def retrieve_response(self, spider, request):
response = super(S3CacheStorage, self).retrieve_response(spider, request)
if response is None: # not in local filesystem cache, so try copying from s3
local_path = self._get_request_path(spider, request)
remote_path = os.path.relpath(local_path, self.tmpcachedir).lower()
bucket = self.conn.get_bucket(self.bucket_name, validate=False)
def _get_key(filename):
key_name = os.path.join(remote_path, filename)
return bucket.get_key(key_name)
# check if the key exists
metadata_key = _get_key('pickled_meta')
if metadata_key is None:
return None # key not found
# check if the cache entry has expired
mtime = parse_ts(metadata_key.last_modified)
if 0 < self.expiration_secs < (datetime.datetime.utcnow() - mtime).total_seconds():
return None # expired
# deserialise the cached response
metadata = pickle.loads(metadata_key.get_contents_as_string())
body = _get_key('response_body').get_contents_as_string()
rawheaders = _get_key('response_headers').get_contents_as_string()
url = metadata.get('response_url')
status = metadata['status']
headers = Headers(headers_raw_to_dict(rawheaders))
respcls = responsetypes.from_args(headers=headers, url=url)
response = respcls(url=url, headers=headers, status=status, body=body)
return response
def retrieve_response(self, spider, request):
"""Return response if present in cache, or None otherwise."""
key = self._get_request_key(spider, request)
expiration_time = self._get_expiration_time(spider)
metadata = self._read_meta(key, expiration_time)
if metadata is None:
return # not cached
spider.log("%s: found cache for %s" %
(self.__class__.__name__, request.url))
res = self._get_cached_data(key, spider)
if res is None:
return None
spider.log("%s: got response from cache for %s" %
(self.__class__.__name__, request.url))
response_url = res['response_url']
response_body = str(res['response_body'])
response_rawheaders = res['response_meta']['headers']
status = res['response_meta']['status']
response_headers = Headers(headers_raw_to_dict(response_rawheaders))
response_cls = responsetypes.from_args(headers=response_headers,
url=response_url)
response = response_cls(url=response_url,
headers=response_headers,
status=status,
body=response_body)
return response
def test_headers_raw_to_dict(self):
raw = b"Content-type: text/html\n\rAccept: gzip\n\r\
Cache-Control: no-cache\n\rCache-Control: no-store\n\n"
dct = {
b'Content-type': [b'text/html'],
b'Accept': [b'gzip'],
b'Cache-Control': [b'no-cache', b'no-store']
}
self.assertEqual(headers_raw_to_dict(raw), dct)
def test_headers_raw_to_dict(self):
raw = b"Content-type: text/html\n\rAccept: gzip\n\r\
Cache-Control: no-cache\n\rCache-Control: no-store\n\n"
dct = {
b"Content-type": [b"text/html"],
b"Accept": [b"gzip"],
b"Cache-Control": [b"no-cache", b"no-store"],
}
self.assertEqual(headers_raw_to_dict(raw), dct)
def retrieve_response(self, spider, request):
"""Return response if present in cache, or None otherwise."""
metadata = self._read_meta(spider, request)
if metadata is None:
return # not cached
rpath = self._get_request_path(spider, request)
with open(join(rpath, 'response_body'), 'rb') as f:
body = f.read()
with open(join(rpath, 'response_headers'), 'rb') as f:
rawheaders = f.read()
url = metadata.get('response_url')
status = metadata['status']
headers = Headers(headers_raw_to_dict(rawheaders))
respcls = responsetypes.from_args(headers=headers, url=url)
response = respcls(url=url, headers=headers, status=status, body=body)
return response
def retrieve_response(self, spider, request):
"""Return response if present in cache, or None otherwise."""
try:
doc = json.loads(self.bucket.get(self._inverse_url(request.url))[2])
except couchbase.exception.MemcachedError:
return
# @TODO expiration
body = doc['response_body']
url = doc['response_url']
status = doc['status']
headers = Headers(headers_raw_to_dict(doc['response_headers']))
encoding = doc['encoding']
respcls = responsetypes.from_args(headers=headers, url=url)
response = respcls(url=url, headers=headers, status=status, body=body,
encoding=encoding)
return response
def retrieve_response(self, spider, request):
"""Return response if present in cache, or None otherwise."""
