Spaces:
Sleeping
Sleeping
File size: 25,091 Bytes
6219ec9 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 |
"""Prepares a distribution for installation
"""
# The following comment should be removed at some point in the future.
# mypy: strict-optional=False
import logging
import mimetypes
import os
import shutil
from typing import Dict, Iterable, List, Optional
from pip._vendor.packaging.utils import canonicalize_name
from pip._internal.distributions import make_distribution_for_install_requirement
from pip._internal.distributions.installed import InstalledDistribution
from pip._internal.exceptions import (
DirectoryUrlHashUnsupported,
HashMismatch,
HashUnpinned,
InstallationError,
MetadataInconsistent,
NetworkConnectionError,
PreviousBuildDirError,
VcsHashUnsupported,
)
from pip._internal.index.package_finder import PackageFinder
from pip._internal.metadata import BaseDistribution, get_metadata_distribution
from pip._internal.models.direct_url import ArchiveInfo
from pip._internal.models.link import Link
from pip._internal.models.wheel import Wheel
from pip._internal.network.download import BatchDownloader, Downloader
from pip._internal.network.lazy_wheel import (
HTTPRangeRequestUnsupported,
dist_from_wheel_url,
)
from pip._internal.network.session import PipSession
from pip._internal.operations.build.build_tracker import BuildTracker
from pip._internal.req.req_install import InstallRequirement
from pip._internal.utils.direct_url_helpers import (
direct_url_for_editable,
direct_url_from_link,
)
from pip._internal.utils.hashes import Hashes, MissingHashes
from pip._internal.utils.logging import indent_log
from pip._internal.utils.misc import (
display_path,
hash_file,
hide_url,
is_installable_dir,
)
from pip._internal.utils.temp_dir import TempDirectory
from pip._internal.utils.unpacking import unpack_file
from pip._internal.vcs import vcs
logger = logging.getLogger(__name__)
def _get_prepared_distribution(
req: InstallRequirement,
build_tracker: BuildTracker,
finder: PackageFinder,
build_isolation: bool,
check_build_deps: bool,
) -> BaseDistribution:
"""Prepare a distribution for installation."""
abstract_dist = make_distribution_for_install_requirement(req)
with build_tracker.track(req):
abstract_dist.prepare_distribution_metadata(
finder, build_isolation, check_build_deps
)
return abstract_dist.get_metadata_distribution()
def unpack_vcs_link(link: Link, location: str, verbosity: int) -> None:
vcs_backend = vcs.get_backend_for_scheme(link.scheme)
assert vcs_backend is not None
vcs_backend.unpack(location, url=hide_url(link.url), verbosity=verbosity)
class File:
def __init__(self, path: str, content_type: Optional[str]) -> None:
self.path = path
if content_type is None:
self.content_type = mimetypes.guess_type(path)[0]
else:
self.content_type = content_type
def get_http_url(
link: Link,
download: Downloader,
download_dir: Optional[str] = None,
hashes: Optional[Hashes] = None,
) -> File:
temp_dir = TempDirectory(kind="unpack", globally_managed=True)
# If a download dir is specified, is the file already downloaded there?
already_downloaded_path = None
if download_dir:
already_downloaded_path = _check_download_dir(link, download_dir, hashes)
if already_downloaded_path:
from_path = already_downloaded_path
content_type = None
else:
# let's download to a tmp dir
from_path, content_type = download(link, temp_dir.path)
if hashes:
hashes.check_against_path(from_path)
return File(from_path, content_type)
def get_file_url(
link: Link, download_dir: Optional[str] = None, hashes: Optional[Hashes] = None
) -> File:
"""Get file and optionally check its hash."""
