def cache_data(
self,
extracted_metadata: dict,
cache_manager: CacheManager,
allow_list: dict,
):
for feature, meta_data in extracted_metadata.items():
if (allow_list[feature]
and Explanation.Cached not in meta_data[EXPLANATION]):
values = []
if feature == ACCESSIBILITY:
values = meta_data[VALUES]
data_to_be_cached = {
VALUES: values,
STAR_CASE: meta_data[STAR_CASE],
TIMESTAMP: get_utc_now(),
EXPLANATION: meta_data[EXPLANATION],
}
create_cache_entry(
cache_manager.get_domain(),
feature,
data_to_be_cached,
self._logger,
)
def print_exceptions(maximum_age_in_seconds: int):
database: Session = ProfilerSession()
query = database.query(
db_models.Record.exception,
db_models.Record.timestamp,
db_models.Record.url,
)
meta_rows = database.execute(query)
failure_urls = []
print_data = {}
for meta_row in meta_rows:
timestamp = meta_row[1]
if timestamp < (get_utc_now() - maximum_age_in_seconds):
continue
exception = meta_row[0]
if exception == "":
continue
if exception not in print_data.keys():
print_data.update({exception: 0})
if "Empty html. Potentially, splash failed." in exception:
print(f"Splash failed for url: '{meta_row[2]}'")
failure_urls.append(meta_row[2])
print_data[exception] += 1
print(
f"----------------- Found exceptions of the last {round(maximum_age_in_seconds / SECONDS_PER_DAY, 2)} days."
)
print(f"All urls which caused exceptions: {get_unique_list(failure_urls)}")
print(f"Total number of found exceptions: {len(print_data.items())}")
for exception, value in print_data.items():
if exception != "":
print(exception, value)
print("-------")
def _processing_values(self, values: list[str], website_data: WebsiteData,
before: float) -> dict:
website_data.values = values
star_case, explanation = self._decide(website_data=website_data)
data = {
self.key: {
TIME_REQUIRED: get_utc_now() - before,
VALUES: values,
STAR_CASE: star_case,
EXPLANATION: explanation,
}
}
if self.tag_list_last_modified != "":
data[self.key].update({
"tag_list_last_modified": self.tag_list_last_modified,
"tag_list_expires": self.tag_list_expires,
})
return data
def extract_meta(input_data: Input):
starting_extraction = get_utc_now()
allowance = _convert_allow_list_to_dict(input_data.allow_list)
database_exception = ""
try:
create_request_record(
starting_extraction, input_data=input_data, allowance=allowance
)
except OperationalError as err:
database_exception += (
"\nDatabase exception: "
+ str(err.args)
+ "".join(traceback.format_exception(None, err, err.__traceback__))
)
uuid = app.communicator.send_message(
{
MESSAGE_URL: input_data.url,
MESSAGE_HTML: input_data.html,
MESSAGE_HEADERS: input_data.headers,
MESSAGE_HAR: input_data.har,
MESSAGE_ALLOW_LIST: allowance,
MESSAGE_SHARED_MEMORY_NAME: shared_status.shm.name,
MESSAGE_BYPASS_CACHE: input_data.bypass_cache,
}
)
meta_data: dict = app.communicator.get_message(uuid)
if meta_data:
extractor_tags = _convert_dict_to_output_model(
meta_data, input_data.debug
)
if MESSAGE_EXCEPTION in meta_data.keys():
exception = meta_data[MESSAGE_EXCEPTION]
else:
exception = ""
else:
extractor_tags = None
exception = f"No response from {METADATA_EXTRACTOR}."
