278346 - Page source does not display properly if page is Big5 or Cyrillic encoded

Status: RESOLVED FIXED

(Camino Graveyard :: Page Layout, defect)

Description

[Yu-Hung Lien]

Well, I found Camino seemed failed to show page source if the line contained
both chinese characters and HTML tags. For example, page source of the site
http://www.oikos.com.tw/modules/news/ displayed inproperly if HTML tags and
chinese characters coexist in the same line. On the other hand, other portions
which containd chinese characters are just right.

Comment 1

[Smokey Ardisson (offline for a while; not following bugs - do not email)]

Does the charset have to be Big5, too?  

Compare the source of http://www.google.com.tw/ which is UTF-8 to the others,
which I see are Big-5.  Google.tw's source displays cleanly for me even with
Chinese and HTML on the same line.

Comment 2

[Yu-Hung Lien]

(In reply to comment #1)

Yes, I suppose you are right. I have checked several big-5 encoding sites and
their page sources all inproperly displayed.

Comment 3

[Smokey Ardisson (offline for a while; not following bugs - do not email)]

I'm working my way through sites with various encodings and have found three
more problematic encodings, all Cyrillic: Windows-1251
(http://www.macintosh.ru), Mac-Cyrillic (http://www.apple.ru), and koi8-r
(http://goa.cet.middlebury.edu/Smitheram/charset/koi8-r.html)

Shift-JIS and ISO-2022-JP, Win-1256 (Arab), euc-kr, gb2312 (Simp. Chinese),
win-1255 (Hebr), windows-1253 (Greek), tis-620 (Thai), windows-1250 and
iso-8859-2 (CE) all seem ok.  There are other charsets to test but I haven't
found pages written in them yet.

Changing summary to reflect currently-known problematic encodings and
confirming.  Big5 has the worst overlap by far of the four; Firefox 1.0 displays
the sources properly in all four cases.  Cm 2005011008 (v0.8+)

Comment 4

[Smokey Ardisson (offline for a while; not following bugs - do not email)]

Attachment: Lo-res screenshot to illustrate overlap

Comment 5

[Mike Pinkerton (not reading bugmail)]

is anything spit out on the console? just curious.

Comment 6

[Yu-Hung Lien]

(In reply to comment #5)
> is anything spit out on the console? just curious.

console? where should I check?

Comment 7

[Smokey Ardisson (offline for a while; not following bugs - do not email)]

/Applications/Utilities/Console

To answer Mike, yes: 
Jan 19 01:36:03 Krak-des-Chevaliers
/Applications/Internet/Camino.app/Contents/MacOS/Camino: *** Warning:
ATSUMeasureText has been deprecated.  Use ATSUGetUnjustifiedBounds instead. ***

(provided that message hasn't been sent already in a Camino session)

Specifically, 
http://tw.login.yahoo.com/cgi-bin/login.cgi?srv=mail&from=http://edit.tpe.yahoo.com/config/mail?.intl=tw
http://www.macintosh.ru/
http://goa.cet.middlebury.edu/Smitheram/charset/koi8-r.html
cause that message when the page itself is loading.

The others
http://www.oikos.com.tw/modules/news/
http://www.apple.ru/
cause the message to be sent only when view-source: begins loading.

(The error also shows up in System.log, too, also once per app session!)

Comment 8

[Smokey Ardisson (offline for a while; not following bugs - do not email)]

I see that Console.log message has its own bug, bug 161332.  Whether fixing that
bug will fix the problem with view source or whether it's just a coincidence
message I couldn't even begin to guess. :-)

Comment 9

[Yu-Hung Lien]

This bug seems to be cured for big5 encoding in 2005020508 build. The cyrillic
encoding is not solved yet.

Comment 10

[Smokey Ardisson (offline for a while; not following bugs - do not email)]

I still see the problem on big5 pages, until I reset my fonts as you mentioned
in  bug 278255 comment 8.  Then big5 source display does appear fixed. 

Comment 11

[Simon Fraser [no longer active]]

Fixed by charset stuff, and/or ATSUI fallback fix? Any remaining issues are
probably bug 175651.