mirror of
https://github.com/scrapy/scrapy.git
synced 2025-02-23 22:03:41 +00:00
We're storing exceptions captured by Twisted on the media pipeline cache, but we're also using the defer.returnValue method with our own methods decorated with @defer.inlineCallbacks. The defer.returnValue method passes returned values forward by throwing a defer._DefGen_Return exception, which in its turn extends the BaseException class and is captured by Twisted. This way, the latest exception stored in the Failure's object may also have an HtmlResponse object in its __context__ attribute. As the Response object also keeps track of the Request object that has originated it, you could figure it out how many RAM we're wasting here. This could easily lead to a Memory Leak problem when running spiders with Media Pipeline enabled and a particular Request set that tends to raise a significant number of exceptions. Example triggers: - media requests with 404 status responses - user land exceptins coming from custom middlewares - etc.
393 lines
16 KiB
Python
393 lines
16 KiB
Python
from __future__ import print_function
|
|
|
|
import sys
|
|
|
|
from testfixtures import LogCapture
|
|
from twisted.trial import unittest
|
|
from twisted.python.failure import Failure
|
|
from twisted.internet import reactor
|
|
from twisted.internet.defer import Deferred, inlineCallbacks, returnValue
|
|
|
|
from scrapy.http import Request, Response
|
|
from scrapy.settings import Settings
|
|
from scrapy.spiders import Spider
|
|
from scrapy.utils.request import request_fingerprint
|
|
from scrapy.pipelines.media import MediaPipeline
|
|
from scrapy.pipelines.files import FileException
|
|
from scrapy.utils.log import failure_to_exc_info
|
|
from scrapy.utils.signal import disconnect_all
|
|
from scrapy import signals
|
|
|
|
|
|
def _mocked_download_func(request, info):
|
|
response = request.meta.get('response')
|
|
return response() if callable(response) else response
|
|
|
|
|
|
class BaseMediaPipelineTestCase(unittest.TestCase):
|
|
|
|
pipeline_class = MediaPipeline
|
|
settings = None
|
|
|
|
def setUp(self):
|
|
self.spider = Spider('media.com')
|
|
self.pipe = self.pipeline_class(download_func=_mocked_download_func,
|
|
settings=Settings(self.settings))
|
|
self.pipe.open_spider(self.spider)
|
|
self.info = self.pipe.spiderinfo
|
|
|
|
def tearDown(self):
|
|
for name, signal in vars(signals).items():
|
|
if not name.startswith('_'):
|
|
disconnect_all(signal)
|
|
|
|
def test_default_media_to_download(self):
|
|
request = Request('http://url')
|
|
assert self.pipe.media_to_download(request, self.info) is None
|
|
|
|
def test_default_get_media_requests(self):
|
|
item = dict(name='name')
|
|
assert self.pipe.get_media_requests(item, self.info) is None
|
|
|
|
def test_default_media_downloaded(self):
|
|
request = Request('http://url')
|
|
response = Response('http://url', body=b'')
|
|
assert self.pipe.media_downloaded(response, request, self.info) is response
|
|
|
|
def test_default_media_failed(self):
|
|
request = Request('http://url')
|
|
fail = Failure(Exception())
|
|
assert self.pipe.media_failed(fail, request, self.info) is fail
|
|
|
|
def test_default_item_completed(self):
|
|
item = dict(name='name')
|
|
assert self.pipe.item_completed([], item, self.info) is item
|
|
|
|
# Check that failures are logged by default
|
|
fail = Failure(Exception())
|
|
results = [(True, 1), (False, fail)]
|
|
|
|
with LogCapture() as l:
|
|
new_item = self.pipe.item_completed(results, item, self.info)
|
|
|
|
assert new_item is item
|
|
assert len(l.records) == 1
|
|
record = l.records[0]
|
|
assert record.levelname == 'ERROR'
|
|
self.assertTupleEqual(record.exc_info, failure_to_exc_info(fail))
|
|
|
|
# disable failure logging and check again
|
|
self.pipe.LOG_FAILED_RESULTS = False
|
|
with LogCapture() as l:
|
|
new_item = self.pipe.item_completed(results, item, self.info)
|
|
assert new_item is item
|
|
assert len(l.records) == 0
|
|
|
|
@inlineCallbacks
|
|
def test_default_process_item(self):
|
|
item = dict(name='name')
|
|
new_item = yield self.pipe.process_item(item, self.spider)
|
|
assert new_item is item
|
|
|
|
def test_modify_media_request(self):
|
|
request = Request('http://url')
|
|
self.pipe._modify_media_request(request)
|
|
assert request.meta == {'handle_httpstatus_all': True}
|
|
|
|
def test_should_remove_req_res_references_before_caching_the_results(self):
|
|
"""Regression test case to prevent a memory leak in the Media Pipeline.
