Initial commit. Basic models mostly done.

This commit is contained in:
tcaxle
2020-04-11 13:03:48 +01:00
commit 840e3c86f9
5761 changed files with 650959 additions and 0 deletions

View File

@@ -0,0 +1,2 @@
"""Contains purely network-related utilities.
"""

View File

@@ -0,0 +1,298 @@
"""Network Authentication Helpers
Contains interface (MultiDomainBasicAuth) and associated glue code for
providing credentials in the context of network requests.
"""
# The following comment should be removed at some point in the future.
# mypy: disallow-untyped-defs=False
import logging
from pip._vendor.requests.auth import AuthBase, HTTPBasicAuth
from pip._vendor.requests.utils import get_netrc_auth
from pip._vendor.six.moves.urllib import parse as urllib_parse
from pip._internal.utils.misc import (
ask,
ask_input,
ask_password,
remove_auth_from_url,
split_auth_netloc_from_url,
)
from pip._internal.utils.typing import MYPY_CHECK_RUNNING
if MYPY_CHECK_RUNNING:
from optparse import Values
from typing import Dict, Optional, Tuple
from pip._internal.vcs.versioncontrol import AuthInfo
Credentials = Tuple[str, str, str]
logger = logging.getLogger(__name__)
try:
import keyring # noqa
except ImportError:
keyring = None
except Exception as exc:
logger.warning(
"Keyring is skipped due to an exception: %s", str(exc),
)
keyring = None
def get_keyring_auth(url, username):
"""Return the tuple auth for a given url from keyring."""
if not url or not keyring:
return None
try:
try:
get_credential = keyring.get_credential
except AttributeError:
pass
else:
logger.debug("Getting credentials from keyring for %s", url)
cred = get_credential(url, username)
if cred is not None:
return cred.username, cred.password
return None
if username:
logger.debug("Getting password from keyring for %s", url)
password = keyring.get_password(url, username)
if password:
return username, password
except Exception as exc:
logger.warning(
"Keyring is skipped due to an exception: %s", str(exc),
)
class MultiDomainBasicAuth(AuthBase):
def __init__(self, prompting=True, index_urls=None):
# type: (bool, Optional[Values]) -> None
self.prompting = prompting
self.index_urls = index_urls
self.passwords = {} # type: Dict[str, AuthInfo]
# When the user is prompted to enter credentials and keyring is
# available, we will offer to save them. If the user accepts,
# this value is set to the credentials they entered. After the
# request authenticates, the caller should call
# ``save_credentials`` to save these.
self._credentials_to_save = None # type: Optional[Credentials]
def _get_index_url(self, url):
"""Return the original index URL matching the requested URL.
Cached or dynamically generated credentials may work against
the original index URL rather than just the netloc.
The provided url should have had its username and password
removed already. If the original index url had credentials then
they will be included in the return value.
Returns None if no matching index was found, or if --no-index
was specified by the user.
"""
if not url or not self.index_urls:
return None
for u in self.index_urls:
prefix = remove_auth_from_url(u).rstrip("/") + "/"
if url.startswith(prefix):
return u
def _get_new_credentials(self, original_url, allow_netrc=True,
allow_keyring=True):
"""Find and return credentials for the specified URL."""
# Split the credentials and netloc from the url.
url, netloc, url_user_password = split_auth_netloc_from_url(
original_url,
)
# Start with the credentials embedded in the url
username, password = url_user_password
if username is not None and password is not None:
logger.debug("Found credentials in url for %s", netloc)
return url_user_password
# Find a matching index url for this request
index_url = self._get_index_url(url)
if index_url:
# Split the credentials from the url.
index_info = split_auth_netloc_from_url(index_url)
if index_info:
index_url, _, index_url_user_password = index_info
logger.debug("Found index url %s", index_url)
# If an index URL was found, try its embedded credentials
if index_url and index_url_user_password[0] is not None:
username, password = index_url_user_password
if username is not None and password is not None:
logger.debug("Found credentials in index url for %s", netloc)
return index_url_user_password
# Get creds from netrc if we still don't have them
if allow_netrc:
netrc_auth = get_netrc_auth(original_url)
if netrc_auth:
logger.debug("Found credentials in netrc for %s", netloc)
return netrc_auth
# If we don't have a password and keyring is available, use it.
if allow_keyring:
# The index url is more specific than the netloc, so try it first
kr_auth = (
get_keyring_auth(index_url, username) or
