157605 - Auto-detection doesn't work with certain case

Status: RESOLVED INVALID

(Core :: Internationalization, defect)

Description

[Yuying Long]

Build: 07-15 1.0.1 branch build with all platforms.

Steps:
1. Create a new profile and launch browser.
2. Load page(This page can be loaded by click the link in home.netscape.com/zh/cn):
http://www.zdnet.com.cn/news/internet/story/0,2000079768,39035221,00.htm
- this with gb2312 meta charset page load fine.  
3. Click on those comments links in the buttom of the page.

Result:
Load display garbled page (no meta charset), set charset auto-detection to
Universal, Chinese, SimpChinese still get garbled page.

Auto-dection works in some other cases, e.g. directly load a non meta-charset
page www.f10.com.cn...etc.

Comment 1

[Yuying Long]

Reassign to Shanjian.

Comment 2

[Shanjian Li]

This website hard chop gb2312 string for titles. Some gb2312 characters was cut
2 piece, and only the first part available on the page. Those partial character
failed in code scheme verification. It is not a problem of our charset detector. 

Comment 3

[Frank Tang]

bobj:    *  Send me any sites which do not work without charset menu (e.g.,
ZDNet China feedback)
ftang:  I don't believe this is the true case. If you set your default encoding
to gb2312 from the pref, and set the auto detector to universial. And then open
a new window to view those page. I can always see the in chinese
ylong: It is true.  There are some problems in their web site, please see
details in bug: http://bugzilla.mozilla.org/show_bug.cgi?id=157605

however, this bug happen when you DOES NOT set the default charset in pref to
GB2312, if you set it and open a new window, then it should work.