mirrors/ArchiveBox
1
0
Fork 0
mirror of https://github.com/ArchiveBox/ArchiveBox.git synced 2025-05-16 16:14:28 -04:00
Code Issues Projects Releases Packages Wiki Activity Actions

show progress during validate_links

Browse source
This commit is contained in:
Nick Sweeting 2019-05-01 02:28:26 -04:00
parent f60b5ed867
commit 1ac99621ab
1 changed files with 7 additions and 12 deletions
Download patch file Download diff file Expand all files Collapse all files

19
archivebox/index/__init__.py
View file

@ -122,18 +122,13 @@ def merge_links(a: Link, b: Link) -> Link:
@enforce_types @enforce_types
def validate_links(links: Iterable[Link]) -> List[Link]: def validate_links(links: Iterable[Link]) -> List[Link]:
links = archivable_links(links) # remove chrome://, about:, mailto: etc. timer = TimedProgress(TIMEOUT * 4)
links = sorted_links(links) # deterministically sort the links based on timstamp, url try:
links = uniquefied_links(links) # merge/dedupe duplicate timestamps & urls links = archivable_links(links) # remove chrome://, about:, mailto: etc.
links = sorted_links(links) # deterministically sort the links based on timstamp, url
if not links: links = uniquefied_links(links) # merge/dedupe duplicate timestamps & urls
stderr('{red}[X] No links found in index.{reset}'.format(**ANSI)) finally:
stderr(' To add a link to your archive, run:') timer.end()
stderr(" archivebox add 'https://example.com'")
stderr()
stderr(' For more usage and examples, run:')
stderr(' archivebox help')
raise SystemExit(1)
return list(links) return list(links)