bellingcat/auto-archiver
1
0
Fork 0
mirror of https://github.com/bellingcat/auto-archiver.git synced 2026-06-13 05:38:29 +03:00
Code Issues Packages Projects Releases Wiki Activity

fix issue with two runs of anitbot extractor

Browse Source
This commit is contained in:
msramalho
2025-07-05 16:00:03 +01:00
parent 0d1447117c
commit eae0da08b3
1 changed files with 3 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

3
src/auto_archiver/modules/antibot_extractor_enricher/antibot_extractor_enricher.py
View File

@@ -81,6 +81,9 @@ class AntibotExtractorEnricher(Extractor, Enricher):
os.makedirs(self.user_data_dir, exist_ok=True) os.makedirs(self.user_data_dir, exist_ok=True)
def enrich(self, to_enrich: Metadata, custom_data_dir: bool = True) -> bool: def enrich(self, to_enrich: Metadata, custom_data_dir: bool = True) -> bool:
if to_enrich.get_media_by_id("html_source_code"):
logger.info("Antibot has already been executed, skipping.")
return True
using_user_data_dir = self.user_data_dir if custom_data_dir else None using_user_data_dir = self.user_data_dir if custom_data_dir else None
url = to_enrich.get_url() url = to_enrich.get_url()