mirrors/ArchiveBox
1
0
Fork 0
mirror of https://github.com/ArchiveBox/ArchiveBox.git synced 2025-05-19 09:25:11 -04:00
Code Issues Projects Releases Packages Wiki Activity Actions

Merge branch 'dev' into plugins-browsertrix

Browse source
This commit is contained in:
Nick Sweeting 2024-01-19 03:43:52 -08:00
commit a6b241f733
2 changed files with 11 additions and 2 deletions
Download patch file Download diff file Expand all files Collapse all files

5
archivebox/extractors/wget.py
View file

@ -202,4 +202,9 @@ def wget_output_path(link: Link) -> Optional[str]:
if search_dir.is_dir(): if search_dir.is_dir():
return domain(link.url).replace(":", "+") return domain(link.url).replace(":", "+")
# fallback to just the domain dir without port
search_dir = Path(link.link_dir) / domain(link.url).split(":", 1)[0]
if search_dir.is_dir():
return domain(link.url).split(":", 1)[0]
return None return None

8
archivebox/index/schema.py
View file

@ -379,11 +379,15 @@ class Link:
output_paths = ( output_paths = (
domain(self.url), domain(self.url),
'output.html',
'output.pdf', 'output.pdf',
'screenshot.png', 'screenshot.png',
'output.html', 'singlefile.html',
'readability/content.html',
'mercury/content.html',
'htmltotext.txt',
'media', 'media',
'singlefile.html' 'git',
) )
return any( return any(