CLERC/Reskreen
6
0
Fork 0
Code Issues 5 Pull Requests Actions Packages Projects 1 Releases 1 Wiki Activity

Change venv

Browse Source
This commit is contained in:
Ambulance Clerc
2023-05-31 08:31:22 +02:00
parent fb6f579089
commit fdbb52c96f
466 changed files with 25899 additions and 64721 deletions
Download Patch File Download Diff File Expand all files Collapse all files

409
venv/Lib/site-packages/pip/_internal/index/collector.py
View File

@@ -2,30 +2,32 @@
The main purpose of this module is to expose LinkCollector.collect_sources().
"""
import cgi
import collections
import email.message
import functools
import html
import itertools
import json
import logging
import os
import re
import urllib.parse
import urllib.request
import xml.etree.ElementTree
from html.parser import HTMLParser
from optparse import Values
from typing import (
TYPE_CHECKING,
Callable,
Dict,
Iterable,
List,
MutableMapping,
NamedTuple,
Optional,
Sequence,
Tuple,
Union,
)
from pip._vendor import html5lib, requests
from pip._vendor import requests
from pip._vendor.requests import Response
from pip._vendor.requests.exceptions import RetryError, SSLError
@@ -35,14 +37,18 @@ from pip._internal.models.search_scope import SearchScope
from pip._internal.network.session import PipSession
from pip._internal.network.utils import raise_for_status
from pip._internal.utils.filetypes import is_archive_file
from pip._internal.utils.misc import pairwise, redact_auth_from_url
from pip._internal.utils.misc import redact_auth_from_url
from pip._internal.vcs import vcs
from .sources import CandidatesFromPage, LinkSource, build_source
if TYPE_CHECKING:
from typing import Protocol
else:
Protocol = object
logger = logging.getLogger(__name__)
HTMLElement = xml.etree.ElementTree.Element
ResponseHeaders = MutableMapping[str, str]
@@ -52,70 +58,90 @@ def _match_vcs_scheme(url: str) -> Optional[str]:
Returns the matched VCS scheme, or None if there's no match.
"""
for scheme in vcs.schemes:
if url.lower().startswith(scheme) and url[len(scheme)] in '+:':
if url.lower().startswith(scheme) and url[len(scheme)] in "+:":
return scheme
return None
class _NotHTML(Exception):
class _NotAPIContent(Exception):
def __init__(self, content_type: str, request_desc: str) -> None:
super().__init__(content_type, request_desc)
self.content_type = content_type
self.request_desc = request_desc
def _ensure_html_header(response: Response) -> None:
"""Check the Content-Type header to ensure the response contains HTML.
Raises `_NotHTML` if the content type is not text/html.
def _ensure_api_header(response: Response) -> None:
"""
content_type = response.headers.get("Content-Type", "")
if not content_type.lower().startswith("text/html"):
raise _NotHTML(content_type, response.request.method)
Check the Content-Type header to ensure the response contains a Simple
API Response.
Raises `_NotAPIContent` if the content type is not a valid content-type.
"""
content_type = response.headers.get("Content-Type", "Unknown")
content_type_l = content_type.lower()
if content_type_l.startswith(
(
"text/html",
"application/vnd.pypi.simple.v1+html",
"application/vnd.pypi.simple.v1+json",
)
):
return
raise _NotAPIContent(content_type, response.request.method)
class _NotHTTP(Exception):
pass
def _ensure_html_response(url: str, session: PipSession) -> None:
"""Send a HEAD request to the URL, and ensure the response contains HTML.
def _ensure_api_response(url: str, session: PipSession) -> None:
"""
Send a HEAD request to the URL, and ensure the response contains a simple
API Response.
Raises `_NotHTTP` if the URL is not available for a HEAD request, or
`_NotHTML` if the content type is not text/html.
`_NotAPIContent` if the content type is not a valid content type.
