import base64 import dataclasses import logging import random import re import time from abc import ABC, abstractmethod from datetime import datetime from ftplib import FTP from io import BytesIO from pathlib import Path from posixpath import split from shutil import rmtree from tempfile import mkdtemp from typing import Any from unittest import mock from unittest.mock import MagicMock import attr import pytest from itemadapter import ItemAdapter from twisted.internet.defer import Deferred from twisted.python.failure import Failure from scrapy.crawler import Crawler from scrapy.exceptions import IgnoreRequest, NotConfigured from scrapy.http import Request, Response from scrapy.item import Field, Item from scrapy.pipelines.files import ( FileException, FilesPipeline, FSFilesStore, FTPFilesStore, GCSFilesStore, S3FilesStore, ) from scrapy.pipelines.media import _MediaRequestFiltered from scrapy.settings import Settings from scrapy.utils.asyncio import call_later from scrapy.utils.defer import maybe_deferred_to_future from scrapy.utils.spider import DefaultSpider from scrapy.utils.test import get_crawler from tests.mockserver.ftp import MockFTPServer from tests.utils.decorators import coroutine_test, inline_callbacks_test from .utils.cloud import mock_google_cloud_storage from .utils.media_pipelines import DUMMY_SPIDER_INFO, mocked_download_func def get_ftp_content_and_delete( path: str, host: str, port: int, username: str, password: str, use_active_mode: bool = False, ) -> bytes: with FTP() as ftp: ftp.connect(host, port) ftp.login(username, password) if use_active_mode: ftp.set_pasv(False) ftp_data: list[bytes] = [] def buffer_data(data: bytes) -> None: ftp_data.append(data) ftp.retrbinary(f"RETR {path}", buffer_data) dirname, filename = split(path) ftp.cwd(dirname) ftp.delete(filename) return b"".join(ftp_data) class DeferredFSFilesStore(FSFilesStore): """A simple store with persist_file() returning a deferred.""" def persist_file(self, path, buf, info, meta=None, headers=None): deferred: Deferred[None] = Deferred() # short-hand super() doesn't work in nested functions parent_persist_file = super().persist_file def cb(): parent_persist_file(path, buf, info, meta=meta, headers=headers) deferred.callback(None) call_later(0.5, cb) return deferred class TestFilesPipeline: def setup_method(self): self.tempdir = mkdtemp() self.pipeline = self._create_pipeline(FilesPipeline) def teardown_method(self): rmtree(self.tempdir) def _create_pipeline(self, pipeline_cls: type[FilesPipeline]) -> FilesPipeline: crawler = get_crawler(DefaultSpider, {"FILES_STORE": self.tempdir}) crawler.spider = crawler._create_spider() crawler.engine = MagicMock(download_async=mocked_download_func) pipeline = pipeline_cls.from_crawler(crawler) pipeline.open_spider() return pipeline def test_file_path_query_parameters(self): file_path = self.pipeline.file_path req1 = Request("http://foo.bar/baz.txt?fizz") assert file_path(req1) == "full/a2b4913a62f65445aeae2bac08cd8c3b41d7195e.txt" req2 = Request("http://foo.bar/get_img.foo?file=photo.jpg") assert file_path(req2) == "full/7fc9461c9fd836515bea6983373097203a7d748e.jpg" def test_file_path(self): file_path = self.pipeline.file_path assert ( file_path(Request("https://dev.mydeco.com/mydeco.pdf")) == "full/c9b564df929f4bc635bdd19fde4f3d4847c757c5.pdf" ) assert ( file_path( Request( "http://www.maddiebrown.co.uk///catalogue-items//image_54642_12175_95307.txt" ) ) == "full/4ce274dd83db0368bafd7e406f382ae088e39219.txt" ) assert ( file_path( Request("https://dev.mydeco.com/two/dirs/with%20spaces%2Bsigns.doc") ) == "full/94ccc495a17b9ac5d40e3eabf3afcb8c2c9b9e1a.doc" ) assert ( file_path( Request( "http://www.dfsonline.co.uk/get_prod_image?img=status_0907_mdm.jpg" ) ) == "full/c67f916ff9d542e822dedf38f9fcb146d1faba78.jpg" ) assert ( file_path(Request("http://www.dorma.co.uk/images/product_details/2532/")) == "full/97ee6f8a46cbbb418ea91502fd24176865cf39b2" ) assert ( file_path(Request("http://www.dorma.co.uk/images/product_details/2532")) == "full/244e0dd7d96a3b7b01f54eded250c9e272577aa1" ) assert ( file_path( Request("http://www.dorma.co.uk/images/product_details/2532"), response=Response("http://www.dorma.co.uk/images/product_details/2532"), info=object(), # type: ignore[arg-type] ) == "full/244e0dd7d96a3b7b01f54eded250c9e272577aa1" ) assert ( file_path( Request( "http://www.dfsonline.co.uk/get_prod_image?img=status_0907_mdm.jpg.bohaha" ) ) == "full/e75f2fa260521b56f6b6a867447b8002d00b5841" ) assert ( file_path( Request( "data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAR0AAACxCAMAAADOHZloAAACClBMVEX/\ //+F0tzCwMK76ZKQ21AMqr7oAAC96JvD5aWM2kvZ78J0N7fmAAC46Y4Ap7y" ) ) == "full/178059cbeba2e34120a67f2dc1afc3ecc09b61cb.png" ) def test_fs_store(self): assert isinstance(self.pipeline.store, FSFilesStore) assert self.pipeline.store.basedir == self.tempdir path = "some/image/key.jpg" fullpath = Path(self.tempdir, "some", "image", "key.jpg") assert self.pipeline.store._get_filesystem_path(path) == fullpath @coroutine_test async def test_file_not_expired(self): item_url = "http://example.com/file.pdf" item = _create_item_with_files(item_url) with ( mock.patch.object(FilesPipeline, "inc_stats", return_value=True), mock.patch.object( FSFilesStore, "stat_file", return_value={"checksum": "abc", "last_modified": time.time()}, ), mock.patch.object( FilesPipeline, "get_media_requests", return_value=[_prepare_request_object(item_url)], ), ): result = await self.pipeline.process_item(item) assert result["files"][0]["checksum"] == "abc" assert result["files"][0]["status"] == "uptodate" @coroutine_test async def test_file_expired(self): item_url = "http://example.com/file2.pdf" item = _create_item_with_files(item_url) with ( mock.patch.object( FSFilesStore, "stat_file", return_value={ "checksum": "abc", "last_modified": time.time() - (self.pipeline.expires * 60 * 60 * 24 * 2), }, ), mock.patch.object( FilesPipeline, "get_media_requests", return_value=[_prepare_request_object(item_url)], ), mock.patch.object(FilesPipeline, "inc_stats", return_value=True), ): result = await self.pipeline.process_item(item) assert result["files"][0]["checksum"] != "abc" assert result["files"][0]["status"] == "downloaded" @coroutine_test async def test_file_cached(self): item_url = "http://example.com/file3.pdf" item = _create_item_with_files(item_url) with ( mock.patch.object(FilesPipeline, "inc_stats", return_value=True), mock.patch.object( FSFilesStore, "stat_file", return_value={ "checksum": "abc", "last_modified": time.time() - (self.pipeline.expires * 60 * 60 * 24 * 2), }, ), mock.patch.object( FilesPipeline, "get_media_requests", return_value=[_prepare_request_object(item_url, flags=["cached"])], ), ): result = await self.pipeline.process_item(item) assert result["files"][0]["checksum"] != "abc" assert result["files"][0]["status"] == "cached" @coroutine_test async def test_file_stat_without_last_modified(self) -> None: """A stat result without a last modification time forces a download.""" item_url = "http://example.com/file4.pdf" item = _create_item_with_files(item_url) with ( mock.patch.object(FilesPipeline, "inc_stats", return_value=True), mock.patch.object( FSFilesStore, "stat_file", return_value={"checksum": "abc"} ), mock.patch.object( FilesPipeline, "get_media_requests", return_value=[_prepare_request_object(item_url)], ), ): result = await self.pipeline.process_item(item) assert result["files"][0]["checksum"] != "abc" assert result["files"][0]["status"] == "downloaded" @coroutine_test async def test_file_empty_content(self, caplog: pytest.LogCaptureFixture) -> None: item_url = "http://example.com/empty.pdf" item = _create_item_with_files(item_url) request = Request( item_url, meta={"response": Response(item_url, status=200, body=b"")} ) with ( caplog.at_level(logging.WARNING), mock.patch.object( FilesPipeline, "get_media_requests", return_value=[request] ), ): result = await self.pipeline.process_item(item) assert result["files"] == [] assert "File (empty-content): Empty file from" in caplog.text @coroutine_test async def test_file_downloaded_file_exception( self, caplog: pytest.LogCaptureFixture ) -> None: """A FileException from file_downloaded() is logged as a warning and kept as is.""" class FailingFilesPipeline(FilesPipeline): def file_downloaded(self, response, request, info, *, item=None): raise FileException("boom") item_url = "http://example.com/file5.pdf" item = _create_item_with_files(item_url) pipeline = self._create_pipeline(FailingFilesPipeline) with ( caplog.at_level(logging.WARNING), mock.patch.object( FilesPipeline, "get_media_requests", return_value=[_prepare_request_object(item_url)], ), ): result = await pipeline.process_item(item) assert result["files"] == [] records = [ r for r