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

warcserver: self-redirect improvement: include trailing slash in self-redirect check, urls differing only by trailing slash should be considered self-redirect, update tests

Browse Source
This commit is contained in:
Ilya Kreymer 2017-11-09 21:22:11 -08:00
parent da2ae0f373
commit 0c74616070
2 changed files with 8 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
pywb/warcserver/resource/responseloader.py
View File

@ -139,8 +139,8 @@ class BaseLoader(object):
host = urlsplit(cdx['url']).netloc host = urlsplit(cdx['url']).netloc
location_url = host + location_url location_url = host + location_url
location_url = location_url.split('://', 1)[-1] location_url = location_url.split('://', 1)[-1].rstrip('/')
request_url = request_url.split('://', 1)[-1] request_url = request_url.split('://', 1)[-1].rstrip('/')
if request_url == location_url: if request_url == location_url:
msg = 'Self Redirect {0} -> {1}' msg = 'Self Redirect {0} -> {1}'

6
tests/test_integration.py
View File

@ -350,6 +350,12 @@ class TestWbIntegration(BaseConfigTest):
assert resp.status_int == 200 assert resp.status_int == 200
assert resp.headers['Content-Location'].endswith('/pywb/20140126200928{0}/http://www.iana.org/domains/root/db'.format(fmod)) assert resp.headers['Content-Location'].endswith('/pywb/20140126200928{0}/http://www.iana.org/domains/root/db'.format(fmod))
def test_non_exact_replay_skip_self_redir_slash(self, fmod):
uri = '/pywb/20140126200927{0}/http://www.iana.org/domains/root/db/'
resp = self.get(uri, fmod)
assert resp.status_int == 200
assert resp.headers['Content-Location'].endswith('/pywb/20140126200928{0}/http://www.iana.org/domains/root/db'.format(fmod))
def test_not_existant_warc_other_capture(self, fmod): def test_not_existant_warc_other_capture(self, fmod):
resp = self.get('/pywb/20140703030321{0}/http://example.com/?example=2', fmod) resp = self.get('/pywb/20140703030321{0}/http://example.com/?example=2', fmod)
assert resp.status_int == 200 assert resp.status_int == 200