"""
scheme, netloc, path, query, fragment = urllib.parse.urlsplit(url)
if scheme not in {'http', 'https'}:
if scheme not in {"http", "https"}:
raise _NotHTTP()
resp = session.head(url, allow_redirects=True)
raise_for_status(resp)
_ensure_html_header(resp)
_ensure_api_header(resp)
def _get_html_response(url: str, session: PipSession) -> Response:
"""Access an HTML page with GET, and return the response.
def _get_simple_response(url: str, session: PipSession) -> Response:
"""Access an Simple API response with GET, and return the response.
This consists of three parts:
1. If the URL looks suspiciously like an archive, send a HEAD first to
check the Content-Type is HTML, to avoid downloading a large file.
Raise `_NotHTTP` if the content type cannot be determined, or
`_NotHTML` if it is not HTML.
check the Content-Type is HTML or Simple API, to avoid downloading a
large file. Raise `_NotHTTP` if the content type cannot be determined, or
`_NotAPIContent` if it is not HTML or a Simple API.
2. Actually perform the request. Raise HTTP exceptions on network failures.
3. Check the Content-Type header to make sure we got HTML, and raise
`_NotHTML` otherwise.
3. Check the Content-Type header to make sure we got a Simple API response,
and raise `_NotAPIContent` otherwise.
"""
if is_archive_file(Link(url).filename):
_ensure_html_response(url, session=session)
_ensure_api_response(url, session=session)
logger.debug('Getting page %s', redact_auth_from_url(url))
logger.debug("Getting page %s", redact_auth_from_url(url))
resp = session.get(
url,
headers={
"Accept": "text/html",
"Accept": ", ".join(
[
"application/vnd.pypi.simple.v1+json",
"application/vnd.pypi.simple.v1+html; q=0.1",
"text/html; q=0.01",
]
),
# We don't want to blindly returned cached data for
# /simple/, because authors generally expecting that
# twine upload && pip install will function, but if
@@ -137,153 +163,52 @@ def _get_html_response(url: str, session: PipSession) -> Response:
# 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.
_ensure_html_header(resp)
# url we cannot know ahead of time for sure if something is a
# Simple API response or not. However we can check after we've
# downloaded it.
_ensure_api_header(resp)
logger.debug(
"Fetched page %s as %s",
redact_auth_from_url(url),
resp.headers.get("Content-Type", "Unknown"),
)
return resp
def _get_encoding_from_headers(headers: ResponseHeaders) -> Optional[str]:
"""Determine if we have any encoding information in our headers.
"""
"""Determine if we have any encoding information in our headers."""
if headers and "Content-Type" in headers:
content_type, params = cgi.parse_header(headers["Content-Type"])
if "charset" in params:
return params['charset']
m = email.message.Message()
m["content-type"] = headers["Content-Type"]
charset = m.get_param("charset")
if charset:
return str(charset)
return None
def _determine_base_url(document: HTMLElement, page_url: str) -> str:
"""Determine the HTML document's base URL.
This looks for a ``<base>`` tag in the HTML document. If present, its href
attribute denotes the base URL of anchor tags in the document. If there is
no such tag (or if it does not have a valid href attribute), the HTML
file's URL is used as the base URL.
:param document: An HTML document representation. The current
implementation expects the result of ``html5lib.parse()``.
:param page_url: The URL of the HTML document.
"""
for base in document.findall(".//base"):
href = base.get("href")
if href is not None:
return href
return page_url
def _clean_url_path_part(part: str) -> str:
"""
Clean a "part" of a URL path (i.e. after splitting on "@" characters).
"""
# We unquote prior to quoting to make sure nothing is double quoted.
return urllib.parse.quote(urllib.parse.unquote(part))
def _clean_file_url_path(part: str) -> str:
"""
Clean the first part of a URL path that corresponds to a local
filesystem path (i.e. the first part after splitting on "@" characters).
"""
# We unquote prior to quoting to make sure nothing is double quoted.
# Also, on Windows the path part might contain a drive letter which
# should not be quoted. On Linux where drive letters do not
# exist, the colon should be quoted. We rely on urllib.request
# to do the right thing here.
return urllib.request.pathname2url(urllib.request.url2pathname(part))
# percent-encoded: /
_reserved_chars_re = re.compile('(@|%2F)', re.IGNORECASE)
def _clean_url_path(path: str, is_local_path: bool) -> str:
"""
Clean the path portion of a URL.
