Opened on 02/27/2015 at 06:52:09 PM
Closed on 03/02/2015 at 04:26:21 PM
#2062 closed defect (fixed)
Trailing "?" is ignored when matching URLs
| Reported by: | mapx | Assignee: | sebastian |
|---|---|---|---|
| Priority: | P1 | Milestone: | Adblock-Plus-1.8.12-for-Chrome-Opera-Safari |
| Module: | Platform | Keywords: | |
| Cc: | sebastian | Blocked By: | |
| Blocking: | Platform: | Chrome | |
| Ready: | yes | Confidential: | no |
| Tester: | Verified working: | ||
| Review URL(s): | |||
Description (last modified by mapx)
Environment
chrome 41
ABP 1.8.11.1364 or ABP 1.8.11 (stable)
How to reproduce
go to
http://thedailyshow.cc.com/full-episodes/d9lx83/february-24--2015---lynsey-addario
or
http://www.spike.com/full-episodes/1hx3mv/framework-chainsaw-nightmare-season-1-ep-108
When the video reaches the first set of ads (it's marked on the seeking bar) you'll see "your content will return shortly" but ...will stay there forever.
I tested in FF 36, last dev build ==> it's playing fine
What I noted:
-the filter
@@||v.fwmrm.net/ad/p/1?*$domain=..…
is correctly read in firefox, but not in chrome
I can see in the chrome console:
POST http://140cc.v.fwmrm.net/ad/p/1? net::ERR_BLOCKED_BY_CLIENT
Anyway, I tested using an old version (1.8.8) and no issues with that version.
Attachments (0)
Change History (6)
comment:2 Changed on 02/27/2015 at 06:58:14 PM by mapx
- Cc sebastian added
comment:4 Changed on 02/27/2015 at 07:26:44 PM by sebastian
- Owner set to sebastian
- Priority changed from Unknown to P1
- Ready set
- Summary changed from ABP 1.8.11 doesn't apply correctly some filters to Trailing "?" is ignored when matching URLs
comment:5 Changed on 02/27/2015 at 07:38:23 PM by sebastian
- Review URL(s) modified (diff)
- Status changed from new to reviewing
comment:6 Changed on 03/02/2015 at 04:26:21 PM by sebastian
- Milestone set to Adblock-Plus-for-Chrome-Opera-Safari-next
- Resolution set to fixed
- Status changed from reviewing to closed
This regression where introduced by #1346 and is visible in Adblock Plus 1.8.11 as well as in the current devbuild. This is a bug in the code putting URL objects into strings, caused by relying on URL.search which returns an empty string if the URL has an empty query string like in http://example.com?.