|
|
|
|
The memory leak is triggered when an exception is raised when a Response
|
|
scheduled by the Media Pipeline is being returned. For example, when a
|
|
FileException('download-error') is raised because the Response status
|
|
code is not 200 OK.
|
|
|
|
It happens because we are keeping a reference to the Response object
|
|
inside the FileException context. This is caused by the way Twisted
|
|
return values from inline callbacks. It raises a custom exception
|
|
encapsulating the original return value.
|
|
|
|
The solution is to remove the exception context when this context is a
|
|
_DefGen_Return instance, the BaseException used by Twisted to pass the
|
|
returned value from those inline callbacks.
|
|
|
|
Maybe there's a better and more reliable way to test the case described
|
|
here, but it would be more complicated and involve running - or at least
|
|
mocking - some async steps from the Media Pipeline. The current test
|
|
case is simple and detects the problem very fast. On the other hand, it
|
|
would not detect another kind of leak happening due to old object
|
|
references being kept inside the Media Pipeline cache.
|
|
|
|
This problem does not occur in Python 2.7 since we don't have Exception
|
|
Chaining (https://www.python.org/dev/peps/pep-3134/).
|
|
"""
|
|
# Create sample pair of Request and Response objects
|
|
request = Request('http://url')
|
|
response = Response('http://url', body=b'', request=request)
|
|
|
|
# Simulate the Media Pipeline behavior to produce a Twisted Failure
|
|
try:
|
|
# Simulate a Twisted inline callback returning a Response
|
|
# The returnValue method raises an exception encapsulating the value
|
|
returnValue(response)
|
|
except BaseException as exc:
|
|
def_gen_return_exc = exc
|
|
try:
|
|
# Simulate the media_downloaded callback raising a FileException
|
|
# This usually happens when the status code is not 200 OK
|
|
raise FileException('download-error')
|
|
except Exception as exc:
|
|
file_exc = exc
|
|
# Simulate Twisted capturing the FileException
|
|
# It encapsulates the exception inside a Twisted Failure
|
|
failure = Failure(file_exc)
|
|
|
|
# The Failure should encapsulate a FileException ...
|
|
self.assertEqual(failure.value, file_exc)
|
|
# ... and if we're running on Python 3 ...
|
|
if sys.version_info.major >= 3:
|
|
# ... it should have the returnValue exception set as its context
|
|
self.assertEqual(failure.value.__context__, def_gen_return_exc)
|
|
|
|
# Let's calculate the request fingerprint and fake some runtime data...
|
|
fp = request_fingerprint(request)
|
|
info = self.pipe.spiderinfo
|
|
info.downloading.add(fp)
|
|
info.waiting[fp] = []
|
|
|
|
# When calling the method that caches the Request's result ...
|
|
self.pipe._cache_result_and_execute_waiters(failure, fp, info)
|
|
# ... it should store the Twisted Failure ...
|
|
self.assertEqual(info.downloaded[fp], failure)
|
|
# ... encapsulating the original FileException ...
|
|
self.assertEqual(info.downloaded[fp].value, file_exc)
|
|
# ... but it should not store the returnValue exception on its context
|
|
context = getattr(info.downloaded[fp].value, '__context__', None)
|
|
self.assertIsNone(context)
|
|
|
|
|
|
class MockedMediaPipeline(MediaPipeline):
|
|
|
|
def __init__(self, *args, **kwargs):
|
|
super(MockedMediaPipeline, self).__init__(*args, **kwargs)
|
|
self._mockcalled = []
|
|
|
|
def download(self, request, info):
|
|
self._mockcalled.append('download')
|
|
return super(MockedMediaPipeline, self).download(request, info)
|
|
|
|
def media_to_download(self, request, info):
|
|
self._mockcalled.append('media_to_download')
|
|
if 'result' in request.meta:
|
|
return request.meta.get('result')
|
|
return super(MockedMediaPipeline, self).media_to_download(request, info)
|
|
|
|
def get_media_requests(self, item, info):
|
|
self._mockcalled.append('get_media_requests')
|
|