"""
if is_local_path:
clean_func = _clean_file_url_path
else:
clean_func = _clean_url_path_part
# Split on the reserved characters prior to cleaning so that
# revision strings in VCS URLs are properly preserved.
parts = _reserved_chars_re.split(path)
cleaned_parts = []
for to_clean, reserved in pairwise(itertools.chain(parts, [''])):
cleaned_parts.append(clean_func(to_clean))
# Normalize %xx escapes (e.g. %2f -> %2F)
cleaned_parts.append(reserved.upper())
return ''.join(cleaned_parts)
def _clean_link(url: str) -> str:
"""
Make sure a link is fully quoted.
For example, if ' ' occurs in the URL, it will be replaced with "%20",
and without double-quoting other characters.
"""
# Split the URL into parts according to the general structure
# `scheme://netloc/path;parameters?query#fragment`.
result = urllib.parse.urlparse(url)
# If the netloc is empty, then the URL refers to a local filesystem path.
is_local_path = not result.netloc
path = _clean_url_path(result.path, is_local_path=is_local_path)
return urllib.parse.urlunparse(result._replace(path=path))
def _create_link_from_element(
anchor: HTMLElement,
page_url: str,
base_url: str,
) -> Optional[Link]:
"""
Convert an anchor element in a simple repository page to a Link.
"""
href = anchor.get("href")
if not href:
return None
url = _clean_link(urllib.parse.urljoin(base_url, href))
pyrequire = anchor.get('data-requires-python')
pyrequire = html.unescape(pyrequire) if pyrequire else None
yanked_reason = anchor.get('data-yanked')
if yanked_reason:
yanked_reason = html.unescape(yanked_reason)
link = Link(
url,
comes_from=page_url,
requires_python=pyrequire,
yanked_reason=yanked_reason,
)
return link
class CacheablePageContent:
def __init__(self, page: "HTMLPage") -> None:
def __init__(self, page: "IndexContent") -> None:
assert page.cache_link_parsing
self.page = page
def __eq__(self, other: object) -> bool:
return (isinstance(other, type(self)) and
self.page.url == other.page.url)
return isinstance(other, type(self)) and self.page.url == other.page.url
def __hash__(self) -> int:
return hash(self.page.url)
def with_cached_html_pages(
fn: Callable[["HTMLPage"], Iterable[Link]],
) -> Callable[["HTMLPage"], List[Link]]:
class ParseLinks(Protocol):
def __call__(self, page: "IndexContent") -> Iterable[Link]:
...
def with_cached_index_content(fn: ParseLinks) -> ParseLinks:
"""
Given a function that parses an Iterable[Link] from an HTMLPage, cache the
function's result (keyed by CacheablePageContent), unless the HTMLPage
Given a function that parses an Iterable[Link] from an IndexContent, cache the
function's result (keyed by CacheablePageContent), unless the IndexContent
`page` has `page.cache_link_parsing == False`.
"""
@@ -292,7 +217,7 @@ def with_cached_html_pages(
return list(fn(cacheable_page.page))
@functools.wraps(fn)
def wrapper_wrapper(page: "HTMLPage") -> List[Link]:
def wrapper_wrapper(page: "IndexContent") -> List[Link]:
if page.cache_link_parsing:
return wrapper(CacheablePageContent(page))
return list(fn(page))
@@ -300,36 +225,42 @@ def with_cached_html_pages(
return wrapper_wrapper
@with_cached_html_pages
def parse_links(page: "HTMLPage") -> Iterable[Link]:
@with_cached_index_content
def parse_links(page: "IndexContent") -> Iterable[Link]:
"""
Parse an HTML document, and yield its anchor elements as Link objects.
Parse a Simple API's Index Content, and yield its anchor elements as Link objects.
