34243 - 8bit chars in header stop display of msg

Status: RESOLVED WORKSFORME

(MailNews Core :: MIME, defect, P3)

Description

[Ben Bucksch (:BenB)]

Similar to the recent View Source problem: If an 8bit char (e.g. "ü") appears in
the header, nothing is displayed from there on, i.e. the body isn't rendered.

I'll attach a test msg.

Comment 1

[Ben Bucksch (:BenB)]

Attachment: Test msg

Comment 2

[Ben Bucksch (:BenB)]

To make clear: This bug appears for normal display.

Yes, I know, 8bit chars shouldn't appear in the header (I already contacted the
responsible person for the generator), but we shouldn't fall over them either
:).

Comment 3

[Ben Bucksch (:BenB)]

Also, I don't see the old header section anymore, but the normal headers look
like the ones in forwarded msgs now.

Comment 4

[rhp (gone)]

Looks like the header broadcaster/listeners are having problem with 8 bit 
data...hmm...weird. 

Any ideas Scott?

- rhp

Comment 5

[Scott MacGregor]

Rich, did you actually make sure the broadcaster was getting called with a
character with 8-bit data (i.e. that the failure didn't happen earlier?). The
broadcaster I wrote uses unicode strings which are multibyte strings so in
theory they don't care if it's an 8-bit character....


Ben, you are using imap right? We did have a bug a while ago where 8 bit data in
the headers was throwing the imap parser out of whack. This could be related?

Comment 6

[Ben Bucksch (:BenB)]

mscott,
View Source works correctly. The problem equally appears, if the folder resides
in Local Folders.

Comment 7

[Scott MacGregor]

ahh thanks for the extra info ben. so it must be in mime then.

Comment 8

[rhp (gone)]

Don't worry Scott. I'll debug it when I am doing debugging work...but now, I'm 
just pushing these out to work on features. Just wondered if anything rang a 
bell.

- rhp

Comment 9

[Ben Bucksch (:BenB)]

In Netscape 6 Beta1 (which still has the separate header section), only the
subject is truncated (from the 8bit char on), but the rest of the msg displays
fine.

Comment 10

[rhp (gone)]

This seems to render fine for me. View source as well.

- rhp

Comment 11

[Ben Bucksch (:BenB)]

mee to