Mirrors/ArchiveBox
2
0
Fork 0
mirror of https://github.com/ArchiveBox/ArchiveBox synced 2024-11-22 20:23:12 +00:00
Code Issues Projects Releases Packages Wiki Activity

disable wget --mirror in favor of timestamping

Browse source
This commit is contained in:
Nick Sweeting 2017-07-04 03:21:47 -05:00
parent 0551e5b9e0
commit e09c704d50
1 changed files with 4 additions and 2 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

6
archive.py
View file

@ -168,7 +168,7 @@ def fetch_wget(out_dir, link, overwrite=False):
if not os.path.exists('{}/{}'.format(out_dir, domain)) or overwrite: if not os.path.exists('{}/{}'.format(out_dir, domain)) or overwrite:
print(' - Downloading Full Site') print(' - Downloading Full Site')
CMD = [ CMD = [
*'wget --mirror --adjust-extension --convert-links --no-parent'.split(' '), *'wget --timestamping --adjust-extension --convert-links --no-parent'.split(' '),
*(('--page-requisites',) if FETCH_WGET_IMAGES else ()), *(('--page-requisites',) if FETCH_WGET_IMAGES else ()),
link['url'], link['url'],
] ]
@ -225,8 +225,9 @@ def archive_dot_org(out_dir, link, overwrite=False):
submit_url = 'https://web.archive.org/save/{}'.format(link['url'].split('?', 1)[0]) submit_url = 'https://web.archive.org/save/{}'.format(link['url'].split('?', 1)[0])
success = False success = False
CMD = ['curl', '-I', submit_url]
try: try:
result = run(['curl', '-I', submit_url], stdout=PIPE, stderr=DEVNULL, cwd=out_dir, timeout=TIMEOUT) # archive.org result = run(CMD, stdout=PIPE, stderr=DEVNULL, cwd=out_dir, timeout=TIMEOUT) # archive.org
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]
if content_location: if content_location:
@ -234,6 +235,7 @@ def archive_dot_org(out_dir, link, overwrite=False):
saved_url = 'https://web.archive.org{}'.format(archive_path) saved_url = 'https://web.archive.org{}'.format(archive_path)
success = True success = True
else: else:
print(' Visit url to see output:', ' '.join(CMD))
raise Exception('Failed to find Content-Location URL in Archive.org response headers.') raise Exception('Failed to find Content-Location URL in Archive.org response headers.')
except Exception as e: except Exception as e:
print(' Exception: {} {}'.format(e.__class__.__name__, e)) print(' Exception: {} {}'.format(e.__class__.__name__, e))