mirrors/ArchiveBox
1
0
Fork 0
mirror of https://github.com/ArchiveBox/ArchiveBox.git synced 2025-05-14 07:04:27 -04:00
Code Issues Projects Releases Packages Wiki Activity Actions

feat: Add warc to list and limit check to succeeded archive results

Browse source
This commit is contained in:
Cristian 2020-11-05 07:54:40 -05:00
parent 33182fd53c
commit 71655220ad
1 changed files with 6 additions and 2 deletions
Download patch file Download diff file Expand all files Collapse all files

8
archivebox/core/utils.py
View file

@ -20,17 +20,21 @@ def get_icons(snapshot: Snapshot) -> str:
"archive_org": "🏛", "archive_org": "🏛",
"readability": "🆁", "readability": "🆁",
"mercury": "🅼", "mercury": "🅼",
"warc": "📦"
} }
exclude = ["favicon"] exclude = ["favicon"]
# Missing specific entry for WARC # Missing specific entry for WARC
for extractor in EXTRACTORS: for extractor in EXTRACTORS:
result = archive_results.filter(extractor=extractor[0]) result = archive_results.filter(extractor=extractor[0], status="succeeded")
try: try:
if extractor[0] not in exclude: if extractor[0] not in exclude:
output += output_template.format(link.archive_path, canon[f"{extractor[0]}_path"], output += output_template.format(link.archive_path, canon[f"{extractor[0]}_path"],
result.exists(), extractor[0], icons.get(extractor[0], "?")) result.exists(), extractor[0], icons.get(extractor[0], "?"))
if extractor[0] == "wget":
extractor = "warc"
output += output_template.format(link.archive_path, canon[f"{extractor}_path"],
result.exists(), extractor, icons.get(extractor, "?"))
except Exception as e: except Exception as e:
print(e) print(e)