"""
document = html5lib.parse(
page.content,
transport_encoding=page.encoding,
namespaceHTMLElements=False,
)
content_type_l = page.content_type.lower()
if content_type_l.startswith("application/vnd.pypi.simple.v1+json"):
data = json.loads(page.content)
for file in data.get("files", []):
link = Link.from_json(file, page.url)
if link is None:
continue
yield link
return
parser = HTMLLinkParser(page.url)
encoding = page.encoding or "utf-8"
parser.feed(page.content.decode(encoding))
url = page.url
base_url = _determine_base_url(document, url)
for anchor in document.findall(".//a"):
link = _create_link_from_element(
anchor,
page_url=url,
base_url=base_url,
)
base_url = parser.base_url or url
for anchor in parser.anchors:
link = Link.from_element(anchor, page_url=url, base_url=base_url)
if link is None:
continue
yield link
class HTMLPage:
"""Represents one page, along with its URL"""
class IndexContent:
"""Represents one response (or page), along with its URL"""
def __init__(
self,
content: bytes,
content_type: str,
encoding: Optional[str],
url: str,
cache_link_parsing: bool = True,
@@ -342,6 +273,7 @@ class HTMLPage:
have this set to False, for example.
"""
self.content = content
self.content_type = content_type
self.encoding = encoding
self.url = url
self.cache_link_parsing = cache_link_parsing
@@ -350,80 +282,115 @@ class HTMLPage:
return redact_auth_from_url(self.url)
def _handle_get_page_fail(
class HTMLLinkParser(HTMLParser):
"""
HTMLParser that keeps the first base HREF and a list of all anchor
elements' attributes.
"""
def __init__(self, url: str) -> None:
super().__init__(convert_charrefs=True)
self.url: str = url
self.base_url: Optional[str] = None
self.anchors: List[Dict[str, Optional[str]]] = []
def handle_starttag(self, tag: str, attrs: List[Tuple[str, Optional[str]]]) -> None:
if tag == "base" and self.base_url is None:
href = self.get_href(attrs)
if href is not None:
self.base_url = href
elif tag == "a":
self.anchors.append(dict(attrs))
def get_href(self, attrs: List[Tuple[str, Optional[str]]]) -> Optional[str]:
for name, value in attrs:
if name == "href":
return value
return None
def _handle_get_simple_fail(
link: Link,
reason: Union[str, Exception],
meth: Optional[Callable[..., None]] = None
meth: Optional[Callable[..., None]] = None,
) -> None:
if meth is None:
meth = logger.debug
meth("Could not fetch URL %s: %s - skipping", link, reason)
def _make_html_page(response: Response, cache_link_parsing: bool = True) -> HTMLPage:
def _make_index_content(
response: Response, cache_link_parsing: bool = True
) -> IndexContent:
encoding = _get_encoding_from_headers(response.headers)
return HTMLPage(
return IndexContent(
response.content,
response.headers["Content-Type"],
encoding=encoding,
url=response.url,
cache_link_parsing=cache_link_parsing)
cache_link_parsing=cache_link_parsing,
)
def _get_html_page(
link: Link, session: Optional[PipSession] = None
) -> Optional["HTMLPage"]:
if session is None:
raise TypeError(
"_get_html_page() missing 1 required keyword argument: 'session'"
)
url = link.url.split('#', 1)[0]
def _get_index_content(link: Link, *, session: PipSession) -> Optional["IndexContent"]:
url = link.url.split("#", 1)[0]