in caplog.records if "Error processing file" in r.getMessage() ] assert len(records) == 1 assert records[0].levelname == "WARNING" assert "boom" in records[0].getMessage() @coroutine_test async def test_file_downloaded_unknown_error( self, caplog: pytest.LogCaptureFixture ) -> None: """Any other exception from file_downloaded() is logged as an error and reported as a FileException.""" class FailingFilesPipeline(FilesPipeline): def file_downloaded(self, response, request, info, *, item=None): raise RuntimeError("boom") item_url = "http://example.com/file6.pdf" item = _create_item_with_files(item_url) pipeline = self._create_pipeline(FailingFilesPipeline) with ( caplog.at_level(logging.WARNING), mock.patch.object( FilesPipeline, "get_media_requests", return_value=[_prepare_request_object(item_url)], ), ): result = await pipeline.process_item(item) assert result["files"] == [] records = [ r for r in caplog.records if "Error processing file" in r.getMessage() ] assert len(records) == 1 assert records[0].levelname == "ERROR" exc_info = records[0].exc_info assert exc_info is not None assert exc_info[0] is RuntimeError @coroutine_test async def test_async_store(self) -> None: """Test that async persist_file() works and is awaited.""" self.pipeline.store = DeferredFSFilesStore(self.tempdir) item_url = "http://example.com/file.pdf" item = _create_item_with_files(item_url) with ( mock.patch.object(FilesPipeline, "inc_stats", return_value=True), mock.patch.object( FilesPipeline, "get_media_requests", return_value=[_prepare_request_object(item_url)], ), ): result = await self.pipeline.process_item(item) assert result["files"][0]["status"] == "downloaded" assert result["files"][0]["checksum"] # check that the file was written by persist_file() path = Path(self.tempdir) / result["files"][0]["path"] assert path.exists() assert path.read_bytes() == b"data" def test_file_path_from_item(self): """ Custom file path based on item data, overriding default implementation """ class CustomFilesPipeline(FilesPipeline): def file_path(self, request, response=None, info=None, item=None) -> str: return f"full/{item.get('path')}" file_path = CustomFilesPipeline.from_crawler( get_crawler(None, {"FILES_STORE": self.tempdir}) ).file_path item = {"path": "path-to-store-file"} request = Request("http://example.com") assert file_path(request, item=item) == "full/path-to-store-file" def test_media_failed_filtered_request( self, caplog: pytest.LogCaptureFixture ) -> None: """A filtered media request (IgnoreRequest) is reported as a _MediaRequestFiltered exception and logged at the DEBUG level, instead of as a download error with a traceback.""" request = Request("http://example.com/file.pdf") reason = "Filtered offsite request to 'example.com'" failure = Failure(IgnoreRequest(reason)) with ( caplog.at_level(logging.DEBUG), pytest.raises(_MediaRequestFiltered, match=re.escape(reason)), ): self.pipeline.media_failed(failure, request, self.pipeline.spiderinfo) assert len(caplog.records) == 1 record = caplog.records[0] assert record.levelname == "DEBUG" assert record.exc_info is None assert reason in record.getMessage() def test_media_failed_download_error( self, caplog: pytest.LogCaptureFixture ) -> None: """A genuine download error is reported as a FileException and logged as a warning.""" request = Request("http://example.com/file.pdf") failure = Failure(Exception("boom")) with caplog.at_level(logging.WARNING), pytest.raises(FileException): self.pipeline.media_failed(failure, request, self.pipeline.spiderinfo) assert len(caplog.records) == 1 assert caplog.records[0].levelname == "WARNING" @coroutine_test async def test_process_item_filtered_request( self, caplog: pytest.LogCaptureFixture ) -> None: """A filtered (e.g. offsite) media request is processed as a failed result without being logged as an error with a traceback.""" item_url = "http://example.com/file.pdf" item = _create_item_with_files(item_url) request = Request( item_url, meta={ "response": IgnoreRequest("Filtered offsite request to 'example.com'") }, ) with ( caplog.at_level(logging.DEBUG), mock.patch.object( FilesPipeline, "get_media_requests", return_value=[request] ), ): result = await self.pipeline.process_item(item) assert result["files"] == [] assert not any(r.levelname in ("WARNING", "ERROR") for r in caplog.records) assert any( "Filtered offsite request to 