Closed
Bug 175345
Opened 23 years ago
Closed 23 years ago
Auto-detect Universal doesn't work well with JavaScript openwindow
Categories
(Core :: Internationalization, defect)
Core
Internationalization
Tracking
()
VERIFIED
WORKSFORME
People
(Reporter: amyy, Assigned: ftang)
References
()
Details
(Keywords: intl, topembed+, Whiteboard: [adt2] [ETA Needed])
Attachments
(2 files)
Build: 10-16 branch and trunk build on all platforms
Steps to reproduce:
1. Load page http://newchat.china.com - this is a page with GB2312 meta-tag charset.
2. View | Character Coding | Auto-detect, set auto-detection to Universal.
3. Enter any chat room by clicking on any link in this page.
4. There should be open a pop-up window by JavaScript, and the frame in the top
of page doesn't has meta-charset.
Result:
There are some text links in the top of page are displayed garbled.
Note:
This problem is only happens on auto-detect Universal. Auto-detect Chinese,
Simplified Chinese and East Asian will show those links properly.
I think to resolve this problem, we should either improve the universal
auto-detector or carry over the parent page charset.
| Reporter | ||
Comment 1•23 years ago
|
||
There is an associated page aa1.html will be followed.
| Reporter | ||
Comment 2•23 years ago
|
||
Please save this page and the page above locally in same directory.
| Reporter | ||
Comment 3•23 years ago
|
||
nominate as nsbeta1 since china.com is one of top 10 Chinese site, their chat
room has a large population.
Comment 4•23 years ago
|
||
If you take a look at the frame source, there is only 2 chinese characters
there. That's simple not enough for auto-detector to reach a reasonable result.
All other characters are from java script in different file. Unless we process
javascript file, there is almost no way to fix this problem.
Status: NEW → ASSIGNED
| Reporter | ||
Comment 5•23 years ago
|
||
Other auto-detectors like Chinese will work.
Also, although there is a little difference with the test case, I think there
are enough characters there, but still has the same problem with universal
detector which is not good. I guess we would see similar cases in some other
web sites with pop-up window by javascript.
Comment 6•23 years ago
|
||
Marking as topembed+ because this will be needed to support a mjor embedding
customer.
Keywords: topembed+
Whiteboard: [adt2] [ETA Needed]
Comment 7•23 years ago
|
||
I do not see any problem with the attached testcase.
My guess about why chinese detector works is because chinese detector could not
figure out what charset it is, so it reports nothing. Frame uses parent charset
this way. For universal detector, the 2 character in page "songti" can be
interprete as more "frequently" (not in the sense of webpage) characters in euc-tw.
We probably shouldn't give any result if data is too small.
| Reporter | ||
Comment 8•23 years ago
|
||
in comment #7:
> I do not see any problem with the attached testcase.
The attached test case display garbled for me.
If there is really hard for universal detector detects small data, then can we
carry over the parent page charset into w/o meta-tag charset sub-frame?
Comment 9•23 years ago
|
||
That's what I am suggesting, don't report charset if data is too small.
Comment 10•23 years ago
|
||
In comment comment #8
> can we carry over the parent page charset into w/o meta-tag charset sub-frame?
But in this case, it is opening a new window. It is not a sub-frame. For any
new unlabeled window, we use the default charset. What is your default charset?
Reassiging to ftang.
Assignee: yokoyama → ftang
Status: ASSIGNED → NEW
| Reporter | ||
Comment 11•23 years ago
|
||
> > can we carry over the parent page charset into w/o meta-tag charset sub-frame?
> But in this case, it is opening a new window. It is not a sub-frame. > For
any new unlabeled window, we use the default charset. What is
> your default charset?
Even though the page is pop-up opened by the link in http://newchat.china.com,
but the bug is about the sub-frame without charset meta-tag at the top of the
page, while the pop-up page itself does has the charset meta-tag.
Btw, the page has been changed since the bug reported, the top frame part has
more data right now, auto-detect universal is working currently.
| Assignee | ||
Comment 12•23 years ago
|
||
I cannot reproduce this problem with the 20030114 window trunk build. Mark it
WFM. If you can still reproduce that, please attach an image and indicate the
problem area when you reopen. Thanks
Status: NEW → RESOLVED
Closed: 23 years ago
Resolution: --- → WORKSFORME
| Reporter | ||
Comment 13•23 years ago
|
||
Mark as verified as works for me - since the web site contents changed,
auto-detect Universal works right now.
Status: RESOLVED → VERIFIED
You need to log in
before you can comment on or make changes to this bug.
Description
•