req
)
if applicable_versions[0][1] is INSTALLED_VERSION:
# We have an existing version, and its the best version
logger.info(
'Installed version (%s) is most up-to-date (past versions: '
'%s)' % (
req.satisfied_by.version,
', '.join([
version for parsed_version, link, version
in applicable_versions[1:]
]) or 'none'))
raise BestVersionAlreadyInstalled
if len(applicable_versions) > 1:
logger.info(
'Using version %s (newest of versions: %s)' %
(
applicable_versions[0][2],
', '.join([
version for parsed_version, link, version
in applicable_versions
])
)
)
selected_version = applicable_versions[0][1]
if (selected_version.internal is not None
and not selected_version.internal):
logger.warn("%s an externally hosted file and may be "
"unreliable" % req.name)
if (selected_version.verifiable is not None
and not selected_version.verifiable):
logger.warn("%s is potentially insecure and "
"unverifiable." % req.name)
if selected_version._deprecated_regex:
logger.deprecated(
"1.7",
"%s discovered using a deprecated method of parsing, "
"in the future it will no longer be discovered" % req.name
)
return selected_version
def get_page(cls, link, req, cache=None, skip_archives=True):
url = link.url
url = url.split('#', 1)[0]
if cache.too_many_failures(url):
return None
# Check for VCS schemes that do not support lookup as web pages.
from pip.vcs import VcsSupport
for scheme in VcsSupport.schemes:
if url.lower().startswith(scheme) and url[len(scheme)] in '+:':
logger.debug('Cannot look at %(scheme)s URL %(link)s' % locals())
return None
if cache is not None:
inst = cache.get_page(url)
if inst is not None:
return inst
try:
if skip_archives:
if cache is not None:
if cache.is_archive(url):
return None
filename = link.filename
for bad_ext in ['.tar', '.tar.gz', '.tar.bz2', '.tgz', '.zip']:
if filename.endswith(bad_ext):
content_type = cls._get_content_type(url)
if content_type.lower().startswith('text/html'):
break
else:
logger.debug('Skipping page %s because of Content-Type: %s' % (link, content_type))
if cache is not None:
cache.set_is_archive(url)
return None
logger.debug('Getting page %s' % url)
# Tack index.html onto file:// URLs that point to directories
(scheme, netloc, path, params, query, fragment) = urlparse.urlparse(url)
if scheme == 'file' and os.path.isdir(url2pathname(path)):
# add trailing slash if not present so urljoin doesn't trim final segment
if not url.endswith('/'):
url += '/'
url = urlparse.urljoin(url, 'index.html')
logger.debug(' file: URL is directory, getting %s' % url)
resp = urlopen(url)
real_url = geturl(resp)
headers = resp.info()
contents = resp.read()
encoding = headers.get('Content-Encoding', None)
#XXX need to handle exceptions and add testing for this
if encoding is not None:
if encoding == 'gzip':
contents = gzip.GzipFile(fileobj=BytesIO(contents)).read()
if encoding == 'deflate':
contents = zlib.decompress(contents)
# The check for archives above only works if the url ends with
# something that looks like an archive. However that is not a
# requirement. For instance http://sourceforge.net/projects/docutils/files/docutils/0.8.1/docutils-0.8.1.tar.gz/download
# redirects to http://superb-dca3.dl.sourceforge.net/project/docutils/docutils/0.8.1/docutils-0.8.1.tar.gz
# Unless we issue a HEAD request on every url we cannot know
# ahead of time for sure if something is HTML or not. However we
# can check after we've downloaded it.
content_type = headers.get('Content-Type', 'unknown')
if not content_type.lower().startswith("text/html"):
logger.debug('Skipping page %s because of Content-Type: %s' %
(link, content_type))
if cache is not None:
cache.set_is_archive(url)
return None
inst = cls(u(contents), real_url, headers, trusted=link.trusted)
except (HTTPError, URLError, socket.timeout, socket.error, OSError, WindowsError):
e = sys.exc_info()[1]
desc = str(e)
if isinstance(e, socket.timeout):
log_meth = logger.info
level =1
desc = 'timed out'
elif isinstance(e, URLError):
#ssl/certificate error
if hasattr(e, 'reason') and (isinstance(e.reason, ssl.SSLError) or isinstance(e.reason, CertificateError)):
desc = 'There was a problem confirming the ssl certificate: %s' % e
log_meth = logger.notify
else:
log_meth = logger.info
if hasattr(e, 'reason') and isinstance(e.reason, socket.timeout):
desc = 'timed out'
level = 1
else:
level = 2
elif isinstance(e, HTTPError) and e.code == 404:
