backup/pywb
1
0
Fork 0
mirror of https://github.com/webrecorder/pywb.git synced 2025-03-24 06:59:52 +01:00
Code Issues Projects Releases Activity

replay: ignore 304 captures

Browse Source
This commit is contained in:
archiveit 2014-03-08 23:46:59 +00:00
parent 584d826f05
commit 4fdcdc98ae
1 changed files with 13 additions and 5 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

18
pywb/core/replay_views.py
View File

@ -178,12 +178,20 @@ class ReplayView:
Check if response is a 3xx redirect to the same url Check if response is a 3xx redirect to the same url
If so, reject this capture to avoid causing redirect loop If so, reject this capture to avoid causing redirect loop
""" """
if status_headers.statusline.startswith('3'): if not status_headers.statusline.startswith('3'):
request_url = wbrequest.wb_url.url.lower() return
location_url = status_headers.get_header('Location').lower()
if (UrlRewriter.strip_protocol(request_url) == UrlRewriter.strip_protocol(location_url)): request_url = wbrequest.wb_url.url.lower()
raise CaptureException('Self Redirect: ' + str(cdx)) location_url = status_headers.get_header('Location')
if not location_url:
if status_headers.statusline.startswith('304'):
raise CaptureException('Skipping 304 Modified: ' + str(cdx))
return
location_url = location_url.lower()
if (UrlRewriter.strip_protocol(request_url) == UrlRewriter.strip_protocol(location_url)):
raise CaptureException('Self Redirect: ' + str(cdx))
def _reject_referrer_self_redirect(self, wbrequest): def _reject_referrer_self_redirect(self, wbrequest):
""" """