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

fix archive.org content-location capitalization

Browse source
This commit is contained in:
Nick Sweeting 2018-04-17 17:16:41 -04:00
parent b926b1affc
commit b3c93c43b2
1 changed files with 3 additions and 3 deletions
Download patch file Download diff file Expand all files Collapse all files

6
archive_methods.py
View file

@ -310,11 +310,11 @@ def archive_dot_org(link_dir, link, timeout=TIMEOUT):
# Parse archive.org response headers # Parse archive.org response headers
headers = result.stdout.splitlines() headers = result.stdout.splitlines()
content_location = [h for h in headers if b'Content-Location: ' in h] content_location = [h for h in headers if b'content-location: ' in h]
errors = [h for h in headers if h and b'X-Archive-Wayback-Runtime-Error: ' in h] errors = [h for h in headers if h and b'X-Archive-Wayback-Runtime-Error: ' in h]
if content_location: if content_location:
archive_path = content_location[0].split(b'Content-Location: ', 1)[-1].decode('utf-8') archive_path = content_location[0].split(b'content-location: ', 1)[-1].decode('utf-8')
saved_url = 'https://web.archive.org{}'.format(archive_path) saved_url = 'https://web.archive.org{}'.format(archive_path)
success = True success = True
@ -324,7 +324,7 @@ def archive_dot_org(link_dir, link, timeout=TIMEOUT):
elif errors: elif errors:
raise Exception(', '.join(e.decode() for e in errors)) raise Exception(', '.join(e.decode() for e in errors))
else: else:
raise Exception('Failed to find "Content-Location" URL header in Archive.org response.') raise Exception('Failed to find "content-location" URL header in Archive.org response.')
except Exception as e: except Exception as e:
end() end()
print(' Visit url to see output:', ' '.join(CMD)) print(' Visit url to see output:', ' '.join(CMD))