## FIXME: notify?
log_meth = logger.info
level = 2
else:
log_meth = logger.info
level = 1
log_meth('Could not fetch URL %s: %s' % (link, desc))
log_meth('Will skip URL %s when looking for download links for %s' % (link.url, req))
if cache is not None:
cache.add_page_failure(url, level)
return None
if cache is not None:
cache.add_page([url, real_url], inst)
return inst
def get_page(cls, link, req, cache=None, skip_archives=True, session=None):
if session is None:
session = PipSession()
url = link.url
url = url.split("#", 1)[0]
if cache.too_many_failures(url):
return None
# Check for VCS schemes that do not support lookup as web pages.
from pip.vcs import VcsSupport
for scheme in VcsSupport.schemes:
if url.lower().startswith(scheme) and url[len(scheme)] in "+:":
logger.debug("Cannot look at %(scheme)s URL %(link)s" % locals())
return None
if cache is not None:
inst = cache.get_page(url)
if inst is not None:
return inst
try:
if skip_archives:
if cache is not None:
if cache.is_archive(url):
return None
filename = link.filename
for bad_ext in [".tar", ".tar.gz", ".tar.bz2", ".tgz", ".zip"]:
if filename.endswith(bad_ext):
content_type = cls._get_content_type(url, session=session)
if content_type.lower().startswith("text/html"):
break
else:
logger.debug("Skipping page %s because of Content-Type: %s" % (link, content_type))
if cache is not None:
cache.set_is_archive(url)
return None
logger.debug("Getting page %s" % url)
# Tack index.html onto file:// URLs that point to directories
(scheme, netloc, path, params, query, fragment) = urlparse.urlparse(url)
if scheme == "file" and os.path.isdir(url2pathname(path)):
# add trailing slash if not present so urljoin doesn't trim final segment
if not url.endswith("/"):
url += "/"
url = urlparse.urljoin(url, "index.html")
logger.debug(" file: URL is directory, getting %s" % url)
resp = session.get(url)
resp.raise_for_status()
# The check for archives above only works if the url ends with
# something that looks like an archive. However that is not a
# requirement. For instance http://sourceforge.net/projects/docutils/files/docutils/0.8.1/docutils-0.8.1.tar.gz/download
# redirects to http://superb-dca3.dl.sourceforge.net/project/docutils/docutils/0.8.1/docutils-0.8.1.tar.gz
# Unless we issue a HEAD request on every url we cannot know
# ahead of time for sure if something is HTML or not. However we
# can check after we've downloaded it.
content_type = resp.headers.get("Content-Type", "unknown")
if not content_type.lower().startswith("text/html"):
logger.debug("Skipping page %s because of Content-Type: %s" % (link, content_type))
if cache is not None:
cache.set_is_archive(url)
return None
inst = cls(resp.text, resp.url, resp.headers, trusted=link.trusted)
except requests.HTTPError as exc:
level = 2 if exc.response.status_code == 404 else 1
cls._handle_fail(req, link, exc, url, cache=cache, level=level)
except requests.Timeout:
cls._handle_fail(req, link, "timed out", url, cache=cache)
except SSLError as exc:
reason = "There was a problem confirming the ssl certificate: " "%s" % exc
cls._handle_fail(req, link, reason, url, cache=cache, level=2, meth=logger.notify)
else:
if cache is not None:
cache.add_page([url, resp.url], inst)
return inst
def find_requirement(self, req, upgrade):
url_name = req.url_name
# Only check main index if index URL is given:
main_index_url = None
if self.index_urls:
# Check that we have the url_name correctly spelled:
main_index_url = Link(posixpath.join(self.index_urls[0], url_name))
# This will also cache the page, so it's okay that we get it again later:
page = self._get_page(main_index_url, req)
if page is None:
url_name = self._find_url_name(Link(self.index_urls[0]),
url_name, req) or req.url_name
# Combine index URLs with mirror URLs here to allow
# adding more index URLs from requirements files
all_index_urls = self.index_urls + self.mirror_urls
def mkurl_pypi_url(url):
loc = posixpath.join(url, url_name)