'example.com'" in r.getMessage() for r in caplog.records ) @pytest.mark.parametrize( "bad_type", [ "http://example.com/file.pdf", ("http://example.com/file.pdf",), {"url": "http://example.com/file.pdf"}, 123, None, ], ) def test_rejects_non_list_file_urls(self, tmp_path, bad_type): pipeline = FilesPipeline.from_crawler( get_crawler(None, {"FILES_STORE": str(tmp_path)}) ) item = ItemWithFiles() item["file_urls"] = bad_type with pytest.raises(TypeError, match="file_urls must be a list of URLs"): list(pipeline.get_media_requests(item, None)) # type: ignore[arg-type] class TestFilesPipelineFieldsMixin(ABC): @property @abstractmethod def item_class(self) -> Any: raise NotImplementedError def test_item_fields_default(self, tmp_path): url = "http://www.example.com/files/1.txt" item = self.item_class(name="item1", file_urls=[url]) pipeline = FilesPipeline.from_crawler( get_crawler(None, {"FILES_STORE": tmp_path}) ) requests = list(pipeline.get_media_requests(item, None)) # type: ignore[arg-type] assert requests[0].url == url results = [(True, {"url": url})] item = pipeline.item_completed(results, item, None) # type: ignore[arg-type] files = ItemAdapter(item).get("files") assert files == [results[0][1]] assert isinstance(item, self.item_class) def test_item_fields_override_settings(self, tmp_path): url = "http://www.example.com/files/1.txt" item = self.item_class(name="item1", custom_file_urls=[url]) pipeline = FilesPipeline.from_crawler( get_crawler( None, { "FILES_STORE": tmp_path, "FILES_URLS_FIELD": "custom_file_urls", "FILES_RESULT_FIELD": "custom_files", }, ) ) requests = list(pipeline.get_media_requests(item, None)) # type: ignore[arg-type] assert requests[0].url == url results = [(True, {"url": url})] item = pipeline.item_completed(results, item, None) # type: ignore[arg-type] custom_files = ItemAdapter(item).get("custom_files") assert custom_files == [results[0][1]] assert isinstance(item, self.item_class) class TestFilesPipelineFieldsDict(TestFilesPipelineFieldsMixin): item_class = dict class FilesPipelineTestItem(Item): name = Field() # default fields file_urls = Field() files = Field() # overridden fields custom_file_urls = Field() custom_files = Field() class TestFilesPipelineFieldsItem(TestFilesPipelineFieldsMixin): item_class = FilesPipelineTestItem @dataclasses.dataclass class FilesPipelineTestDataClass: name: str # default fields file_urls: list[str] = dataclasses.field(default_factory=list) files: list[dict[str, str]] = dataclasses.field(default_factory=list) # overridden fields custom_file_urls: list[str] = dataclasses.field(default_factory=list) custom_files: list[dict[str, str]] = dataclasses.field(default_factory=list) class TestFilesPipelineFieldsDataClass(TestFilesPipelineFieldsMixin): item_class = FilesPipelineTestDataClass @attr.s class FilesPipelineTestAttrsItem: name = attr.ib(default="") # default fields file_urls: list[str] = attr.ib(default=list) files: list[dict[str, str]] = attr.ib(default=list) # overridden fields custom_file_urls: list[str] = attr.ib(default=list) custom_files: list[dict[str, str]] = attr.ib(default=list) class TestFilesPipelineFieldsAttrsItem(TestFilesPipelineFieldsMixin): item_class = FilesPipelineTestAttrsItem class TestFilesPipelineCustomSettings: default_cls_settings = { "EXPIRES": 90, "FILES_URLS_FIELD": "file_urls", "FILES_RESULT_FIELD": "files", } file_cls_attr_settings_map = { ("EXPIRES", "FILES_EXPIRES", "expires"), ("FILES_URLS_FIELD", "FILES_URLS_FIELD", "files_urls_field"), ("FILES_RESULT_FIELD", "FILES_RESULT_FIELD", "files_result_field"), } def _generate_fake_settings( self, tmp_path: Path, prefix: str | None = None ) -> dict[str, Any]: def random_string() -> str: return "".join([chr(random.randint(97, 123)) for _ in range(10)]) settings = { "FILES_EXPIRES": random.randint(100, 1000), "FILES_URLS_FIELD": random_string(), "FILES_RESULT_FIELD": random_string(), "FILES_STORE": tmp_path, } if not prefix: return settings return { prefix.upper() + "_" + k if k != "FILES_STORE" else k: v for k, v in settings.items() } def _generate_fake_pipeline(self) -> type[FilesPipeline]: class UserDefinedFilePipeline(FilesPipeline): EXPIRES = 1001 FILES_URLS_FIELD = "alfa" FILES_RESULT_FIELD = "beta" return UserDefinedFilePipeline def test_different_settings_for_different_instances(self, tmp_path): """ If there are different instances with different settings they