metadata = self._read_meta(spider, request)
if metadata is None:
return # not cached
rpath = self._get_request_path(spider, request)
with self._open(os.path.join(rpath, 'response_body'), 'rb') as f:
body = f.read()
with self._open(os.path.join(rpath, 'response_headers'), 'rb') as f:
rawheaders = f.read()
url = metadata.get('response_url')
status = metadata['status']
headers = Headers(headers_raw_to_dict(rawheaders))
respcls = responsetypes.from_args(headers=headers, url=url)
response = respcls(url=url, headers=headers, status=status, body=body)
return response
def retrieve_response(rpath):
"""
Return response if present in cache, or None otherwise
"""
metadata = read_meta(rpath)
if metadata is None:
return # not cached
with open(join(rpath, 'response_body'), 'rb') as f:
body = f.read()
with open(join(rpath, 'response_headers'), 'rb') as f:
rawheaders = f.read()
url = metadata.get('response_url')
status = metadata['status']
headers = Headers(headers_raw_to_dict(rawheaders))
respcls = responsetypes.from_args(headers=headers, url=url)
response = respcls(url=url, headers=headers, status=status, body=body)
return response
def retrieve_response(self, spider, request):
"""Return response if present in cache, or None otherwise."""
metadata = self._read_meta(spider, request)
if metadata is None:
return # not cached
rpath = self._get_request_path(spider, request)
with open(join(rpath, 'response_body'), 'rb') as f:
body = f.read()
with open(join(rpath, 'response_headers'), 'rb') as f:
rawheaders = f.read()
# We failback to metadata['url'] to support old generated caches. TODO: remove for Scrapy 0.11
url = metadata.get('response_url') or metadata['url']
status = metadata['status']
headers = Headers(headers_raw_to_dict(rawheaders))
respcls = responsetypes.from_args(headers=headers, url=url)
response = respcls(url=url, headers=headers, status=status, body=body)
return response
def retrieve_response(self, spider, request):
"""Return response if present in cache, or None otherwise."""
metadata = self._read_meta(spider, request)
if metadata is None:
return # not cached
rpath = self._get_request_path(spider, request)
with open(join(rpath, 'response_body'), 'rb') as f:
body = f.read()
with open(join(rpath, 'response_headers'), 'rb') as f:
rawheaders = f.read()
# We failback to metadata['url'] to support old generated caches. TODO: remove for Scrapy 0.11
url = metadata.get('response_url') or metadata['url']
status = metadata['status']
headers = Headers(headers_raw_to_dict(rawheaders))
respcls = responsetypes.from_args(headers=headers, url=url)
response = respcls(url=url, headers=headers, status=status, body=body)
return response
def retrieve_response(self, spider: TSpider,
request: TRequest) -> Optional[TResponse]:
"""Return response if present in cache, or None otherwise."""
metadata = self._read_meta(spider, request)
if metadata is None:
return # not cached
rpath = self._get_request_path(spider, request)
with self._open(os.path.join(rpath, "response_body"), "rb") as f:
body = f.read()
with self._open(os.path.join(rpath, "response_headers"), "rb") as f:
rawheaders = f.read()
url = metadata.get("response_url")
status = metadata["status"]
headers = Headers(headers_raw_to_dict(rawheaders))
respcls = responsetypes.from_args(headers=headers, url=url)
response = respcls(url=url, headers=headers, status=status, body=body)
return response
def retrieve_response(self, spider, request):
"""Return response if present in cache, or None otherwise."""
if self.dont_retrieve:
return
keyname = self._get_request_path(request)
keydata = self.get_object_from_key(self.bucket_name, keyname)
if not keydata:
return # not cached
keydata = pickle.loads(keydata)
metadata = keydata['meta']
body = keydata['response_body']
rawheaders = keydata['response_headers']
url = metadata.get('response_url')
status = metadata['status']
headers = Headers(headers_raw_to_dict(rawheaders))
respcls = responsetypes.from_args(headers=headers, url=url)
response = respcls(url=url, headers=headers, status=status, body=body)
return response
def retrieve_response(self, spider, request):
key = request_fingerprint(request)
value = self.collection.find_one({"_id": key})
if not value:
return
stored_data = value["value"]
metadata = stored_data['metadata']
url = str(metadata['response_url'])
rawheaders = stored_data['response_headers']
rawheaders = headers_raw_to_dict(rawheaders)
if "Content-Encoding" in rawheaders:
del rawheaders["Content-Encoding"]
body = stored_data["response_body"]
body = body.encode("utf-8", "w3lib_replace")
status = str(metadata['status'])
headers = Headers(rawheaders)
respcls = responsetypes.from_args(headers=headers, url=url)
response = respcls(url=url, headers=headers, status=status, body=body)
return response
def retrieve_response(self, spider, request):
"""Return response if present in cache, or None otherwise."""