# Check for VCS schemes that do not support lookup as web pages.
vcs_scheme = _match_vcs_scheme(url)
if vcs_scheme:
logger.warning('Cannot look at %s URL %s because it does not support '
'lookup as web pages.', vcs_scheme, link)
logger.warning(
"Cannot look at %s URL %s because it does not support lookup as web pages.",
vcs_scheme,
link,
)
return None
# Tack index.html onto file:// URLs that point to directories
scheme, _, path, _, _, _ = urllib.parse.urlparse(url)
if (scheme == 'file' and os.path.isdir(urllib.request.url2pathname(path))):
if scheme == "file" and os.path.isdir(urllib.request.url2pathname(path)):
# add trailing slash if not present so urljoin doesn't trim
# final segment
if not url.endswith('/'):
url += '/'
url = urllib.parse.urljoin(url, 'index.html')
logger.debug(' file: URL is directory, getting %s', url)
if not url.endswith("/"):
url += "/"
# TODO: In the future, it would be nice if pip supported PEP 691
# style responses in the file:// URLs, however there's no
# standard file extension for application/vnd.pypi.simple.v1+json
# so we'll need to come up with something on our own.
url = urllib.parse.urljoin(url, "index.html")
logger.debug(" file: URL is directory, getting %s", url)
try:
resp = _get_html_response(url, session=session)
resp = _get_simple_response(url, session=session)
except _NotHTTP:
logger.warning(
'Skipping page %s because it looks like an archive, and cannot '
'be checked by a HTTP HEAD request.', link,
"Skipping page %s because it looks like an archive, and cannot "
"be checked by a HTTP HEAD request.",
link,
)
except _NotHTML as exc:
except _NotAPIContent as exc:
logger.warning(
'Skipping page %s because the %s request got Content-Type: %s.'
'The only supported Content-Type is text/html',
link, exc.request_desc, exc.content_type,
"Skipping page %s because the %s request got Content-Type: %s. "
"The only supported Content-Types are application/vnd.pypi.simple.v1+json, "
"application/vnd.pypi.simple.v1+html, and text/html",
link,
exc.request_desc,
exc.content_type,
)
except NetworkConnectionError as exc:
_handle_get_page_fail(link, exc)
_handle_get_simple_fail(link, exc)
except RetryError as exc:
_handle_get_page_fail(link, exc)
_handle_get_simple_fail(link, exc)
except SSLError as exc:
reason = "There was a problem confirming the ssl certificate: "
reason += str(exc)
_handle_get_page_fail(link, reason, meth=logger.info)
_handle_get_simple_fail(link, reason, meth=logger.info)
except requests.ConnectionError as exc:
_handle_get_page_fail(link, f"connection error: {exc}")
_handle_get_simple_fail(link, f"connection error: {exc}")
except requests.Timeout:
_handle_get_page_fail(link, "timed out")
_handle_get_simple_fail(link, "timed out")
else:
return _make_html_page(resp,
cache_link_parsing=link.cache_link_parsing)
return _make_index_content(resp, cache_link_parsing=link.cache_link_parsing)
return None
@@ -451,9 +418,10 @@ class LinkCollector:
@classmethod
def create(
cls, session: PipSession,
cls,
session: PipSession,
options: Values,
suppress_no_index: bool = False
suppress_no_index: bool = False,
) -> "LinkCollector":
"""
:param session: The Session to use to make requests.
@@ -463,8 +431,8 @@ class LinkCollector:
index_urls = [options.index_url] + options.extra_index_urls
if options.no_index and not suppress_no_index:
logger.debug(
'Ignoring indexes: %s',
','.join(redact_auth_from_url(url) for url in index_urls),
"Ignoring indexes: %s",
",".join(redact_auth_from_url(url) for url in index_urls),
)
index_urls = []
@@ -472,10 +440,13 @@ class LinkCollector:
find_links = options.find_links or []
search_scope = SearchScope.create(
find_links=find_links, index_urls=index_urls,
find_links=find_links,
index_urls=index_urls,
no_index=options.no_index,
)
link_collector = LinkCollector(
session=session, search_scope=search_scope,
session=session,
search_scope=search_scope,
)
return link_collector
@@ -483,11 +454,11 @@ class LinkCollector:
def find_links(self) -> List[str]:
return self.search_scope.find_links
def fetch_page(self, location: Link) -> Optional[HTMLPage]:
def fetch_response(self, location: Link) -> Optional[IndexContent]:
"""
Fetch an HTML page containing package links.
"""
return _get_html_page(location, session=self.session)
return _get_index_content(location, session=self.session)
def collect_sources(
self,