# If a download dir is specified, is the file already there and valid?
already_downloaded_path = None
if download_dir:
already_downloaded_path = _check_download_dir(link, download_dir, hashes)
if already_downloaded_path:
from_path = already_downloaded_path
else:
from_path = link.file_path
# If --require-hashes is off, `hashes` is either empty, the
# link's embedded hash, or MissingHashes; it is required to
# match. If --require-hashes is on, we are satisfied by any
# hash in `hashes` matching: a URL-based or an option-based
# one; no internet-sourced hash will be in `hashes`.
if hashes:
hashes.check_against_path(from_path)
return File(from_path, None)
def unpack_url(
link: Link,
location: str,
download: Downloader,
verbosity: int,
download_dir: Optional[str] = None,
hashes: Optional[Hashes] = None,
) -> Optional[File]:
"""Unpack link into location, downloading if required.
:param hashes: A Hashes object, one of whose embedded hashes must match,
or HashMismatch will be raised. If the Hashes is empty, no matches are
required, and unhashable types of requirements (like VCS ones, which
would ordinarily raise HashUnsupported) are allowed.
"""
# non-editable vcs urls
if link.is_vcs:
unpack_vcs_link(link, location, verbosity=verbosity)
return None
assert not link.is_existing_dir()
# file urls
if link.is_file:
file = get_file_url(link, download_dir, hashes=hashes)
# http urls
else:
file = get_http_url(
link,
download,
download_dir,
hashes=hashes,
)
# unpack the archive to the build dir location. even when only downloading
# archives, they have to be unpacked to parse dependencies, except wheels
if not link.is_wheel:
unpack_file(file.path, location, file.content_type)
return file
def _check_download_dir(
link: Link, download_dir: str, hashes: Optional[Hashes]
) -> Optional[str]:
"""Check download_dir for previously downloaded file with correct hash
If a correct file is found return its path else None
"""
download_path = os.path.join(download_dir, link.filename)
if not os.path.exists(download_path):
return None
# If already downloaded, does its hash match?
logger.info("File was already downloaded %s", download_path)
if hashes:
try:
hashes.check_against_path(download_path)
except HashMismatch:
logger.warning(
"Previously-downloaded file %s has bad hash. Re-downloading.",
download_path,
)
os.unlink(download_path)
return None
return download_path
class RequirementPreparer:
"""Prepares a Requirement"""
def __init__(
self,
build_dir: str,
download_dir: Optional[str],
src_dir: str,
build_isolation: bool,
check_build_deps: bool,
build_tracker: BuildTracker,
session: PipSession,
progress_bar: str,
finder: PackageFinder,
require_hashes: bool,
use_user_site: bool,
lazy_wheel: bool,
verbosity: int,
) -> None:
super().__init__()
self.src_dir = src_dir
self.build_dir = build_dir
self.build_tracker = build_tracker
self._session = session
self._download = Downloader(session, progress_bar)
self._batch_download = BatchDownloader(session, progress_bar)
self.finder = finder
# Where still-packed archives should be written to. If None, they are
# not saved, and are deleted immediately after unpacking.
self.download_dir = download_dir
# Is build isolation allowed?
self.build_isolation = build_isolation
# Should check build dependencies?
self.check_build_deps = check_build_deps
# Should hash-checking be required?
self.require_hashes = require_hashes
# Should install in user site-packages?
self.use_user_site = use_user_site
# Should wheels be downloaded lazily?
self.use_lazy_wheel = lazy_wheel
# How verbose should underlying tooling be?
self.verbosity = verbosity
# Memoized downloaded files, as mapping of url: path.
self._downloaded: Dict[str, str] = {}
# Previous "header" printed for a link-based InstallRequirement
self._previous_requirement_header = ("", "")
def _log_preparing_link(self, req: InstallRequirement) -> None:
"""Provide context for the requirement being prepared."""
if req.link.is_file and not req.original_link_is_in_wheel_cache:
message = "Processing %s"
information = str(display_path(req.link.file_path))
else:
message = "Collecting %s"
information = str(req.req or req)
if (message, information) != self._previous_requirement_header:
self._previous_requirement_header = (message, information)
logger.info(message, information)
if req.original_link_is_in_wheel_cache:
with indent_log():
logger.info("Using cached %s", req.link.filename)
def _ensure_link_req_src_dir(
self, req: InstallRequirement, parallel_builds: bool
) -> None:
"""Ensure source_dir of a linked InstallRequirement."""
