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: change strip_scheme() to strip_scheme_www() to also strip away www. prefix for self-redirect checking, #73

Browse Source
This commit is contained in:
Ilya Kreymer 2015-02-22 22:51:35 -08:00
parent 83f8d7d29b
commit 5d80d2d891
1 changed files with 23 additions and 19 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

42
pywb/webapp/replay_views.py
View File

@ -34,7 +34,7 @@ class CaptureException(WbException):
#================================================================= #=================================================================
class ReplayView(object): class ReplayView(object):
STRIP_SCHEME = re.compile('^([\w]+:[/]*)?(.*?)$') STRIP_SCHEME_WWW = re.compile('^([\w]+:[/]*(?:www[\d]*\.)?)?(.*?)$')
def __init__(self, content_loader, config): def __init__(self, content_loader, config):
self.content_loader = content_loader self.content_loader = content_loader
@ -286,8 +286,8 @@ class ReplayView(object):
host = urlsplit(cdx['original']).netloc host = urlsplit(cdx['original']).netloc
location_url = host + location_url location_url = host + location_url
if (ReplayView.strip_scheme(request_url) == if (ReplayView.strip_scheme_www(request_url) ==
ReplayView.strip_scheme(location_url)): ReplayView.strip_scheme_www(location_url)):
raise CaptureException('Self Redirect: ' + str(cdx)) raise CaptureException('Self Redirect: ' + str(cdx))
# TODO: reevaluate this, as it may reject valid refreshes of a page # TODO: reevaluate this, as it may reject valid refreshes of a page
@ -307,39 +307,43 @@ class ReplayView(object):
request_url = (wbrequest.host_prefix + request_url = (wbrequest.host_prefix +
wbrequest.rel_prefix + str(wbrequest.wb_url)) wbrequest.rel_prefix + str(wbrequest.wb_url))
if (ReplayView.strip_scheme(request_url) == if (ReplayView.strip_scheme_www(request_url) ==
ReplayView.strip_scheme(wbrequest.referrer)): ReplayView.strip_scheme_www(wbrequest.referrer)):
raise CaptureException('Self Redirect via Referrer: ' + raise CaptureException('Self Redirect via Referrer: ' +
str(wbrequest.wb_url)) str(wbrequest.wb_url))
@staticmethod @staticmethod
def strip_scheme(url): def strip_scheme_www(url):
""" """
>>> ReplayView.strip_scheme('https://example.com') ==\ >>> ReplayView.strip_scheme_www('https://example.com') ==\
ReplayView.strip_scheme('http://example.com') ReplayView.strip_scheme_www('http://example.com')
True True
>>> ReplayView.strip_scheme('https://example.com') ==\ >>> ReplayView.strip_scheme_www('https://example.com') ==\
ReplayView.strip_scheme('http:/example.com') ReplayView.strip_scheme_www('http:/example.com')
True True
>>> ReplayView.strip_scheme('https://example.com') ==\ >>> ReplayView.strip_scheme_www('https://example.com') ==\
ReplayView.strip_scheme('example.com') ReplayView.strip_scheme_www('example.com')
True True
>>> ReplayView.strip_scheme('about://example.com') ==\ >>> ReplayView.strip_scheme_www('https://example.com') ==\
ReplayView.strip_scheme('example.com') ReplayView.strip_scheme_www('http://www2.example.com')
True True
>>> ReplayView.strip_scheme('http://') ==\ >>> ReplayView.strip_scheme_www('about://example.com') ==\
ReplayView.strip_scheme('') ReplayView.strip_scheme_www('example.com')
True True
>>> ReplayView.strip_scheme('#!@?') ==\ >>> ReplayView.strip_scheme_www('http://') ==\
ReplayView.strip_scheme('#!@?') ReplayView.strip_scheme_www('')
True
>>> ReplayView.strip_scheme_www('#!@?') ==\
ReplayView.strip_scheme_www('#!@?')
True True
""" """
m = ReplayView.STRIP_SCHEME.match(url) m = ReplayView.STRIP_SCHEME_WWW.match(url)
match = m.group(2) match = m.group(2)
return match return match