get_keyring_auth(netloc, username)
)
if kr_auth:
logger.debug("Found credentials in keyring for %s", netloc)
return kr_auth
return username, password
def _get_url_and_credentials(self, original_url):
"""Return the credentials to use for the provided URL.
If allowed, netrc and keyring may be used to obtain the
correct credentials.
Returns (url_without_credentials, username, password). Note
that even if the original URL contains credentials, this
function may return a different username and password.
"""
url, netloc, _ = split_auth_netloc_from_url(original_url)
# Use any stored credentials that we have for this netloc
username, password = self.passwords.get(netloc, (None, None))
if username is None and password is None:
# No stored credentials. Acquire new credentials without prompting
# the user. (e.g. from netrc, keyring, or the URL itself)
username, password = self._get_new_credentials(original_url)
if username is not None or password is not None:
# Convert the username and password if they're None, so that
# this netloc will show up as "cached" in the conditional above.
# Further, HTTPBasicAuth doesn't accept None, so it makes sense to
# cache the value that is going to be used.
username = username or ""
password = password or ""
# Store any acquired credentials.
self.passwords[netloc] = (username, password)
assert (
# Credentials were found
(username is not None and password is not None) or
# Credentials were not found
(username is None and password is None)
), "Could not load credentials from url: {}".format(original_url)
return url, username, password
def __call__(self, req):
# Get credentials for this request
url, username, password = self._get_url_and_credentials(req.url)
# Set the url of the request to the url without any credentials
req.url = url
if username is not None and password is not None:
# Send the basic auth with this request
req = HTTPBasicAuth(username, password)(req)
# Attach a hook to handle 401 responses
req.register_hook("response", self.handle_401)
return req
# Factored out to allow for easy patching in tests
def _prompt_for_password(self, netloc):
username = ask_input("User for %s: " % netloc)
if not username:
return None, None
auth = get_keyring_auth(netloc, username)
if auth:
return auth[0], auth[1], False
password = ask_password("Password: ")
return username, password, True
# Factored out to allow for easy patching in tests
def _should_save_password_to_keyring(self):
if not keyring:
return False
return ask("Save credentials to keyring [y/N]: ", ["y", "n"]) == "y"
def handle_401(self, resp, **kwargs):
# We only care about 401 responses, anything else we want to just
# pass through the actual response
if resp.status_code != 401:
return resp
# We are not able to prompt the user so simply return the response
if not self.prompting:
return resp
parsed = urllib_parse.urlparse(resp.url)
# Prompt the user for a new username and password
username, password, save = self._prompt_for_password(parsed.netloc)
# Store the new username and password to use for future requests
self._credentials_to_save = None
if username is not None and password is not None:
self.passwords[parsed.netloc] = (username, password)
# Prompt to save the password to keyring
if save and self._should_save_password_to_keyring():
self._credentials_to_save = (parsed.netloc, username, password)
# Consume content and release the original connection to allow our new
# request to reuse the same one.
resp.content
resp.raw.release_conn()
# Add our new username and password to the request
req = HTTPBasicAuth(username or "", password or "")(resp.request)
req.register_hook("response", self.warn_on_401)
# On successful request, save the credentials that were used to
# keyring. (Note that if the user responded "no" above, this member
# is not set and nothing will be saved.)
if self._credentials_to_save:
req.register_hook("response", self.save_credentials)
# Send our new request
new_resp = resp.connection.send(req, **kwargs)
new_resp.history.append(resp)
return new_resp
def warn_on_401(self, resp, **kwargs):
"""Response callback to warn about incorrect credentials."""
if resp.status_code == 401:
logger.warning(
'401 Error, Credentials not correct for %s', resp.request.url,
)
def save_credentials(self, resp, **kwargs):
"""Response callback to save credentials on success."""
assert keyring is not None, "should never reach here without keyring"
if not keyring:
return
creds = self._credentials_to_save
self._credentials_to_save = None
if creds and resp.status_code < 400:
try:
logger.info('Saving credentials to keyring')
keyring.set_password(*creds)
except Exception:
logger.exception('Failed to save credentials')