end_time = get_utc_now()
out = Output(
url=input_data.url,
meta=extractor_tags,
exception=exception + database_exception,
time_until_complete=end_time - starting_extraction,
)
try:
create_response_record(
starting_extraction,
end_time,
input_data=input_data,
allowance=allowance,
output=out,
)
except OperationalError as err:
database_exception += (
"\nDatabase exception: "
+ str(err.args)
+ "".join(traceback.format_exception(None, err, err.__traceback__))
)
out.exception += database_exception
if exception != "":
raise HTTPException(
status_code=400,
detail={
MESSAGE_URL: input_data.url,
"meta": meta_data,
MESSAGE_EXCEPTION: exception,
"time_until_complete": end_time - starting_extraction,
},
)
return out
def is_cached_value_recent(timestamp: float) -> bool:
return timestamp >= (get_utc_now() -
CACHE_RETENTION_TIME_DAYS * SECONDS_PER_DAY)
def rester():
allow_list = {
"advertisement": True,
"easy_privacy": True,
"malicious_extensions": True,
"extracted_links": True,
"extract_from_files": True,
"fanboy_annoyance": True,
"fanboy_notification": True,
"fanboy_social_media": True,
"anti_adblock": True,
"easylist_germany": True,
"easylist_adult": True,
"paywall": True,
"security": True,
"iframe_embeddable": True,
"pop_up": True,
"reg_wall": True,
"log_in_out": True,
"accessibility": True,
"cookies": True,
"g_d_p_r": True,
"javascript": True,
}
extractor_url = "http://0.0.0.0:5057/extract_meta"
result = {}
try:
os.remove(RESULT_FILE_PATH)
except FileNotFoundError:
pass
logs, file_path = load_file_list()
for counter, raw in enumerate(load_scraped_data(logs, file_path)):
before = time.perf_counter()
print(f"Working file {counter + 1} of {len(logs)}".center(80, "-"))
print(raw["url"])
starting_extraction = get_utc_now()
headers = {"Content-Type": "application/json"}
payload = {
MESSAGE_HTML: raw["html"],
MESSAGE_HEADERS: raw["headers"],
MESSAGE_URL: raw["url"],
MESSAGE_ALLOW_LIST: allow_list,
MESSAGE_HAR: raw["har"],
"debug": True,
}
response = requests.request("POST",
extractor_url,
headers=headers,
data=json.dumps(payload))
try:
output = json.loads(response.content)
except JSONDecodeError as e:
print(response.content)
print(f"Exception: {e}, {e.args}")
output = {}
output.update(
{"time_for_extraction": get_utc_now() - starting_extraction})
result.update({raw["url"]: output})
with open(RESULT_FILE_PATH, "w") as fp:
json.dump(result, fp)
print(output)
after = time.perf_counter()
print(f"Total time needed in series: {after - before}")
def _prepare_start(self) -> tuple[float, WebsiteData]:
self._logger.info(f"Starting {self.__class__.__name__}.")
before = get_utc_now()
website_data = self._prepare_website_data()
return before, website_data
def start(self, message: dict) -> dict:
self._logger.debug(
f"Start metadata_manager at {time.perf_counter() - global_start} since start"
)
shared_status = shared_memory.ShareableList(
name=message[MESSAGE_SHARED_MEMORY_NAME])
url = message[MESSAGE_URL]
if len(url) > 1024:
url = url[0:1024]
shared_status[1] = url
website_manager = WebsiteManager.get_instance()
self._logger.debug(
f"WebsiteManager initialized at {time.perf_counter() - global_start} since start"
)
website_manager.load_website_data(message=message)
self._logger.debug(
f"WebsiteManager loaded at {time.perf_counter() - global_start} since start"
)
cache_manager = CacheManager.get_instance()
cache_manager.update_to_current_domain(
website_manager.website_data.domain,
bypass=message[MESSAGE_BYPASS_CACHE],
)
now = time.perf_counter()
self._logger.debug(
f"starting_extraction at {now - global_start} since start")
starting_extraction = get_utc_now()
if website_manager.website_data.html == "":
exception = "Empty html. Potentially, splash failed."
extracted_meta_data = {MESSAGE_EXCEPTION: exception}
else:
try:
extracted_meta_data = asyncio.run(
self._extract_meta_data(
allow_list=message[MESSAGE_ALLOW_LIST],
cache_manager=cache_manager,
shared_memory_name=message[MESSAGE_SHARED_MEMORY_NAME],
))
self.cache_data(
extracted_meta_data,
cache_manager,
allow_list=message[MESSAGE_ALLOW_LIST],
)
except ConnectionError as e:
exception = f"Connection error extracting metadata: '{e.args}'"
self._logger.exception(
exception,
exc_info=True,
)
extracted_meta_data = {MESSAGE_EXCEPTION: exception}
except Exception as e:
exception = (
f"Unknown exception from extracting metadata: '{e.args}'. "
f"{''.join(traceback.format_exception(None, e, e.__traceback__))}"
)
self._logger.exception(
exception,
exc_info=True,
)
extracted_meta_data = {MESSAGE_EXCEPTION: exception}
self._logger.debug(
f"extracted_meta_data at {time.perf_counter() - global_start} since start"
)
extracted_meta_data.update({
"time_for_extraction": get_utc_now() - starting_extraction,
**website_manager.get_website_data_to_log(),
})
website_manager.reset()
cache_manager.reset()
shared_status[1] = ""
self._logger.debug(
f"website_manager.reset() at {time.perf_counter() - global_start} since start"
)
return extracted_meta_data