return item.get('requests')
|
|
|
|
def media_downloaded(self, response, request, info):
|
|
self._mockcalled.append('media_downloaded')
|
|
return super(MockedMediaPipeline, self).media_downloaded(response, request, info)
|
|
|
|
def media_failed(self, failure, request, info):
|
|
self._mockcalled.append('media_failed')
|
|
return super(MockedMediaPipeline, self).media_failed(failure, request, info)
|
|
|
|
def item_completed(self, results, item, info):
|
|
self._mockcalled.append('item_completed')
|
|
item = super(MockedMediaPipeline, self).item_completed(results, item, info)
|
|
item['results'] = results
|
|
return item
|
|
|
|
|
|
class MediaPipelineTestCase(BaseMediaPipelineTestCase):
|
|
|
|
pipeline_class = MockedMediaPipeline
|
|
|
|
@inlineCallbacks
|
|
def test_result_succeed(self):
|
|
cb = lambda _: self.pipe._mockcalled.append('request_callback') or _
|
|
eb = lambda _: self.pipe._mockcalled.append('request_errback') or _
|
|
rsp = Response('http://url1')
|
|
req = Request('http://url1', meta=dict(response=rsp), callback=cb, errback=eb)
|
|
item = dict(requests=req)
|
|
new_item = yield self.pipe.process_item(item, self.spider)
|
|
self.assertEqual(new_item['results'], [(True, rsp)])
|
|
self.assertEqual(self.pipe._mockcalled,
|
|
['get_media_requests', 'media_to_download',
|
|
'media_downloaded', 'request_callback', 'item_completed'])
|
|
|
|
@inlineCallbacks
|
|
def test_result_failure(self):
|
|
self.pipe.LOG_FAILED_RESULTS = False
|
|
cb = lambda _: self.pipe._mockcalled.append('request_callback') or _
|
|
eb = lambda _: self.pipe._mockcalled.append('request_errback') or _
|
|
fail = Failure(Exception())
|
|
req = Request('http://url1', meta=dict(response=fail), callback=cb, errback=eb)
|
|
item = dict(requests=req)
|
|
new_item = yield self.pipe.process_item(item, self.spider)
|
|
self.assertEqual(new_item['results'], [(False, fail)])
|
|
self.assertEqual(self.pipe._mockcalled,
|
|
['get_media_requests', 'media_to_download',
|
|
'media_failed', 'request_errback', 'item_completed'])
|
|
|
|
@inlineCallbacks
|
|
def test_mix_of_success_and_failure(self):
|
|
self.pipe.LOG_FAILED_RESULTS = False
|
|
rsp1 = Response('http://url1')
|
|
req1 = Request('http://url1', meta=dict(response=rsp1))
|
|
fail = Failure(Exception())
|
|
req2 = Request('http://url2', meta=dict(response=fail))
|
|
item = dict(requests=[req1, req2])
|
|
new_item = yield self.pipe.process_item(item, self.spider)
|
|
self.assertEqual(new_item['results'], [(True, rsp1), (False, fail)])
|
|
m = self.pipe._mockcalled
|
|
# only once
|
|
self.assertEqual(m[0], 'get_media_requests') # first hook called
|
|
self.assertEqual(m.count('get_media_requests'), 1)
|
|
self.assertEqual(m.count('item_completed'), 1)
|
|
self.assertEqual(m[-1], 'item_completed') # last hook called
|
|
# twice, one per request
|
|
self.assertEqual(m.count('media_to_download'), 2)
|
|
# one to handle success and other for failure
|
|
self.assertEqual(m.count('media_downloaded'), 1)
|
|
self.assertEqual(m.count('media_failed'), 1)
|
|
|
|
@inlineCallbacks
|
|
def test_get_media_requests(self):
|
|
# returns single Request (without callback)
|
|
req = Request('http://url')
|
|
item = dict(requests=req) # pass a single item
|
|
new_item = yield self.pipe.process_item(item, self.spider)
|
|
assert new_item is item
|
|
assert request_fingerprint(req) in self.info.downloaded
|
|
|
|
# returns iterable of Requests
|
|
req1 = Request('http://url1')
|
|
req2 = Request('http://url2')
|
|
item = dict(requests=iter([req1, req2]))
|
|
new_item = yield self.pipe.process_item(item, self.spider)
|
|
assert new_item is item
|
|
assert request_fingerprint(req1) in self.info.downloaded
|
|
assert request_fingerprint(req2) in self.info.downloaded
|
|
|
|
@inlineCallbacks
|
|
def test_results_are_cached_across_multiple_items(self):
|
|
rsp1 = Response('http://url1')
|
|
req1 = Request('http://url1', meta=dict(response=rsp1))
