siphyo / adblock-plus-japanese-filter

Automatically exported from code.google.com/p/adblock-plus-japanese-filter
0 stars 0 forks source link

False positive: bbc.co.uk #11

Closed GoogleCodeExporter closed 8 years ago

GoogleCodeExporter commented 8 years ago
What steps will reproduce the problem?
1. Subscribe to your filter list in Firefox 3.6.*
2. Go to http://news.bbc.co.uk/2/hi/world/asia_pacific/10156834.stm

What is the expected output? What do you see instead?

Expected output is the site itself but it is hidden by one of your filters.

Please provide any additional information below.

Hi, I am one of of the maintainers of EasyList/EasyPrivacy filter subscription. 
[1] Recently a user reported a problem with your filter list on bbc.co.uk. The 
original bug report can 
be found in the official Adblock Plus forum. [2]. The filter 
##*[class*="sponsor"] breaks the site by hiding the legitimate content (i.e. 
the whole site).

Your sincerely
Erunno

[1] http://easylist.adblockplus.org/
[2] https://adblockplus.org/forum/viewtopic.php?f=1&t=5589&start=0

Original issue reported on code.google.com by erunno.o...@googlemail.com on 27 May 2010 at 7:59

GoogleCodeExporter commented 8 years ago
Thanks erunno for reporting. 

I read the original post on your forum, and checked the BBC page with our 
filter ON.
But the page is still readable. Why?

##*[class*="sponsor"] hides the DIV element with id 
"bbccom_storyprintsponsorship".
This element is located after the article body and too small to make the page 
blank.

I don't think our filter causes the problem.

Best regards,
k2japan

Original comment by k2ja...@gmail.com on 27 May 2010 at 7:48

GoogleCodeExporter commented 8 years ago
Hi,

the <body> element contains "bbccom_slot_storyprintsponsorship" in its class 
declaration. ##*[class*="sponsor"] 
triggers because of the "sponsor" substring and therefore hides the whole side.

Best regards,
Erunno

Original comment by erunno.o...@googlemail.com on 27 May 2010 at 8:27

GoogleCodeExporter commented 8 years ago
Hi, Erunno

I was a little bit confused. I thought ##*[id*="sponsor"] was wrong.
It's not ID but CLASS, which cause false positive. I misunderstood that.

Anyway, I tried to reproduce the problem again because one of our contributers
pointed out that the problem happens with JavaScript enabled.

The problem is reproduced with JavaScript. The page dynamically adds some class
attributes, which include "bbccom_slot_storyprintsponsorship".

Now, ##*[class*="sponsor"] has exception as follows to avoid the false positive.
~bbc.co.uk##*[class*="sponsor"]

Thanks again for reporting, Erunno

Best regards,
k2japan

Original comment by k2ja...@gmail.com on 28 May 2010 at 3:11