should keep different settings. """ custom_settings = self._generate_fake_settings(tmp_path) another_pipeline = FilesPipeline.from_crawler( get_crawler(None, custom_settings) ) one_pipeline = FilesPipeline(tmp_path, crawler=get_crawler(None)) for pipe_attr, settings_attr, pipe_ins_attr in self.file_cls_attr_settings_map: default_value = self.default_cls_settings[pipe_attr] assert getattr(one_pipeline, pipe_attr) == default_value custom_value = custom_settings[settings_attr] assert default_value != custom_value assert getattr(another_pipeline, pipe_ins_attr) == custom_value def test_subclass_attributes_preserved_if_no_settings(self, tmp_path): """ If subclasses override class attributes and there are no special settings those values should be kept. """ pipe_cls = self._generate_fake_pipeline() pipe = pipe_cls.from_crawler(get_crawler(None, {"FILES_STORE": tmp_path})) for pipe_attr, _, pipe_ins_attr in self.file_cls_attr_settings_map: custom_value = getattr(pipe, pipe_ins_attr) assert custom_value != self.default_cls_settings[pipe_attr] assert getattr(pipe, pipe_ins_attr) == getattr(pipe, pipe_attr) def test_subclass_attrs_preserved_custom_settings(self, tmp_path): """ If file settings are defined but they are not defined for subclass settings should be preserved. """ pipeline_cls = self._generate_fake_pipeline() settings = self._generate_fake_settings(tmp_path) pipeline = pipeline_cls.from_crawler(get_crawler(None, settings)) for pipe_attr, settings_attr, pipe_ins_attr in self.file_cls_attr_settings_map: value = getattr(pipeline, pipe_ins_attr) setting_value = settings.get(settings_attr) assert value != self.default_cls_settings[pipe_attr] assert value == setting_value def test_no_custom_settings_for_subclasses(self, tmp_path): """ If there are no settings for subclass and no subclass attributes, pipeline should use attributes of base class. """ class UserDefinedFilesPipeline(FilesPipeline): pass user_pipeline = UserDefinedFilesPipeline.from_crawler( get_crawler(None, {"FILES_STORE": tmp_path}) ) for pipe_attr, _, pipe_ins_attr in self.file_cls_attr_settings_map: # Values from settings for custom pipeline should be set on pipeline instance. custom_value = self.default_cls_settings.get(pipe_attr.upper()) assert getattr(user_pipeline, pipe_ins_attr) == custom_value def test_custom_settings_for_subclasses(self, tmp_path): """ If there are custom settings for subclass and NO class attributes, pipeline should use custom settings. """ class UserDefinedFilesPipeline(FilesPipeline): pass prefix = UserDefinedFilesPipeline.__name__.upper() settings = self._generate_fake_settings(tmp_path, prefix=prefix) user_pipeline = UserDefinedFilesPipeline.from_crawler( get_crawler(None, settings) ) for pipe_attr, settings_attr, pipe_inst_attr in self.file_cls_attr_settings_map: # Values from settings for custom pipeline should be set on pipeline instance. custom_value = settings.get(prefix + "_" + settings_attr) assert custom_value != self.default_cls_settings[pipe_attr] assert getattr(user_pipeline, pipe_inst_attr) == custom_value def test_custom_settings_and_class_attrs_for_subclasses(self, tmp_path): """ If there are custom settings for subclass AND class attributes setting keys are preferred and override attributes. """ pipeline_cls = self._generate_fake_pipeline() prefix = pipeline_cls.__name__.upper() settings = self._generate_fake_settings(tmp_path, prefix=prefix) user_pipeline = pipeline_cls.from_crawler(get_crawler(None, settings)) for ( pipe_cls_attr, settings_attr, pipe_inst_attr, ) in self.file_cls_attr_settings_map: custom_value = settings.get(prefix + "_" + settings_attr) assert custom_value != self.default_cls_settings[pipe_cls_attr] assert getattr(user_pipeline, pipe_inst_attr) == custom_value def test_cls_attrs_with_DEFAULT_prefix(self, tmp_path): class UserDefinedFilesPipeline(FilesPipeline): DEFAULT_FILES_RESULT_FIELD = "this" DEFAULT_FILES_URLS_FIELD = "that" pipeline = UserDefinedFilesPipeline.from_crawler( get_crawler(None, {"FILES_STORE": tmp_path}) ) assert ( pipeline.files_result_field == UserDefinedFilesPipeline.DEFAULT_FILES_RESULT_FIELD ) assert ( pipeline.files_urls_field == UserDefinedFilesPipeline.DEFAULT_FILES_URLS_FIELD ) def test_user_defined_subclass_default_key_names(self, tmp_path): """Test situation when user defines subclass of FilesPipeline, but uses attribute names for default