|
|
item = dict(requests=req1)
|
|
new_item = yield self.pipe.process_item(item, self.spider)
|
|
self.assertTrue(new_item is item)
|
|
self.assertEqual(new_item['results'], [(True, rsp1)])
|
|
|
|
# rsp2 is ignored, rsp1 must be in results because request fingerprints are the same
|
|
req2 = Request(req1.url, meta=dict(response=Response('http://donot.download.me')))
|
|
item = dict(requests=req2)
|
|
new_item = yield self.pipe.process_item(item, self.spider)
|
|
self.assertTrue(new_item is item)
|
|
self.assertEqual(request_fingerprint(req1), request_fingerprint(req2))
|
|
self.assertEqual(new_item['results'], [(True, rsp1)])
|
|
|
|
@inlineCallbacks
|
|
def test_results_are_cached_for_requests_of_single_item(self):
|
|
rsp1 = Response('http://url1')
|
|
req1 = Request('http://url1', meta=dict(response=rsp1))
|
|
req2 = Request(req1.url, meta=dict(response=Response('http://donot.download.me')))
|
|
item = dict(requests=[req1, req2])
|
|
new_item = yield self.pipe.process_item(item, self.spider)
|
|
self.assertTrue(new_item is item)
|
|
self.assertEqual(new_item['results'], [(True, rsp1), (True, rsp1)])
|
|
|
|
@inlineCallbacks
|
|
def test_wait_if_request_is_downloading(self):
|
|
def _check_downloading(response):
|
|
fp = request_fingerprint(req1)
|
|
self.assertTrue(fp in self.info.downloading)
|
|
self.assertTrue(fp in self.info.waiting)
|
|
self.assertTrue(fp not in self.info.downloaded)
|
|
self.assertEqual(len(self.info.waiting[fp]), 2)
|
|
return response
|
|
|
|
rsp1 = Response('http://url')
|
|
def rsp1_func():
|
|
dfd = Deferred().addCallback(_check_downloading)
|
|
reactor.callLater(.1, dfd.callback, rsp1)
|
|
return dfd
|
|
|
|
def rsp2_func():
|
|
self.fail('it must cache rsp1 result and must not try to redownload')
|
|
|
|
req1 = Request('http://url', meta=dict(response=rsp1_func))
|
|
req2 = Request(req1.url, meta=dict(response=rsp2_func))
|
|
item = dict(requests=[req1, req2])
|
|
new_item = yield self.pipe.process_item(item, self.spider)
|
|
self.assertEqual(new_item['results'], [(True, rsp1), (True, rsp1)])
|
|
|
|
@inlineCallbacks
|
|
def test_use_media_to_download_result(self):
|
|
req = Request('http://url', meta=dict(result='ITSME', response=self.fail))
|
|
item = dict(requests=req)
|
|
new_item = yield self.pipe.process_item(item, self.spider)
|
|
self.assertEqual(new_item['results'], [(True, 'ITSME')])
|
|
self.assertEqual(self.pipe._mockcalled, \
|
|
['get_media_requests', 'media_to_download', 'item_completed'])
|
|
|
|
|
|
class MediaPipelineAllowRedirectSettingsTestCase(unittest.TestCase):
|
|
|
|
def _assert_request_no3xx(self, pipeline_class, settings):
|
|
pipe = pipeline_class(settings=Settings(settings))
|
|
request = Request('http://url')
|
|
pipe._modify_media_request(request)
|
|
|
|
self.assertIn('handle_httpstatus_list', request.meta)
|
|
for status, check in [
|
|
(200, True),
|
|
|
|
# These are the status codes we want
|
|
# the downloader to handle itself
|
|
(301, False),
|
|
(302, False),
|
|
(302, False),
|
|
(307, False),
|
|
(308, False),
|
|
|
|
# we still want to get 4xx and 5xx
|
|
(400, True),
|
|
(404, True),
|
|
(500, True)]:
|
|
if check:
|
|
self.assertIn(status, request.meta['handle_httpstatus_list'])
|
|
else:
|
|
self.assertNotIn(status, request.meta['handle_httpstatus_list'])
|
|
|
|
def test_standard_setting(self):
|
|
self._assert_request_no3xx(
|
|
MediaPipeline,
|
|
{
|
|
'MEDIA_ALLOW_REDIRECTS': True
|
|
})
|
|
|
|
def test_subclass_standard_setting(self):
|
|
|
|
class UserDefinedPipeline(MediaPipeline):
|
|
pass
|
|
|
|
self._assert_request_no3xx(
|
|
UserDefinedPipeline,
|
|
{
|
|
'MEDIA_ALLOW_REDIRECTS': True
|
|
})
|
|
|
|
def test_subclass_specific_setting(self):
|
|
|
|
class UserDefinedPipeline(MediaPipeline):
|
|
pass
|
|
|
|
self._assert_request_no3xx(
|
|
UserDefinedPipeline,
|
|
{
|
|
'USERDEFINEDPIPELINE_MEDIA_ALLOW_REDIRECTS': True
|
|
})
|