338252 - sourceTextEncoding attribute isn't parsed correctly in all cases (certain Sherlock plugins don't display correctly)

Status: RESOLVED FIXED

(Firefox :: Search, defect, P1)

Description

[:Gavin Sharp [email: [e-mail]]]

If the sourceTextEncoding attribute in a Sherlock file has a space after the equals sign, it fails to parse and the file's encoding isn't recognized.

We should keep the byte array around until we start the actual parsing, and then re-convert later if we see a sourceTextEncoding attribute. That way we can avoid regex hacks and just use the normal Sherlock attribute parser to parse sourceTextEncoding.

Comment 1

[:Gavin Sharp [email: [e-mail]]]

Attachment: patch

I've tested it with the plugins from bug 335878 comment 14, and from bug 335878 comment 0, but more extensive testing is always appreciated. I want to double check a few things before asking for review.

Comment 2

[Hiro]

(In reply to comment #1)
> Created an attachment (id=222579) [edit]
> patch
> 
> I've tested it with the plugins from bug 335878 comment 14, and from bug 335878
> comment 0, but more extensive testing is always appreciated. I want to double
> check a few things before asking for review.
> 

I tried this patch.
The problem of bug 335878 comment 14 is solved, too. 
There seems to be no problem though other search plugin was tried. 
Please go ahead.

Comment 3

[:Gavin Sharp [email: [e-mail]]]

Comment on attachment 222579 [details] [diff] [review]
patch

Thanks again for the testing Hiro!

Comment 4

[Mike Beltzner [:beltzner, not reading bugmail]]

--> beta1

Comment 5

[Mike Connor [:mconnor]]

Comment on attachment 222579 [details] [diff] [review]
patch

} else

should be

}
else

otherwise, r+a=me

Comment 6

[:Gavin Sharp [email: [e-mail]]]

mozilla/browser/components/search/nsSearchService.js 	1.1.2.24
mozilla/browser/components/search/nsSearchService.js 	1.28