[Fixed] Parsing of filter|* seems broken
Posted: Sun Dec 09, 2012 3:35 am
See first reply for a fix
At http://livemanager.eurovision.edgesuite ... index.html I was trying to whitelist the resourceafter receiving a FP report from a user (didn't find the issue though). I tried adding the following rule:
However, it doesn't match. It works fine if I change it into any of
.
Since there is a * following the |, the | should be parsed as a literal |, and not as the end of an URL. Therefore I believe this is a bug.
https://reports.adblockplus.org/12928b4 ... f6c4e9c54b
sidenote: in a very short (and therefore unsuccessful, although the latter could also be because it's 3:30 AM) attempt to help find the issue, I noticed the following (filterclasses.js)\x80 (unicode 128) however isn't an ascii character anymore
(and a typo in charaters )
At http://livemanager.eurovision.edgesuite ... index.html I was trying to whitelist the resource
Code: Select all
http://adserver.adtech.de/adserv|3.0|1215|4025198|0|1908|ADTECH;grp=%5Bgroup%5D
Code: Select all
@@||adserver.adtech.de/adserv|*
Code: Select all
@@||adserver.adtech.de/adserv
@@||adserver.adtech.de/adserv|3
Since there is a * following the |, the | should be parsed as a literal |, and not as the end of an URL. Therefore I believe this is a bug.
https://reports.adblockplus.org/12928b4 ... f6c4e9c54b
sidenote: in a very short (and therefore unsuccessful, although the latter could also be because it's 3:30 AM) attempt to help find the issue, I noticed the following (filterclasses.js)
Code: Select all
// process separator placeholders (all ANSI charaters but alphanumeric characters and _%.-)
.replace(/\\\^/g, "(?:[\\x00-\\x24\\x26-\\x2C\\x2F\\x3A-\\x40\\x5B-\\x5E\\x60\\x7B-\\x80]|$)")
