cli-ih 0.6.2.1__py3-none-any.whl → 0.6.3.1__py3-none-any.whl

venv/Lib/site-packages/pip/_internal/network/download.py

@@ -0,0 +1,342 @@
+"""Download files with progress indicators."""
+
+from __future__ import annotations
+
+import email.message
+import logging
+import mimetypes
+import os
+from collections.abc import Iterable, Mapping
+from dataclasses import dataclass
+from http import HTTPStatus
+from typing import BinaryIO
+
+from pip._vendor.requests import PreparedRequest
+from pip._vendor.requests.models import Response
+from pip._vendor.urllib3 import HTTPResponse as URLlib3Response
+from pip._vendor.urllib3._collections import HTTPHeaderDict
+from pip._vendor.urllib3.exceptions import ReadTimeoutError
+
+from pip._internal.cli.progress_bars import BarType, get_download_progress_renderer
+from pip._internal.exceptions import IncompleteDownloadError, NetworkConnectionError
+from pip._internal.models.index import PyPI
+from pip._internal.models.link import Link
+from pip._internal.network.cache import SafeFileCache, is_from_cache
+from pip._internal.network.session import CacheControlAdapter, PipSession
+from pip._internal.network.utils import HEADERS, raise_for_status, response_chunks
+from pip._internal.utils.misc import format_size, redact_auth_from_url, splitext
+
+logger = logging.getLogger(__name__)
+
+
+def _get_http_response_size(resp: Response) -> int | None:
+    try:
+        return int(resp.headers["content-length"])
+    except (ValueError, KeyError, TypeError):
+        return None
+
+
+def _get_http_response_etag_or_last_modified(resp: Response) -> str | None:
+    """
+    Return either the ETag or Last-Modified header (or None if neither exists).
+    The return value can be used in an If-Range header.
+    """
+    return resp.headers.get("etag", resp.headers.get("last-modified"))
+
+
+def _log_download(
+    resp: Response,
+    link: Link,
+    progress_bar: BarType,
+    total_length: int | None,
+    range_start: int | None = 0,
+) -> Iterable[bytes]:
+    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:
+        if range_start:
+            logged_url = (
+                f"{logged_url} ({format_size(range_start)}/{format_size(total_length)})"
+            )
+        else:
+            logged_url = f"{logged_url} ({format_size(total_length)})"
+
+    if is_from_cache(resp):
+        logger.info("Using cached %s", logged_url)
+    elif range_start:
+        logger.info("Resuming download %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 > (512 * 1024):
+        show_progress = True
+    else:
+        show_progress = False
+
+    chunks = response_chunks(resp)
+
+    if not show_progress:
+        return chunks
+
+    renderer = get_download_progress_renderer(
+        bar_type=progress_bar, size=total_length, initial_progress=range_start
+    )
+    return renderer(chunks)
+
+
+def sanitize_content_filename(filename: str) -> str:
+    """
+    Sanitize the "filename" value from a Content-Disposition header.
+    """
+    return os.path.basename(filename)
+
+
+def parse_content_disposition(content_disposition: str, default_filename: str) -> str:
+    """
+    Parse the "filename" value from a Content-Disposition header, and
+    return the default filename if the result is empty.
+    """
+    m = email.message.Message()
+    m["content-type"] = content_disposition
+    filename = m.get_param("filename")
+    if filename:
+        # We need to sanitize the filename to prevent directory traversal
+        # in case the filename contains ".." path parts.
+        filename = sanitize_content_filename(str(filename))
+    return filename or default_filename
+
+
+def _get_http_response_filename(resp: Response, link: 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: str | None = splitext(filename)[1]
+    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
+
+
+@dataclass
+class _FileDownload:
+    """Stores the state of a single link download."""
+
+    link: Link
+    output_file: BinaryIO
+    size: int | None
+    bytes_received: int = 0
+    reattempts: int = 0
+
+    def is_incomplete(self) -> bool:
+        return bool(self.size is not None and self.bytes_received < self.size)
+
+    def write_chunk(self, data: bytes) -> None:
+        self.bytes_received += len(data)
+        self.output_file.write(data)
+
+    def reset_file(self) -> None:
+        """Delete any saved data and reset progress to zero."""
+        self.output_file.seek(0)
+        self.output_file.truncate()
+        self.bytes_received = 0
+
+
+class Downloader:
+    def __init__(
+        self,
+        session: PipSession,
+        progress_bar: BarType,
+        resume_retries: int,
+    ) -> None:
+        assert (
+            resume_retries >= 0
+        ), "Number of max resume retries must be bigger or equal to zero"
+        self._session = session
+        self._progress_bar = progress_bar
+        self._resume_retries = resume_retries
+
+    def batch(
+        self, links: Iterable[Link], location: str
+    ) -> Iterable[tuple[Link, tuple[str, str]]]:
+        """Convenience method to download multiple links."""