View File

@@ -0,0 +1,81 @@
"""HTTP cache implementation.
"""
# The following comment should be removed at some point in the future.
# mypy: disallow-untyped-defs=False
import os
from contextlib import contextmanager
from pip._vendor.cachecontrol.cache import BaseCache
from pip._vendor.cachecontrol.caches import FileCache
from pip._vendor.requests.models import Response
from pip._internal.utils.filesystem import adjacent_tmp_file, replace
from pip._internal.utils.misc import ensure_dir
from pip._internal.utils.typing import MYPY_CHECK_RUNNING
if MYPY_CHECK_RUNNING:
from typing import Optional
def is_from_cache(response):
# type: (Response) -> bool
return getattr(response, "from_cache", False)
@contextmanager
def suppressed_cache_errors():
"""If we can't access the cache then we can just skip caching and process
requests as if caching wasn't enabled.
"""
try:
yield
except (OSError, IOError):
pass
class SafeFileCache(BaseCache):
"""
A file based cache which is safe to use even when the target directory may
not be accessible or writable.
"""
def __init__(self, directory):
# type: (str) -> None
assert directory is not None, "Cache directory must not be None."
super(SafeFileCache, self).__init__()
self.directory = directory
def _get_cache_path(self, name):
# type: (str) -> str
# From cachecontrol.caches.file_cache.FileCache._fn, brought into our
# class for backwards-compatibility and to avoid using a non-public
# method.
hashed = FileCache.encode(name)
parts = list(hashed[:5]) + [hashed]
return os.path.join(self.directory, *parts)
def get(self, key):
# type: (str) -> Optional[bytes]
path = self._get_cache_path(key)
with suppressed_cache_errors():
with open(path, 'rb') as f:
return f.read()
def set(self, key, value):
# type: (str, bytes) -> None
path = self._get_cache_path(key)
with suppressed_cache_errors():
ensure_dir(os.path.dirname(path))
with adjacent_tmp_file(path) as f:
f.write(value)
replace(f.name, path)
def delete(self, key):
# type: (str) -> None
path = self._get_cache_path(key)
with suppressed_cache_errors():
os.remove(path)