# For maximum compatibility with easy_install, ensure the path
# ends in a trailing slash. Although this isn't in the spec
# (and PyPI can handle it without the slash) some other index
# implementations might break if they relied on easy_install's behavior.
if not loc.endswith('/'):
loc = loc + '/'
return loc
if url_name is not None:
locations = [mkurl_pypi_url(url)
for url in all_index_urls] + self.find_links
else:
locations = list(self.find_links)
locations.extend(self.dependency_links)
for version in req.absolute_versions:
if url_name is not None and main_index_url is not None:
locations = [posixpath.join(main_index_url.url, version)
] + locations
file_locations, url_locations = self._sort_locations(locations)
locations = [Link(url) for url in url_locations]
logger.debug('URLs to search for versions for %s:' % req)
for location in locations:
logger.debug('* %s' % location)
found_versions = []
found_versions.extend(
self._package_versions(
[Link(url, '-f') for url in self.find_links],
req.name.lower()))
page_versions = []
for page in self._get_pages(locations, req):
logger.debug('Analyzing links from page %s' % page.url)
logger.indent += 2
try:
page_versions.extend(
self._package_versions(page.links, req.name.lower()))
finally:
logger.indent -= 2
dependency_versions = list(
self._package_versions(
[Link(url) for url in self.dependency_links],
req.name.lower()))
if dependency_versions:
logger.info('dependency_links found: %s' % ', '.join(
[link.url for parsed, link, version in dependency_versions]))
file_versions = list(
self._package_versions([Link(url) for url in file_locations],
req.name.lower()))
if not found_versions and not page_versions and not dependency_versions and not file_versions:
logger.fatal(
'Could not find any downloads that satisfy the requirement %s'
% req)
raise DistributionNotFound('No distributions at all found for %s' %
req)
if req.satisfied_by is not None:
found_versions.append((req.satisfied_by.parsed_version, Inf,
req.satisfied_by.version))
if file_versions:
file_versions.sort(reverse=True)
logger.info('Local files found: %s' % ', '.join([
url_to_path(link.url)
for parsed, link, version in file_versions
]))
found_versions = file_versions + found_versions
all_versions = found_versions + page_versions + dependency_versions
applicable_versions = []
for (parsed_version, link, version) in all_versions:
if version not in req.req:
logger.info("Ignoring link %s, version %s doesn't match %s" %
(link, version, ','.join(
[''.join(s) for s in req.req.specs])))
continue
applicable_versions.append((link, version))
applicable_versions = sorted(
applicable_versions,
key=lambda v: pkg_resources.parse_version(v[1]),
reverse=True)
existing_applicable = bool(
[link for link, version in applicable_versions if link is Inf])
if not upgrade and existing_applicable:
if applicable_versions[0][1] is Inf:
logger.info(
'Existing installed version (%s) is most up-to-date and satisfies requirement'
% req.satisfied_by.version)
raise BestVersionAlreadyInstalled
else:
logger.info(
'Existing installed version (%s) satisfies requirement (most up-to-date version is %s)'
% (req.satisfied_by.version, applicable_versions[0][1]))
return None
if not applicable_versions:
logger.fatal(
'Could not find a version that satisfies the requirement %s (from versions: %s)'
% (req, ', '.join([
version for parsed_version, link, version in found_versions
])))
raise DistributionNotFound(
'No distributions matching the version for %s' % req)
if applicable_versions[0][0] is Inf:
# We have an existing version, and its the best version
logger.info(
'Installed version (%s) is most up-to-date (past versions: %s)'
% (req.satisfied_by.version, ', '.join(
[version
for link, version in applicable_versions[1:]]) or 'none'))
raise BestVersionAlreadyInstalled
if len(applicable_versions) > 1:
logger.info('Using version %s (newest of versions: %s)' %
(applicable_versions[0][1], ', '.join(
[version
for link, version in applicable_versions])))
return applicable_versions[0][0]
def release_in_index(self, dist, index_url):
import xmlrpclib
server = xmlrpclib.Server(index_url)
req = pkg_resources.Requirement.parse(dist)
logger.debug('Requirement for %r is %r', dist, req)
releases = server.package_releases(req.project_name)
logger.debug('Checking if %r are in %r', releases, req)
good_releases = sorted((r for r in releases if r in req),
key=pkg_resources.parse_version)
if not good_releases:
logger.debug('Oops, none of %r are in %r', releases, req)
raise DistributionNotFound()
logger.debug('Yay, %r are in %r', good_releases, req)
best_release = good_releases.pop(-1)
logger.debug('Yay, asking for %r %r', req.project_name, best_release)
release = server.release_data(req.project_name, best_release)
release['_pip_index_url'] = index_url
release['_pip_other_versions'] = [r for r in releases if r != best_release]
logger.debug('Other versions are %r', release['_pip_other_versions'])
return release