+        for link in links:
+            filepath, content_type = self(link, location)
+            yield link, (filepath, content_type)
+
+    def __call__(self, link: Link, location: str) -> tuple[str, str]:
+        """Download a link and save it under location."""
+        resp = self._http_get(link)
+        download_size = _get_http_response_size(resp)
+
+        filepath = os.path.join(location, _get_http_response_filename(resp, link))
+        with open(filepath, "wb") as content_file:
+            download = _FileDownload(link, content_file, download_size)
+            self._process_response(download, resp)
+            if download.is_incomplete():
+                self._attempt_resumes_or_redownloads(download, resp)
+
+        content_type = resp.headers.get("Content-Type", "")
+        return filepath, content_type
+
+    def _process_response(self, download: _FileDownload, resp: Response) -> None:
+        """Download and save chunks from a response."""
+        chunks = _log_download(
+            resp,
+            download.link,
+            self._progress_bar,
+            download.size,
+            range_start=download.bytes_received,
+        )
+        try:
+            for chunk in chunks:
+                download.write_chunk(chunk)
+        except ReadTimeoutError as e:
+            # If the download size is not known, then give up downloading the file.
+            if download.size is None:
+                raise e
+
+            logger.warning("Connection timed out while downloading.")
+
+    def _attempt_resumes_or_redownloads(
+        self, download: _FileDownload, first_resp: Response
+    ) -> None:
+        """Attempt to resume/restart the download if connection was dropped."""
+
+        while download.reattempts < self._resume_retries and download.is_incomplete():
+            assert download.size is not None
+            download.reattempts += 1
+            logger.warning(
+                "Attempting to resume incomplete download (%s/%s, attempt %d)",
+                format_size(download.bytes_received),
+                format_size(download.size),
+                download.reattempts,
+            )
+
+            try:
+                resume_resp = self._http_get_resume(download, should_match=first_resp)
+                # Fallback: if the server responded with 200 (i.e., the file has
+                # since been modified or range requests are unsupported) or any
+                # other unexpected status, restart the download from the beginning.
+                must_restart = resume_resp.status_code != HTTPStatus.PARTIAL_CONTENT
+                if must_restart:
+                    download.reset_file()
+                    download.size = _get_http_response_size(resume_resp)
+                    first_resp = resume_resp
+
+                self._process_response(download, resume_resp)
+            except (ConnectionError, ReadTimeoutError, OSError):
+                continue
+
+        # No more resume attempts. Raise an error if the download is still incomplete.
+        if download.is_incomplete():
+            os.remove(download.output_file.name)
+            raise IncompleteDownloadError(download)
+
+        # If we successfully completed the download via resume, manually cache it
+        # as a complete response to enable future caching
+        if download.reattempts > 0:
+            self._cache_resumed_download(download, first_resp)
+
+    def _cache_resumed_download(
+        self, download: _FileDownload, original_response: Response
+    ) -> None:
+        """
+        Manually cache a file that was successfully downloaded via resume retries.
+
+        cachecontrol doesn't cache 206 (Partial Content) responses, since they
+        are not complete files. This method manually adds the final file to the
+        cache as though it was downloaded in a single request, so that future
+        requests can use the cache.
+        """
+        url = download.link.url_without_fragment
+        adapter = self._session.get_adapter(url)
+
+        # Check if the adapter is the CacheControlAdapter (i.e. caching is enabled)
+        if not isinstance(adapter, CacheControlAdapter):
+            logger.debug(
+                "Skipping resume download caching: no cache controller for %s", url
+            )
+            return
+
+        # Check SafeFileCache is being used
+        assert isinstance(
+            adapter.cache, SafeFileCache
+        ), "separate body cache not in use!"
+
+        synthetic_request = PreparedRequest()
+        synthetic_request.prepare(method="GET", url=url, headers={})
+
+        synthetic_response_headers = HTTPHeaderDict()
+        for key, value in original_response.headers.items():
+            if key.lower() not in ["content-range", "content-length"]:
+                synthetic_response_headers[key] = value
+        synthetic_response_headers["content-length"] = str(download.size)
+
+        synthetic_response = URLlib3Response(
+            body="",
+            headers=synthetic_response_headers,
+            status=200,
+            preload_content=False,
+        )
+
+        # Save metadata and then stream the file contents to cache.