try:
doc = json.loads(
self.bucket.get(self._inverse_url(request.url))[2])
except couchbase.exception.MemcachedError:
return
# @TODO expiration
body = doc['response_body']
url = doc['response_url']
status = doc['status']
headers = Headers(headers_raw_to_dict(doc['response_headers']))
encoding = doc['encoding']
respcls = responsetypes.from_args(headers=headers, url=url)
response = respcls(url=url,
headers=headers,
status=status,
body=body,
encoding=encoding)
return response
def retrieve_response(self, spider, request):
"""Return response if present in cache, or None otherwise."""
meta_data_key = self.key_for(spider, request, "meta_data")
raw_meta_data = self.client.get(meta_data_key)
meta_data = json.loads(raw_meta_data) if raw_meta_data else None
if not meta_data:
return
body, raw_headers = (self.client.get(
self.key_for(spider, request, type_))
for type_ in ("response_body",
"response_headers"))
if not all((body, raw_headers)):
return
kwargs = {
"url": meta_data.get("response_url"),
"headers": Headers(headers_raw_to_dict(raw_headers)),
}
response_class = responsetypes.from_args(**kwargs)
kwargs.update({"status": meta_data.get("status"), "body": body})
return response_class(**kwargs)
def retrieve_response(self, spider, request):
path = functools.partial(storage_path, request)
try:
_metadata = get_s3_text(self.bucket, path('pickled_meta'))
body = get_s3_text(self.bucket, path('response_body'))
rawheaders = get_s3_text(self.bucket, path('response_headers'))
except botocore.exceptions.ClientError as e:
if e.response['Error']['Code'] == "404":
return None
else:
raise
metadata = pickle.loads(_metadata)
url = metadata.get('response_url')
status = metadata['status']
headers = Headers(headers_raw_to_dict(rawheaders))
respcls = responsetypes.from_args(headers=headers, url=url)
response = respcls(url=url, headers=headers, status=status, body=body)
if response.status == 302:
self.bucket.delete_objects(Delete=make_delete_objects(path))
return None
return response
def retrieve_response(self, spider, request):
response = super(S3CacheStorage,
self).retrieve_response(spider, request)
if response is None: # not in local filesystem cache, so try copying from s3
local_path = self._get_request_path(spider, request)
remote_path = os.path.relpath(local_path, self.tmpcachedir).lower()
bucket = self.conn.get_bucket(self.bucket_name, validate=False)
def _get_key(filename):
key_name = os.path.join(remote_path, filename)
return bucket.get_key(key_name)
# check if the key exists
metadata_key = _get_key('pickled_meta')
if metadata_key is None:
return None # key not found
# check if the cache entry has expired
mtime = parse_ts(metadata_key.last_modified)
if 0 < self.expiration_secs < (datetime.datetime.utcnow() -
mtime).total_seconds():
return None # expired
# deserialise the cached response
metadata = pickle.loads(metadata_key.get_contents_as_string())
body = _get_key('response_body').get_contents_as_string()
rawheaders = _get_key('response_headers').get_contents_as_string()
url = metadata.get('response_url')
status = metadata['status']
headers = Headers(headers_raw_to_dict(rawheaders))
respcls = responsetypes.from_args(headers=headers, url=url)
response = respcls(url=url,
headers=headers,
status=status,
body=body)
return response
def test_headers_raw_dict_none(self):
self.assertIsNone(headers_raw_to_dict(None))
self.assertIsNone(headers_dict_to_raw(None))
def test_headers_raw_to_dict(self):
raw = b"Content-type: text/html\n\rAccept: gzip\n\n"
dct = {b'Content-type': [b'text/html'], b'Accept': [b'gzip']}
self.assertEqual(headers_raw_to_dict(raw), dct)
def test_headers_raw_dict_none(self):
self.assertIsNone(headers_raw_to_dict(None))
self.assertIsNone(headers_dict_to_raw(None))
def test_headers_raw_to_dict(self):
raw = b"Content-type: text/html\n\rAccept: gzip\n\n"
dct = {b'Content-type': [b'text/html'], b'Accept': [b'gzip']}
self.assertEqual(headers_raw_to_dict(raw), dct)
def test_headers_raw_to_dict(self):
raw = b"Content-type: text/html\n\rAccept: gzip\n\r\
Cache-Control: no-cache\n\rCache-Control: no-store\n\n"
dct = {b'Content-type': [b'text/html'], b'Accept': [b'gzip'],
b'Cache-Control': [b'no-cache', b'no-store']}
self.assertEqual(headers_raw_to_dict(raw), dct)