View File

@@ -0,0 +1,200 @@
"""Download files with progress indicators.
"""
import cgi
import logging
import mimetypes
import os
from pip._vendor import requests
from pip._vendor.requests.models import CONTENT_CHUNK_SIZE
from pip._internal.models.index import PyPI
from pip._internal.network.cache import is_from_cache
from pip._internal.network.utils import response_chunks
from pip._internal.utils.misc import (
format_size,
redact_auth_from_url,
splitext,
)
from pip._internal.utils.typing import MYPY_CHECK_RUNNING
from pip._internal.utils.ui import DownloadProgressProvider
if MYPY_CHECK_RUNNING:
from typing import Iterable, Optional
from pip._vendor.requests.models import Response
from pip._internal.models.link import Link
from pip._internal.network.session import PipSession
logger = logging.getLogger(__name__)
def _get_http_response_size(resp):
# type: (Response) -> Optional[int]
try:
return int(resp.headers['content-length'])
except (ValueError, KeyError, TypeError):
return None
def _prepare_download(
resp, # type: Response
link, # type: Link
progress_bar # type: str
):
# type: (...) -> Iterable[bytes]
total_length = _get_http_response_size(resp)
if link.netloc == PyPI.file_storage_domain:
url = link.show_url
else:
url = link.url_without_fragment
logged_url = redact_auth_from_url(url)
if total_length:
logged_url = '{} ({})'.format(logged_url, format_size(total_length))
if is_from_cache(resp):
logger.info("Using cached %s", logged_url)
else:
logger.info("Downloading %s", logged_url)
if logger.getEffectiveLevel() > logging.INFO:
show_progress = False
elif is_from_cache(resp):
show_progress = False
elif not total_length:
show_progress = True
elif total_length > (40 * 1000):
show_progress = True
else:
show_progress = False
chunks = response_chunks(resp, CONTENT_CHUNK_SIZE)
if not show_progress:
return chunks
return DownloadProgressProvider(
progress_bar, max=total_length
)(chunks)
def sanitize_content_filename(filename):
# type: (str) -> str
"""
Sanitize the "filename" value from a Content-Disposition header.
"""
return os.path.basename(filename)
def parse_content_disposition(content_disposition, default_filename):
# type: (str, str) -> str
"""
Parse the "filename" value from a Content-Disposition header, and
return the default filename if the result is empty.
"""
_type, params = cgi.parse_header(content_disposition)
filename = params.get('filename')
if filename:
# We need to sanitize the filename to prevent directory traversal
# in case the filename contains ".." path parts.
filename = sanitize_content_filename(filename)
return filename or default_filename
def _get_http_response_filename(resp, link):
# type: (Response, Link) -> str
"""Get an ideal filename from the given HTTP response, falling back to
the link filename if not provided.
"""
filename = link.filename # fallback
# Have a look at the Content-Disposition header for a better guess
content_disposition = resp.headers.get('content-disposition')
if content_disposition:
filename = parse_content_disposition(content_disposition, filename)
ext = splitext(filename)[1] # type: Optional[str]
if not ext:
ext = mimetypes.guess_extension(
resp.headers.get('content-type', '')
)
if ext:
filename += ext
if not ext and link.url != resp.url:
ext = os.path.splitext(resp.url)[1]
if ext:
filename += ext
return filename
def _http_get_download(session, link):
# type: (PipSession, Link) -> Response
target_url = link.url.split('#', 1)[0]
resp = session.get(
target_url,
# We use Accept-Encoding: identity here because requests
# defaults to accepting compressed responses. This breaks in
# a variety of ways depending on how the server is configured.
# - Some servers will notice that the file isn't a compressible
# file and will leave the file alone and with an empty
# Content-Encoding
# - Some servers will notice that the file is already
# compressed and will leave the file alone and will add a
# Content-Encoding: gzip header
# - Some servers won't notice anything at all and will take
# a file that's already been compressed and compress it again
# and set the Content-Encoding: gzip header
# By setting this to request only the identity encoding We're
# hoping to eliminate the third case. Hopefully there does not
# exist a server which when given a file will notice it is
# already compressed and that you're not asking for a
# compressed file and will then decompress it before sending
# because if that's the case I don't think it'll ever be
# possible to make this work.
headers={"Accept-Encoding": "identity"},
stream=True,
)
resp.raise_for_status()
return resp
class Download(object):
def __init__(
self,
response, # type: Response
filename, # type: str
chunks, # type: Iterable[bytes]
):
# type: (...) -> None
self.response = response
self.filename = filename
self.chunks = chunks
class Downloader(object):
def __init__(
self,
session, # type: PipSession
progress_bar, # type: str
):
# type: (...) -> None
self._session = session
self._progress_bar = progress_bar
def __call__(self, link):
# type: (Link) -> Download
try:
resp = _http_get_download(self._session, link)
except requests.HTTPError as e:
logger.critical(
"HTTP error %s while getting %s", e.response.status_code, link
)
raise
return Download(
resp,
_get_http_response_filename(resp, link),
_prepare_download(resp, link, self._progress_bar),
)