+        cache_url = adapter.controller.cache_url(url)
+        metadata_blob = adapter.controller.serializer.dumps(
+            synthetic_request, synthetic_response, b""
+        )
+        adapter.cache.set(cache_url, metadata_blob)
+        download.output_file.flush()
+        with open(download.output_file.name, "rb") as f:
+            adapter.cache.set_body_from_io(cache_url, f)
+
+        logger.debug(
+            "Cached resumed download as complete response for future use: %s", url
+        )
+
+    def _http_get_resume(
+        self, download: _FileDownload, should_match: Response
+    ) -> Response:
+        """Issue a HTTP range request to resume the download."""
+        # To better understand the download resumption logic, see the mdn web docs:
+        # https://developer.mozilla.org/en-US/docs/Web/HTTP/Guides/Range_requests
+        headers = HEADERS.copy()
+        headers["Range"] = f"bytes={download.bytes_received}-"
+        # If possible, use a conditional range request to avoid corrupted
+        # downloads caused by the remote file changing in-between.
+        if identifier := _get_http_response_etag_or_last_modified(should_match):
+            headers["If-Range"] = identifier
+        return self._http_get(download.link, headers)
+
+    def _http_get(self, link: Link, headers: Mapping[str, str] = HEADERS) -> Response:
+        target_url = link.url_without_fragment
+        try:
+            resp = self._session.get(target_url, headers=headers, stream=True)
+            raise_for_status(resp)
+        except NetworkConnectionError as e:
+            assert e.response is not None
+            logger.critical(
+                "HTTP error %s while getting %s", e.response.status_code, link
+            )
+            raise
+        return resp

venv/Lib/site-packages/pip/_internal/network/lazy_wheel.py

@@ -0,0 +1,213 @@
+"""Lazy ZIP over HTTP"""
+
+from __future__ import annotations
+
+__all__ = ["HTTPRangeRequestUnsupported", "dist_from_wheel_url"]
+
+from bisect import bisect_left, bisect_right
+from collections.abc import Generator
+from contextlib import contextmanager
+from tempfile import NamedTemporaryFile
+from typing import Any
+from zipfile import BadZipFile, ZipFile
+
+from pip._vendor.packaging.utils import canonicalize_name
+from pip._vendor.requests.models import CONTENT_CHUNK_SIZE, Response
+
+from pip._internal.metadata import BaseDistribution, MemoryWheel, get_wheel_distribution
+from pip._internal.network.session import PipSession
+from pip._internal.network.utils import HEADERS, raise_for_status, response_chunks
+
+
+class HTTPRangeRequestUnsupported(Exception):
+    pass
+
+
+def dist_from_wheel_url(name: str, url: str, session: PipSession) -> BaseDistribution:
+    """Return a distribution object from the given wheel URL.
+
+    This uses HTTP range requests to only fetch the portion of the wheel
+    containing metadata, just enough for the object to be constructed.
+    If such requests are not supported, HTTPRangeRequestUnsupported
+    is raised.
+    """
+    with LazyZipOverHTTP(url, session) as zf:
+        # For read-only ZIP files, ZipFile only needs methods read,
+        # seek, seekable and tell, not the whole IO protocol.
+        wheel = MemoryWheel(zf.name, zf)  # type: ignore
+        # After context manager exit, wheel.name
+        # is an invalid file by intention.
+        return get_wheel_distribution(wheel, canonicalize_name(name))
+
+
+class LazyZipOverHTTP:
+    """File-like object mapped to a ZIP file over HTTP.
+
+    This uses HTTP range requests to lazily fetch the file's content,
+    which is supposed to be fed to ZipFile.  If such requests are not
+    supported by the server, raise HTTPRangeRequestUnsupported
+    during initialization.
+    """
+
+    def __init__(
+        self, url: str, session: PipSession, chunk_size: int = CONTENT_CHUNK_SIZE
+    ) -> None:
+        head = session.head(url, headers=HEADERS)
+        raise_for_status(head)
+        assert head.status_code == 200
+        self._session, self._url, self._chunk_size = session, url, chunk_size
+        self._length = int(head.headers["Content-Length"])
+        self._file = NamedTemporaryFile()
+        self.truncate(self._length)
+        self._left: list[int] = []
+        self._right: list[int] = []
+        if "bytes" not in head.headers.get("Accept-Ranges", "none"):
+            raise HTTPRangeRequestUnsupported("range request is not supported")
+        self._check_zip()
+
+    @property
+    def mode(self) -> str:
+        """Opening mode, which is always rb."""