pipeline (without prefixing them with pipeline class name). """ settings = self._generate_fake_settings(tmp_path) class UserPipe(FilesPipeline): pass pipeline_cls = UserPipe.from_crawler(get_crawler(None, settings)) for _, settings_attr, pipe_inst_attr in self.file_cls_attr_settings_map: expected_value = settings.get(settings_attr) assert getattr(pipeline_cls, pipe_inst_attr) == expected_value def test_file_pipeline_using_pathlike_objects(self, tmp_path): class CustomFilesPipelineWithPathLikeDir(FilesPipeline): def file_path(self, request, response=None, info=None, *, item=None) -> str: return str(Path("subdir") / Path(request.url).name) pipeline = CustomFilesPipelineWithPathLikeDir.from_crawler( get_crawler(None, {"FILES_STORE": tmp_path}) ) request = Request("http://example.com/image01.jpg") assert pipeline.file_path(request) == str(Path("subdir/image01.jpg")) class TestFSFilesStore: def test_constructor_with_pathlike_object(self, tmp_path: Path) -> None: assert FSFilesStore(tmp_path).basedir == str(tmp_path) def test_constructor_with_uri(self, tmp_path: Path) -> None: assert FSFilesStore(f"file://{tmp_path}").basedir == str(tmp_path) def test_stat_file(self, tmp_path: Path) -> None: store = FSFilesStore(tmp_path) store.persist_file("full/filename", BytesIO(b"data"), DUMMY_SPIDER_INFO) stat = store.stat_file("full/filename", DUMMY_SPIDER_INFO) assert stat["checksum"] == "8d777f385d3dfec8815d20f7496026dc" assert stat["last_modified"] == pytest.approx(time.time(), abs=60) def test_stat_missing_file(self, tmp_path: Path) -> None: store = FSFilesStore(tmp_path) assert store.stat_file("full/filename", DUMMY_SPIDER_INFO) == {} @pytest.mark.requires_botocore class TestS3FilesStore: @inline_callbacks_test def test_persist(self): bucket = "mybucket" key = "export.csv" uri = f"s3://{bucket}/{key}" buffer = mock.MagicMock() meta = {"foo": "bar"} path = "" content_type = "image/png" store = S3FilesStore(uri) from botocore.stub import Stubber # noqa: PLC0415 with Stubber(store.s3_client) as stub: stub.add_response( "put_object", expected_params={ "ACL": S3FilesStore.POLICY, "Body": buffer, "Bucket": bucket, "CacheControl": S3FilesStore.HEADERS["Cache-Control"], "ContentType": content_type, "Key": key, "Metadata": meta, }, service_response={}, ) yield store.persist_file( path, buffer, info=DUMMY_SPIDER_INFO, meta=meta, headers={"Content-Type": content_type}, ) stub.assert_no_pending_responses() # The call to read does not happen with Stubber assert buffer.method_calls == [mock.call.seek(0)] @inline_callbacks_test def test_persist_without_headers(self): """Without custom headers only the default ones are sent.""" bucket = "mybucket" key = "export.csv" buffer = mock.MagicMock() store = S3FilesStore(f"s3://{bucket}/{key}") from botocore.stub import Stubber # noqa: PLC0415 with Stubber(store.s3_client) as stub: stub.add_response( "put_object", expected_params={ "ACL": S3FilesStore.POLICY, "Body": buffer, "Bucket": bucket, "CacheControl": S3FilesStore.HEADERS["Cache-Control"], "Key": key, "Metadata": {}, }, service_response={}, ) yield store.persist_file("", buffer, info=DUMMY_SPIDER_INFO) stub.assert_no_pending_responses() def test_missing_botocore(self): with ( mock.patch( "scrapy.pipelines.files.is_botocore_available", return_value=False ), pytest.raises(NotConfigured, match="missing botocore library"), ): S3FilesStore("s3://mybucket/key") def test_wrong_uri_scheme(self): with pytest.raises( ValueError, match=re.escape( "Incorrect URI scheme in ftp://mybucket/key, expected 's3'" ), ): S3FilesStore("ftp://mybucket/key") def test_unsupported_header(self): store = S3FilesStore("s3://mybucket/key") with pytest.raises( TypeError, match='Header "X-Custom" is not supported by botocore' ): store._headers_to_botocore_kwargs({"X-Custom": "value"}) @inline_callbacks_test def test_stat(self): bucket = "mybucket" key = "export.csv" uri = f"s3://{bucket}/{key}" checksum = "3187896a9657a28163abb31667df64c8" last_modified = datetime(2019, 12, 1) store = S3FilesStore(uri) from botocore.stub import Stubber # noqa: PLC0415 with Stubber(store.s3_client) as stub: stub.add_response( "head_object", expected_params={ "Bucket": bucket, "Key": key, }, service_response={ "ETag": f'"{checksum}"', "LastModified": last_modified, }, ) file_stats = yield store.stat_file("", info=DUMMY_SPIDER_INFO) assert file_stats == { "checksum": checksum, "last_modified": last_modified.timestamp(), } stub.assert_no_pending_responses() def test_default_max_pool_connections(self) -> None: store = S3FilesStore("s3://mybucket/prefix/") config: Any = store.s3_client.meta.config assert config.max_pool_connections == 10 @pytest.mark.parametrize( ("settings", "expected"), [ ({}, 10), ({"REACTOR_THREADPOOL_MAXSIZE": 20}, 20), ({"AWS_MAX_POOL_CONNECTIONS": 30}, 30), ({"AWS_MAX_POOL_CONNECTIONS": 30, "REACTOR_THREADPOOL_MAXSIZE": 20}, 30), ], ) def test_max_pool_connections( self, monkeypatch: pytest.MonkeyPatch, settings: dict[str, Any], expected: int ) -> None: # restores the value that FilesPipeline.from_crawler() sets on the class monkeypatch.setattr(S3FilesStore, "AWS_MAX_POOL_CONNECTIONS", None) crawler = get_crawler( settings_dict={"FILES_STORE": "s3://mybucket/prefix/", **settings} ) store = FilesPipeline.from_crawler(crawler).store assert isinstance(store, S3FilesStore) config: Any = store.s3_client.meta.config assert config.max_pool_connections == expected class TestGCSFilesStore: @staticmethod def build_gcs_files_store( *, permissions: tuple[str, ...] = ( "storage.objects.get", "storage.objects.create", ), ) -> tuple[GCSFilesStore, Any, Any]: """Build a :class:`GCSFilesStore` mock. Returns ``(store, bucket_mock, blob_mock)``. Skips the test if google-cloud-storage is not installed. ``permissions`` is what ``Bucket.test_iam_permissions`` will return. """ pytest.importorskip("google.cloud.storage") client_mock, bucket_mock, blob_mock = mock_google_cloud_storage() bucket_mock.test_iam_permissions.return_value = list(permissions) with mock.patch("google.cloud.storage.Client", return_value=client_mock): store = GCSFilesStore("gs://my_bucket/my_prefix/") return store, bucket_mock, blob_mock def test_init(self, caplog: pytest.LogCaptureFixture) -> None: with caplog.at_level(logging.WARNING): self.build_gcs_files_store() assert not caplog.records def test_get_perm_missing(self, caplog: pytest.LogCaptureFixture) -> None: self.build_gcs_files_store(permissions=("storage.objects.create",)) assert ( "No 'storage.objects.get' permission for GCS bucket my_bucket" in caplog.text ) def test_create_perm_missing(self, caplog: pytest.LogCaptureFixture) -> None: self.build_gcs_files_store(permissions=("storage.objects.get",)) assert ( "No 'storage.objects.create' permission for GCS bucket my_bucket" in caplog.text ) def test_update_stores(self, monkeypatch: pytest.MonkeyPatch) -> None: monkeypatch.setattr(GCSFilesStore, "GCS_PROJECT_ID", None) monkeypatch.setattr(GCSFilesStore, "POLICY", None) settings = Settings( {"GCS_PROJECT_ID": "my-project", "FILES_STORE_GCS_ACL": "publicRead"} ) FilesPipeline._update_stores(settings) assert GCSFilesStore.GCS_PROJECT_ID == "my-project" assert GCSFilesStore.POLICY == "publicRead" # An empty FILES_STORE_GCS_ACL is normalised to None. settings = Settings({"GCS_PROJECT_ID": "my-project", "FILES_STORE_GCS_ACL": ""}) FilesPipeline._update_stores(settings) assert GCSFilesStore.POLICY is None @coroutine_test async def test_persist(self) -> None: store, bucket, blob = self.build_gcs_files_store() await maybe_deferred_to_future( store.persist_file( "full/filename", BytesIO(b"hello"), info=DUMMY_SPIDER_INFO, meta={"foo": 1}, headers={"Content-Type": "image/png"}, ) ) bucket.blob.assert_called_once_with("my_prefix/full/filename") assert blob.cache_control == GCSFilesStore.CACHE_CONTROL assert blob.metadata == {"foo": "1"} blob.upload_from_string.assert_called_once_with( data=b"hello", content_type="image/png", predefined_acl=store.POLICY, ) @coroutine_test async def test_persist_defaults(self) -> None: store, _, blob = self.build_gcs_files_store() await maybe_deferred_to_future( store.persist_file( "full/filename", BytesIO(b"hello"), info=DUMMY_SPIDER_INFO, ) ) blob.upload_from_string.assert_called_once_with( data=b"hello", content_type="application/octet-stream", predefined_acl=store.POLICY, ) assert blob.metadata == {} @coroutine_test async def test_stat(self) -> None: store, bucket, blob = self.build_gcs_files_store() checksum = "cdcda85605e46d0af6110752770dce3c" blob.md5_hash = base64.b64encode(bytes.fromhex(checksum)).decode() updated = datetime(2019, 12, 1) blob.updated = updated bucket.get_blob.return_value = blob stat = await maybe_deferred_to_future( store.stat_file("full/filename", info=DUMMY_SPIDER_INFO) ) bucket.get_blob.assert_called_once_with("my_prefix/full/filename") assert