140930 - Auto-detect Russian detect a meta charset windows 1257 page as KIO8-R

Status: VERIFIED DUPLICATE of bug 140234

(Core :: Internationalization, defect)

Description

[Saulius Gurklys]

From Bugzilla Helper:
User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.0.0+)
Gecko/20020428
BuildID:    2002042808

Page's (http://www.vtex.lt/sagu/eb.php) info window shows wrong page encoding,
no matter <meta http-equiv="Content-Type" content="text/html;
charset=Windows-1257"> is used.

Reproducible: Always
Steps to Reproduce:
1.goto http://www.vtex.lt/sagu/eb.php
2.Ctrl+I 
 or
 View->Page Info
 or
 Right-mouse -> View Page info
 

Actual Results:  Page info window shows "Encoding:" KOI-8R ,while meta
"Content-Type" shows Windows-1257.

Expected Results:  Encoding should be Windows-1257

Comment 1

[Christian :Biesinger (don't email me, ping me on IRC)]

*** Bug 140932 has been marked as a duplicate of this bug. ***

Comment 2

[Christopher Hoess (gone)]

->Page Info (maybe il8n)

Comment 3

[Christopher Hoess (gone)]

WFM, a Linux CVS trunk build from today.

Comment 4

[Boris Zbarsky [:bzbarsky]]

Um...  Page info shows KOI-8.  So does the charset submenu in the view menu...
So while we're showing the page as Windows-1257 (as can be verified by manually
changing it to KOI-8 via the View menu), the document _thinks_ it's in KOI-8.
Sounds like the content sink is not propagating charset info back correctly or
something... over to intl (this happens on Linux too, by the way).

Comment 5

[Yuying Long]

I can not reproduce this - the page info. for: http://www.vtex.lt/sagu/eb.php is
showing windows 1257 on 04-29 branch build.

Comment 6

[Saulius Gurklys]

The problem persists with build 0430, though... 

Comment 7

[Saulius Gurklys]

Attachment: Picture shows page info in 0430 build

Comment 8

[Frank Tang]

cannot reproduce with 2002041814 windod trunk build on en winme

Comment 9

[Saulius Gurklys]

I think, it is something related with a form on the page. 
There are 2 more pages:

http://www.vtex.lt/sagu/ebwf.php - Page without form, everything OK

and

http://www.vtex.lt/sagu/ebcf.php - Page with commented form(see source),
encoding is wrong

Comment 10

[Yuying Long]

Both of them display as windows 1257 in page info. for me.

Comment 11

[Saulius Gurklys]

I've found the "problem". I had character autodetection(View->Character
Coding->Auto-Detect) set as Russian. When I set autodetection - OFF, everythings
work just fine. Question - is it right behaviour? I think no matter
what autodetection is set, encoding specified on the page must be used !

Comment 12

[Roy Yokoyama]

>is it right behaviour? I think no matter
>what autodetection is set, encoding specified on the page must be used !
Your answer is correct. Auto Detection is a fallback when it's on.
Specified encoding in the page takes precedence. 

cc shanjian

Comment 13

[Yuying Long]

So, this is a auto-detect Russian problem, not a Page info. problem - charset is
marked as koi8-r in both charset menu and page info.

Change the summary and assign to Shanjian, and QA to myself.

Comment 14

[Shanjian Li]

This is a dup of 140234. I applied that patch and it fixed the problem. 

*** This bug has been marked as a duplicate of 140234 ***

Comment 15

[Yuying Long]

Mark as verified, please re-open if still see it after fix for bug 140234
checked-in.