View File

@@ -0,0 +1,405 @@
"""PipSession and supporting code, containing all pip-specific
network request configuration and behavior.
"""
# The following comment should be removed at some point in the future.
# mypy: disallow-untyped-defs=False
import email.utils
import json
import logging
import mimetypes
import os
import platform
import sys
import warnings
from pip._vendor import requests, six, urllib3
from pip._vendor.cachecontrol import CacheControlAdapter
from pip._vendor.requests.adapters import BaseAdapter, HTTPAdapter
from pip._vendor.requests.models import Response
from pip._vendor.requests.structures import CaseInsensitiveDict
from pip._vendor.six.moves.urllib import parse as urllib_parse
from pip._vendor.urllib3.exceptions import InsecureRequestWarning
from pip import __version__
from pip._internal.network.auth import MultiDomainBasicAuth
from pip._internal.network.cache import SafeFileCache
# Import ssl from compat so the initial import occurs in only one place.
from pip._internal.utils.compat import has_tls, ipaddress
from pip._internal.utils.glibc import libc_ver
from pip._internal.utils.misc import (
build_url_from_netloc,
get_installed_version,
parse_netloc,
)
from pip._internal.utils.typing import MYPY_CHECK_RUNNING
from pip._internal.utils.urls import url_to_path
if MYPY_CHECK_RUNNING:
from typing import (
Iterator, List, Optional, Tuple, Union,
)
from pip._internal.models.link import Link
SecureOrigin = Tuple[str, str, Optional[Union[int, str]]]
logger = logging.getLogger(__name__)
# Ignore warning raised when using --trusted-host.
warnings.filterwarnings("ignore", category=InsecureRequestWarning)
SECURE_ORIGINS = [
# protocol, hostname, port
# Taken from Chrome's list of secure origins (See: http://bit.ly/1qrySKC)
("https", "*", "*"),
("*", "localhost", "*"),
("*", "127.0.0.0/8", "*"),
("*", "::1/128", "*"),
("file", "*", None),
# ssh is always secure.
("ssh", "*", "*"),
] # type: List[SecureOrigin]
# These are environment variables present when running under various
# CI systems. For each variable, some CI systems that use the variable
# are indicated. The collection was chosen so that for each of a number
# of popular systems, at least one of the environment variables is used.
# This list is used to provide some indication of and lower bound for
# CI traffic to PyPI. Thus, it is okay if the list is not comprehensive.
# For more background, see: https://github.com/pypa/pip/issues/5499
CI_ENVIRONMENT_VARIABLES = (
# Azure Pipelines
'BUILD_BUILDID',
# Jenkins
'BUILD_ID',
# AppVeyor, CircleCI, Codeship, Gitlab CI, Shippable, Travis CI
'CI',
# Explicit environment variable.
'PIP_IS_CI',
)
def looks_like_ci():
# type: () -> bool
"""
Return whether it looks like pip is running under CI.
"""
# We don't use the method of checking for a tty (e.g. using isatty())
# because some CI systems mimic a tty (e.g. Travis CI). Thus that
# method doesn't provide definitive information in either direction.
return any(name in os.environ for name in CI_ENVIRONMENT_VARIABLES)
def user_agent():
"""
Return a string representing the user agent.
"""
data = {
"installer": {"name": "pip", "version": __version__},
"python": platform.python_version(),
"implementation": {
"name": platform.python_implementation(),
},
}
if data["implementation"]["name"] == 'CPython':
data["implementation"]["version"] = platform.python_version()
elif data["implementation"]["name"] == 'PyPy':
if sys.pypy_version_info.releaselevel == 'final':
pypy_version_info = sys.pypy_version_info[:3]
else:
pypy_version_info = sys.pypy_version_info
data["implementation"]["version"] = ".".join(
[str(x) for x in pypy_version_info]
)
elif data["implementation"]["name"] == 'Jython':
# Complete Guess
data["implementation"]["version"] = platform.python_version()
elif data["implementation"]["name"] == 'IronPython':
# Complete Guess
data["implementation"]["version"] = platform.python_version()
if sys.platform.startswith("linux"):
from pip._vendor import distro
distro_infos = dict(filter(
lambda x: x[1],