stat == { "checksum": checksum, "last_modified": time.mktime(updated.timetuple()), } @coroutine_test async def test_stat_missing_blob(self) -> None: store, bucket, _ = self.build_gcs_files_store() bucket.get_blob.return_value = None stat = await maybe_deferred_to_future( store.stat_file("full/filename", info=DUMMY_SPIDER_INFO) ) assert stat == {} @coroutine_test async def test_blob_path_consistency(self) -> None: """Test to make sure that paths used to store files is the same as the one used to get already uploaded files. """ store, bucket, _ = self.build_gcs_files_store() bucket.get_blob.return_value = None path = "full/my_data.txt" await maybe_deferred_to_future( store.persist_file(path, BytesIO(b""), info=DUMMY_SPIDER_INFO) ) await maybe_deferred_to_future(store.stat_file(path, info=DUMMY_SPIDER_INFO)) expected_blob_path = store.prefix + path bucket.blob.assert_called_with(expected_blob_path) bucket.get_blob.assert_called_with(expected_blob_path) class TestFTPFileStore: @inline_callbacks_test def test_persist(self): data = b"TestFTPFilesStore: \xe2\x98\x83" buf = BytesIO(data) meta = {"foo": "bar"} path = "full/filename" with MockFTPServer() as ftp_server: # normally set via FilesPipeline.from_crawler() FTPFilesStore.FTP_USERNAME = "anonymous" FTPFilesStore.FTP_PASSWORD = "guest" store = FTPFilesStore(ftp_server.url("/")) empty_dict = yield store.stat_file(path, info=DUMMY_SPIDER_INFO) assert empty_dict == {} yield store.persist_file( path, buf, info=DUMMY_SPIDER_INFO, meta=meta, headers=None ) stat = yield store.stat_file(path, info=DUMMY_SPIDER_INFO) assert "last_modified" in stat assert "checksum" in stat assert stat["checksum"] == "d113d66b2ec7258724a268bd88eef6b6" path = f"{store.basedir}/{path}" content = get_ftp_content_and_delete( path, store.host, store.port, store.username, store.password, bool(store.USE_ACTIVE_MODE), ) assert data == content @inline_callbacks_test def test_persist_active_mode(self, monkeypatch: pytest.MonkeyPatch): data = b"active mode" path = "full/filename" monkeypatch.setattr(FTPFilesStore, "FTP_USERNAME", "anonymous") monkeypatch.setattr(FTPFilesStore, "FTP_PASSWORD", "guest") monkeypatch.setattr(FTPFilesStore, "USE_ACTIVE_MODE", True) with MockFTPServer() as ftp_server: store = FTPFilesStore(ftp_server.url("/")) yield store.persist_file(path, BytesIO(data), info=DUMMY_SPIDER_INFO) stat = yield store.stat_file(path, info=DUMMY_SPIDER_INFO) assert stat["checksum"] == "ff1575649a39a27c13faa0d37c84bab3" def test_wrong_uri_scheme(self): with pytest.raises( ValueError, match=re.escape( "Incorrect URI scheme in http://example.com/, expected 'ftp'" ), ): FTPFilesStore("http://example.com/") class ItemWithFiles(Item): file_urls = Field() files = Field() def _create_item_with_files(*files: str) -> ItemWithFiles: item = ItemWithFiles() item["file_urls"] = files return item def _prepare_request_object(item_url: str, flags: list[str] | None = None) -> Request: return Request( item_url, meta={"response": Response(item_url, status=200, body=b"data", flags=flags)}, ) # this is separate from the one in test_pipeline_media.py to specifically test FilesPipeline subclasses class TestBuildFromCrawler: def setup_method(self): self.tempdir = mkdtemp() self.crawler = get_crawler(None, {"FILES_STORE": self.tempdir}) def teardown_method(self): rmtree(self.tempdir) def test_simple(self): class Pipeline(FilesPipeline): pass pipe = Pipeline.from_crawler(self.crawler) assert pipe.crawler == self.crawler assert pipe._fingerprinter assert pipe.store def test_has_from_crawler_and_init(self): class Pipeline(FilesPipeline): _from_crawler_called = False @classmethod def from_crawler(cls, crawler: Crawler) -> "Pipeline": settings = crawler.settings store_uri = settings["FILES_STORE"] o = cls(store_uri, crawler=crawler) o._from_crawler_called = True return o pipe = Pipeline.from_crawler(self.crawler) assert pipe.crawler == self.crawler assert pipe._fingerprinter assert pipe.store assert pipe._from_crawler_called @pytest.mark.parametrize("store", [None, ""]) def test_files_pipeline_raises_notconfigured_when_files_store_invalid(store): settings = Settings() settings.clear() settings.set("FILES_STORE", store, priority="cmdline") crawler = get_crawler(settings_dict=dict(settings)) with pytest.raises(NotConfigured): FilesPipeline.from_crawler(crawler)