+        return "rb"
+
+    @property
+    def name(self) -> str:
+        """Path to the underlying file."""
+        return self._file.name
+
+    def seekable(self) -> bool:
+        """Return whether random access is supported, which is True."""
+        return True
+
+    def close(self) -> None:
+        """Close the file."""
+        self._file.close()
+
+    @property
+    def closed(self) -> bool:
+        """Whether the file is closed."""
+        return self._file.closed
+
+    def read(self, size: int = -1) -> bytes:
+        """Read up to size bytes from the object and return them.
+
+        As a convenience, if size is unspecified or -1,
+        all bytes until EOF are returned.  Fewer than
+        size bytes may be returned if EOF is reached.
+        """
+        download_size = max(size, self._chunk_size)
+        start, length = self.tell(), self._length
+        stop = length if size < 0 else min(start + download_size, length)
+        start = max(0, stop - download_size)
+        self._download(start, stop - 1)
+        return self._file.read(size)
+
+    def readable(self) -> bool:
+        """Return whether the file is readable, which is True."""
+        return True
+
+    def seek(self, offset: int, whence: int = 0) -> int:
+        """Change stream position and return the new absolute position.
+
+        Seek to offset relative position indicated by whence:
+        * 0: Start of stream (the default).  pos should be >= 0;
+        * 1: Current position - pos may be negative;
+        * 2: End of stream - pos usually negative.
+        """
+        return self._file.seek(offset, whence)
+
+    def tell(self) -> int:
+        """Return the current position."""
+        return self._file.tell()
+
+    def truncate(self, size: int | None = None) -> int:
+        """Resize the stream to the given size in bytes.
+
+        If size is unspecified resize to the current position.
+        The current stream position isn't changed.
+
+        Return the new file size.
+        """
+        return self._file.truncate(size)
+
+    def writable(self) -> bool:
+        """Return False."""
+        return False
+
+    def __enter__(self) -> LazyZipOverHTTP:
+        self._file.__enter__()
+        return self
+
+    def __exit__(self, *exc: Any) -> None:
+        self._file.__exit__(*exc)
+
+    @contextmanager
+    def _stay(self) -> Generator[None, None, None]:
+        """Return a context manager keeping the position.
+
+        At the end of the block, seek back to original position.
+        """
+        pos = self.tell()
+        try:
+            yield
+        finally:
+            self.seek(pos)
+
+    def _check_zip(self) -> None:
+        """Check and download until the file is a valid ZIP."""
+        end = self._length - 1
+        for start in reversed(range(0, end, self._chunk_size)):
+            self._download(start, end)
+            with self._stay():
+                try:
+                    # For read-only ZIP files, ZipFile only needs
+                    # methods read, seek, seekable and tell.
+                    ZipFile(self)
+                except BadZipFile:
+                    pass
+                else:
+                    break
+
+    def _stream_response(
+        self, start: int, end: int, base_headers: dict[str, str] = HEADERS
+    ) -> Response:
+        """Return HTTP response to a range request from start to end."""
+        headers = base_headers.copy()
+        headers["Range"] = f"bytes={start}-{end}"
+        # TODO: Get range requests to be correctly cached
+        headers["Cache-Control"] = "no-cache"
+        return self._session.get(self._url, headers=headers, stream=True)
+
+    def _merge(
+        self, start: int, end: int, left: int, right: int
+    ) -> Generator[tuple[int, int], None, None]:
+        """Return a generator of intervals to be fetched.
+
+        Args:
+            start (int): Start of needed interval
+            end (int): End of needed interval
+            left (int): Index of first overlapping downloaded data
+            right (int): Index after last overlapping downloaded data
+        """
+        lslice, rslice = self._left[left:right], self._right[left:right]
+        i = start = min([start] + lslice[:1])
+        end = max([end] + rslice[-1:])
+        for j, k in zip(lslice, rslice):
+            if j > i:
+                yield i, j - 1
+            i = k + 1
+        if i <= end:
+            yield i, end
+        self._left[left:right], self._right[left:right] = [start], [end]
+
+    def _download(self, start: int, end: int) -> None:
+        """Download bytes from start to end inclusively."""
+        with self._stay():
+            left = bisect_left(self._right, start)
+            right = bisect_right(self._left, end)
+            for start, end in self._merge(start, end, left, right):
+                response = self._stream_response(start, end)
+                response.raise_for_status()
+                self.seek(start)
+                for chunk in response_chunks(response, self._chunk_size):
+                    self._file.write(chunk)