zip(["name", "version", "id"], distro.linux_distribution()),
))
libc = dict(filter(
lambda x: x[1],
zip(["lib", "version"], libc_ver()),
))
if libc:
distro_infos["libc"] = libc
if distro_infos:
data["distro"] = distro_infos
if sys.platform.startswith("darwin") and platform.mac_ver()[0]:
data["distro"] = {"name": "macOS", "version": platform.mac_ver()[0]}
if platform.system():
data.setdefault("system", {})["name"] = platform.system()
if platform.release():
data.setdefault("system", {})["release"] = platform.release()
if platform.machine():
data["cpu"] = platform.machine()
if has_tls():
import _ssl as ssl
data["openssl_version"] = ssl.OPENSSL_VERSION
setuptools_version = get_installed_version("setuptools")
if setuptools_version is not None:
data["setuptools_version"] = setuptools_version
# Use None rather than False so as not to give the impression that
# pip knows it is not being run under CI. Rather, it is a null or
# inconclusive result. Also, we include some value rather than no
# value to make it easier to know that the check has been run.
data["ci"] = True if looks_like_ci() else None
user_data = os.environ.get("PIP_USER_AGENT_USER_DATA")
if user_data is not None:
data["user_data"] = user_data
return "{data[installer][name]}/{data[installer][version]} {json}".format(
data=data,
json=json.dumps(data, separators=(",", ":"), sort_keys=True),
)
class LocalFSAdapter(BaseAdapter):
def send(self, request, stream=None, timeout=None, verify=None, cert=None,
proxies=None):
pathname = url_to_path(request.url)
resp = Response()
resp.status_code = 200
resp.url = request.url
try:
stats = os.stat(pathname)
except OSError as exc:
resp.status_code = 404
resp.raw = exc
else:
modified = email.utils.formatdate(stats.st_mtime, usegmt=True)
content_type = mimetypes.guess_type(pathname)[0] or "text/plain"
resp.headers = CaseInsensitiveDict({
"Content-Type": content_type,
"Content-Length": stats.st_size,
"Last-Modified": modified,
})
resp.raw = open(pathname, "rb")
resp.close = resp.raw.close
return resp
def close(self):
pass
class InsecureHTTPAdapter(HTTPAdapter):
def cert_verify(self, conn, url, verify, cert):
super(InsecureHTTPAdapter, self).cert_verify(
conn=conn, url=url, verify=False, cert=cert
)
class PipSession(requests.Session):
timeout = None # type: Optional[int]
def __init__(self, *args, **kwargs):
"""
:param trusted_hosts: Domains not to emit warnings for when not using
HTTPS.
"""
retries = kwargs.pop("retries", 0)
cache = kwargs.pop("cache", None)
trusted_hosts = kwargs.pop("trusted_hosts", []) # type: List[str]
index_urls = kwargs.pop("index_urls", None)
super(PipSession, self).__init__(*args, **kwargs)
# Namespace the attribute with "pip_" just in case to prevent
# possible conflicts with the base class.
self.pip_trusted_origins = [] # type: List[Tuple[str, Optional[int]]]
# Attach our User Agent to the request
self.headers["User-Agent"] = user_agent()
# Attach our Authentication handler to the session
self.auth = MultiDomainBasicAuth(index_urls=index_urls)
# Create our urllib3.Retry instance which will allow us to customize
# how we handle retries.
retries = urllib3.Retry(
# Set the total number of retries that a particular request can
# have.
total=retries,
# A 503 error from PyPI typically means that the Fastly -> Origin
# connection got interrupted in some way. A 503 error in general
# is typically considered a transient error so we'll go ahead and
# retry it.
# A 500 may indicate transient error in Amazon S3
# A 520 or 527 - may indicate transient error in CloudFlare
status_forcelist=[500, 503, 520, 527],
# Add a small amount of back off between failed requests in
# order to prevent hammering the service.
backoff_factor=0.25,
)
# We want to _only_ cache responses on securely fetched origins. We do
# this because we can't validate the response of an insecurely fetched
# origin, and we don't want someone to be able to poison the cache and
# require manual eviction from the cache to fix it.
if cache:
secure_adapter = CacheControlAdapter(
cache=SafeFileCache(cache),