# Since source_dir is only set for editable requirements.
if req.link.is_wheel:
# We don't need to unpack wheels, so no need for a source
# directory.
return
assert req.source_dir is None
if req.link.is_existing_dir():
# build local directories in-tree
req.source_dir = req.link.file_path
return
# We always delete unpacked sdists after pip runs.
req.ensure_has_source_dir(
self.build_dir,
autodelete=True,
parallel_builds=parallel_builds,
)
# If a checkout exists, it's unwise to keep going. version
# inconsistencies are logged later, but do not fail the
# installation.
# FIXME: this won't upgrade when there's an existing
# package unpacked in `req.source_dir`
# TODO: this check is now probably dead code
if is_installable_dir(req.source_dir):
raise PreviousBuildDirError(
"pip can't proceed with requirements '{}' due to a"
"pre-existing build directory ({}). This is likely "
"due to a previous installation that failed . pip is "
"being responsible and not assuming it can delete this. "
"Please delete it and try again.".format(req, req.source_dir)
)
def _get_linked_req_hashes(self, req: InstallRequirement) -> Hashes:
# By the time this is called, the requirement's link should have
# been checked so we can tell what kind of requirements req is
# and raise some more informative errors than otherwise.
# (For example, we can raise VcsHashUnsupported for a VCS URL
# rather than HashMissing.)
if not self.require_hashes:
return req.hashes(trust_internet=True)
# We could check these first 2 conditions inside unpack_url
# and save repetition of conditions, but then we would
# report less-useful error messages for unhashable
# requirements, complaining that there's no hash provided.
if req.link.is_vcs:
raise VcsHashUnsupported()
if req.link.is_existing_dir():
raise DirectoryUrlHashUnsupported()
# Unpinned packages are asking for trouble when a new version
# is uploaded. This isn't a security check, but it saves users
# a surprising hash mismatch in the future.
# file:/// URLs aren't pinnable, so don't complain about them
# not being pinned.
if req.original_link is None and not req.is_pinned:
raise HashUnpinned()
# If known-good hashes are missing for this requirement,
# shim it with a facade object that will provoke hash
# computation and then raise a HashMissing exception
# showing the user what the hash should be.
return req.hashes(trust_internet=False) or MissingHashes()
def _fetch_metadata_only(
self,
req: InstallRequirement,
) -> Optional[BaseDistribution]:
if self.require_hashes:
logger.debug(
"Metadata-only fetching is not used as hash checking is required",
)
return None
# Try PEP 658 metadata first, then fall back to lazy wheel if unavailable.
return self._fetch_metadata_using_link_data_attr(
req
) or self._fetch_metadata_using_lazy_wheel(req.link)
def _fetch_metadata_using_link_data_attr(
self,
req: InstallRequirement,
) -> Optional[BaseDistribution]:
"""Fetch metadata from the data-dist-info-metadata attribute, if possible."""
# (1) Get the link to the metadata file, if provided by the backend.
metadata_link = req.link.metadata_link()
if metadata_link is None:
return None
assert req.req is not None
logger.info(
"Obtaining dependency information for %s from %s",
req.req,
metadata_link,
)
# (2) Download the contents of the METADATA file, separate from the dist itself.
metadata_file = get_http_url(
metadata_link,
self._download,
hashes=metadata_link.as_hashes(),
)
with open(metadata_file.path, "rb") as f:
metadata_contents = f.read()
# (3) Generate a dist just from those file contents.
metadata_dist = get_metadata_distribution(
metadata_contents,
req.link.filename,
req.req.name,
)
# (4) Ensure the Name: field from the METADATA file matches the name from the
# install requirement.
#
# NB: raw_name will fall back to the name from the install requirement if
# the Name: field is not present, but it's noted in the raw_name docstring
# that that should NEVER happen anyway.
if metadata_dist.raw_name != req.req.name:
raise MetadataInconsistent(
req, "Name", req.req.name, metadata_dist.raw_name
)
return metadata_dist
def _fetch_metadata_using_lazy_wheel(
self,
link: Link,
) -> Optional[BaseDistribution]:
"""Fetch metadata using lazy wheel, if possible."""