max_retries=retries,
)
else:
secure_adapter = HTTPAdapter(max_retries=retries)
# Our Insecure HTTPAdapter disables HTTPS validation. It does not
# support caching (see above) so we'll use it for all http:// URLs as
# well as any https:// host that we've marked as ignoring TLS errors
# for.
insecure_adapter = InsecureHTTPAdapter(max_retries=retries)
# Save this for later use in add_insecure_host().
self._insecure_adapter = insecure_adapter
self.mount("https://", secure_adapter)
self.mount("http://", insecure_adapter)
# Enable file:// urls
self.mount("file://", LocalFSAdapter())
for host in trusted_hosts:
self.add_trusted_host(host, suppress_logging=True)
def add_trusted_host(self, host, source=None, suppress_logging=False):
# type: (str, Optional[str], bool) -> None
"""
:param host: It is okay to provide a host that has previously been
added.
:param source: An optional source string, for logging where the host
string came from.
"""
if not suppress_logging:
msg = 'adding trusted host: {!r}'.format(host)
if source is not None:
msg += ' (from {})'.format(source)
logger.info(msg)
host_port = parse_netloc(host)
if host_port not in self.pip_trusted_origins:
self.pip_trusted_origins.append(host_port)
self.mount(build_url_from_netloc(host) + '/', self._insecure_adapter)
if not host_port[1]:
# Mount wildcard ports for the same host.
self.mount(
build_url_from_netloc(host) + ':',
self._insecure_adapter
)
def iter_secure_origins(self):
# type: () -> Iterator[SecureOrigin]
for secure_origin in SECURE_ORIGINS:
yield secure_origin
for host, port in self.pip_trusted_origins:
yield ('*', host, '*' if port is None else port)
def is_secure_origin(self, location):
# type: (Link) -> bool
# Determine if this url used a secure transport mechanism
parsed = urllib_parse.urlparse(str(location))
origin_protocol, origin_host, origin_port = (
parsed.scheme, parsed.hostname, parsed.port,
)
# The protocol to use to see if the protocol matches.
# Don't count the repository type as part of the protocol: in
# cases such as "git+ssh", only use "ssh". (I.e., Only verify against
# the last scheme.)
origin_protocol = origin_protocol.rsplit('+', 1)[-1]
# Determine if our origin is a secure origin by looking through our
# hardcoded list of secure origins, as well as any additional ones
# configured on this PackageFinder instance.
for secure_origin in self.iter_secure_origins():
secure_protocol, secure_host, secure_port = secure_origin
if origin_protocol != secure_protocol and secure_protocol != "*":
continue
try:
addr = ipaddress.ip_address(
None
if origin_host is None
else six.ensure_text(origin_host)
)
network = ipaddress.ip_network(
six.ensure_text(secure_host)
)
except ValueError:
# We don't have both a valid address or a valid network, so
# we'll check this origin against hostnames.
if (
origin_host and
origin_host.lower() != secure_host.lower() and
secure_host != "*"
):
continue
else:
# We have a valid address and network, so see if the address
# is contained within the network.
if addr not in network:
continue
# Check to see if the port matches.
if (
origin_port != secure_port and
secure_port != "*" and
secure_port is not None
):
continue
# If we've gotten here, then this origin matches the current
# secure origin and we should return True
return True
# If we've gotten to this point, then the origin isn't secure and we
# will not accept it as a valid location to search. We will however
# log a warning that we are ignoring it.
logger.warning(
"The repository located at %s is not a trusted or secure host and "
"is being ignored. If this repository is available via HTTPS we "
"recommend you use HTTPS instead, otherwise you may silence "
"this warning and allow it anyway with '--trusted-host %s'.",
origin_host,
origin_host,
)
return False
def request(self, method, url, *args, **kwargs):
# Allow setting a default timeout on a session
kwargs.setdefault("timeout", self.timeout)
# Dispatch the actual request
return super(PipSession, self).request(method, url, *args, **kwargs)