# --use-feature=fast-deps must be provided.
if not self.use_lazy_wheel:
return None
if link.is_file or not link.is_wheel:
logger.debug(
"Lazy wheel is not used as %r does not point to a remote wheel",
link,
)
return None
wheel = Wheel(link.filename)
name = canonicalize_name(wheel.name)
logger.info(
"Obtaining dependency information from %s %s",
name,
wheel.version,
)
url = link.url.split("#", 1)[0]
try:
return dist_from_wheel_url(name, url, self._session)
except HTTPRangeRequestUnsupported:
logger.debug("%s does not support range requests", url)
return None
def _complete_partial_requirements(
self,
partially_downloaded_reqs: Iterable[InstallRequirement],
parallel_builds: bool = False,
) -> None:
"""Download any requirements which were only fetched by metadata."""
# Download to a temporary directory. These will be copied over as
# needed for downstream 'download', 'wheel', and 'install' commands.
temp_dir = TempDirectory(kind="unpack", globally_managed=True).path
# Map each link to the requirement that owns it. This allows us to set
# `req.local_file_path` on the appropriate requirement after passing
# all the links at once into BatchDownloader.
links_to_fully_download: Dict[Link, InstallRequirement] = {}
for req in partially_downloaded_reqs:
assert req.link
links_to_fully_download[req.link] = req
batch_download = self._batch_download(
links_to_fully_download.keys(),
temp_dir,
)
for link, (filepath, _) in batch_download:
logger.debug("Downloading link %s to %s", link, filepath)
req = links_to_fully_download[link]
req.local_file_path = filepath
# This step is necessary to ensure all lazy wheels are processed
# successfully by the 'download', 'wheel', and 'install' commands.
for req in partially_downloaded_reqs:
self._prepare_linked_requirement(req, parallel_builds)
def prepare_linked_requirement(
self, req: InstallRequirement, parallel_builds: bool = False
) -> BaseDistribution:
"""Prepare a requirement to be obtained from req.link."""
assert req.link
self._log_preparing_link(req)
with indent_log():
# Check if the relevant file is already available
# in the download directory
file_path = None
if self.download_dir is not None and req.link.is_wheel:
hashes = self._get_linked_req_hashes(req)
file_path = _check_download_dir(req.link, self.download_dir, hashes)
if file_path is not None:
# The file is already available, so mark it as downloaded
self._downloaded[req.link.url] = file_path
else:
# The file is not available, attempt to fetch only metadata
metadata_dist = self._fetch_metadata_only(req)
if metadata_dist is not None:
req.needs_more_preparation = True
return metadata_dist
# None of the optimizations worked, fully prepare the requirement
return self._prepare_linked_requirement(req, parallel_builds)
def prepare_linked_requirements_more(
self, reqs: Iterable[InstallRequirement], parallel_builds: bool = False
) -> None:
"""Prepare linked requirements more, if needed."""
reqs = [req for req in reqs if req.needs_more_preparation]
for req in reqs:
# Determine if any of these requirements were already downloaded.
if self.download_dir is not None and req.link.is_wheel:
hashes = self._get_linked_req_hashes(req)
file_path = _check_download_dir(req.link, self.download_dir, hashes)
if file_path is not None:
self._downloaded[req.link.url] = file_path
req.needs_more_preparation = False
# Prepare requirements we found were already downloaded for some
# reason. The other downloads will be completed separately.
partially_downloaded_reqs: List[InstallRequirement] = []
for req in reqs:
if req.needs_more_preparation:
partially_downloaded_reqs.append(req)
else:
self._prepare_linked_requirement(req, parallel_builds)
# TODO: separate this part out from RequirementPreparer when the v1
# resolver can be removed!
self._complete_partial_requirements(
partially_downloaded_reqs,
parallel_builds=parallel_builds,
)
def _prepare_linked_requirement(
self, req: InstallRequirement, parallel_builds: bool
) -> BaseDistribution:
assert req.link
link = req.link
self._ensure_link_req_src_dir(req, parallel_builds)
hashes = self._get_linked_req_hashes(req)
if link.is_existing_dir():
local_file = None
elif link.url not in self._downloaded:
try:
local_file = unpack_url(
link,
req.source_dir,
self._download,
self.verbosity,
self.download_dir,
hashes,
)
except NetworkConnectionError as exc:
raise InstallationError(
"Could not install requirement {} because of HTTP "
"error {} for URL {}".format(req, exc, link)
)
else:
file_path = self._downloaded[link.url]
if hashes:
hashes.check_against_path(file_path)
local_file = File(file_path, content_type=None)
# If download_info is set, we got it from the wheel cache.
if req.download_info is None:
# Editables don't go through this function (see
# prepare_editable_requirement).
assert not req.editable
req.download_info = direct_url_from_link(link, req.source_dir)
# Make sure we have a hash in download_info. If we got it as part of the
# URL, it will have been verified and we can rely on it. Otherwise we
# compute it from the downloaded file.
if (
isinstance(req.download_info.info, ArchiveInfo)
and not req.download_info.info.hash
and local_file
):
hash = hash_file(local_file.path)[0].hexdigest()
req.download_info.info.hash = f"sha256={hash}"
# For use in later processing,
# preserve the file path on the requirement.
if local_file:
req.local_file_path = local_file.path
dist = _get_prepared_distribution(
req,
self.build_tracker,
self.finder,
self.build_isolation,
self.check_build_deps,
)
return dist
def save_linked_requirement(self, req: InstallRequirement) -> None:
assert self.download_dir is not None
assert req.link is not None
link = req.link
if link.is_vcs or (link.is_existing_dir() and req.editable):
# Make a .zip of the source_dir we already created.
req.archive(self.download_dir)
return
if link.is_existing_dir():
logger.debug(
"Not copying link to destination directory "
"since it is a directory: %s",
link,
)
return
if req.local_file_path is None:
# No distribution was downloaded for this requirement.
return
download_location = os.path.join(self.download_dir, link.filename)
if not os.path.exists(download_location):
shutil.copy(req.local_file_path, download_location)
download_path = display_path(download_location)
logger.info("Saved %s", download_path)
def prepare_editable_requirement(
self,
req: InstallRequirement,
) -> BaseDistribution:
"""Prepare an editable requirement."""
assert req.editable, "cannot prepare a non-editable req as editable"
logger.info("Obtaining %s", req)
with indent_log():
if self.require_hashes:
raise InstallationError(
"The editable requirement {} cannot be installed when "
"requiring hashes, because there is no single file to "
"hash.".format(req)
)
req.ensure_has_source_dir(self.src_dir)
req.update_editable()
assert req.source_dir
req.download_info = direct_url_for_editable(req.unpacked_source_directory)
dist = _get_prepared_distribution(
req,
self.build_tracker,
self.finder,
self.build_isolation,
self.check_build_deps,
)
req.check_if_exists(self.use_user_site)
return dist
def prepare_installed_requirement(
self,
req: InstallRequirement,
skip_reason: str,
) -> BaseDistribution:
"""Prepare an already-installed requirement."""
assert req.satisfied_by, "req should have been satisfied but isn't"
assert skip_reason is not None, (
"did not get skip reason skipped but req.satisfied_by "
"is set to {}".format(req.satisfied_by)
)
logger.info(
"Requirement %s: %s (%s)", skip_reason, req, req.satisfied_by.version
)
with indent_log():
if self.require_hashes:
logger.debug(
"Since it is already installed, we are trusting this "
"package without checking its hash. To ensure a "
"completely repeatable environment, install into an "
"empty virtualenv."
)
return InstalledDistribution(req).get_metadata_distribution()
|