View File

@@ -0,0 +1,48 @@
from pip._vendor.requests.models import CONTENT_CHUNK_SIZE, Response
from pip._internal.utils.typing import MYPY_CHECK_RUNNING
if MYPY_CHECK_RUNNING:
from typing import Iterator
def response_chunks(response, chunk_size=CONTENT_CHUNK_SIZE):
# type: (Response, int) -> Iterator[bytes]
"""Given a requests Response, provide the data chunks.
"""
try:
# Special case for urllib3.
for chunk in response.raw.stream(
chunk_size,
# We use decode_content=False here because we don't
# want urllib3 to mess with the raw bytes we get
# from the server. If we decompress inside of
# urllib3 then we cannot verify the checksum
# because the checksum will be of the compressed
# file. This breakage will only occur if the
# server adds a Content-Encoding header, which
# depends on how the server was configured:
# - Some servers will notice that the file isn't a
# compressible file and will leave the file alone
# and with an empty Content-Encoding
# - Some servers will notice that the file is
# already compressed and will leave the file
# alone and will add a Content-Encoding: gzip
# header
# - Some servers won't notice anything at all and
# will take a file that's already been compressed
# and compress it again and set the
# Content-Encoding: gzip header
#
# By setting this not to decode automatically we
# hope to eliminate problems with the second case.
decode_content=False,
):
yield chunk
except AttributeError:
# Standard file-like object.
while True:
chunk = response.raw.read(chunk_size)
if not chunk:
break
yield chunk

View File

@@ -0,0 +1,44 @@
"""xmlrpclib.Transport implementation
"""
# The following comment should be removed at some point in the future.
# mypy: disallow-untyped-defs=False
import logging
from pip._vendor import requests
# NOTE: XMLRPC Client is not annotated in typeshed as on 2017-07-17, which is
# why we ignore the type on this import
from pip._vendor.six.moves import xmlrpc_client # type: ignore
from pip._vendor.six.moves.urllib import parse as urllib_parse
logger = logging.getLogger(__name__)
class PipXmlrpcTransport(xmlrpc_client.Transport):
"""Provide a `xmlrpclib.Transport` implementation via a `PipSession`
object.
"""
def __init__(self, index_url, session, use_datetime=False):
xmlrpc_client.Transport.__init__(self, use_datetime)
index_parts = urllib_parse.urlparse(index_url)
self._scheme = index_parts.scheme
self._session = session
def request(self, host, handler, request_body, verbose=False):
parts = (self._scheme, host, handler, None, None, None)
url = urllib_parse.urlunparse(parts)
try:
headers = {'Content-Type': 'text/xml'}
response = self._session.post(url, data=request_body,
headers=headers, stream=True)
response.raise_for_status()
self.verbose = verbose
return self.parse_response(response.raw)
except requests.HTTPError as exc:
logger.critical(
"HTTP error %s while getting %s",
exc.response.status_code, url,
)
raise