288850 - For non-mail products, don't process 'mailedit' charsets

Status: RESOLVED DUPLICATE of bug 943252

(Core :: Internationalization, defect)

Description

[Jungshik Shin]

While testing my debug build of firefox, I got an assertion in nsCharsetMenu.cpp
because mailedit charset pref. is not available in non-mailnews products. The
part of the code to process mailedit charset needs to be excluded from the
compilation for non-mailnews products. 

Scott, are MOZ_MAIL_NEWS and MOZ_THUNDERBIRD the right macro to check mailnews/tb?

Comment 1

[Jungshik Shin]

Attachment: patch (not yet tested)

I realized that it's not so straightforward as I thought. nsCharsetMenu is an
inplementation of nsICurrentCharsetListener and the conditional compilation of
the part of the interface and its implementation means that iid and cid don't
uniquely identify the interface and the implementation. Giving different iid
and cid don't solve the problem either because the extension mechanism(?) would
probably break. I need some advice from xpcom gurus..

Comment 2

[Jungshik Shin]

copied from bug 323714 (report by Gavin Sharp)

Firefox's preferences dialog notifies "charsetmenu-selected" with the data
"other":
http://bonsai.mozilla.org/cvsblame.cgi?file=mozilla/browser/components/preferences/fonts.js&rev=1.5&mark=141#136
but firefox.js doesn't include the kMaileditPrefKey
("intl.charsetmenu.mailedit") pref, so this assertion in InitMailEditMenu is
hit opening the "Advanced" button in Tools->Options->Content->Fonts and Colors
for the first time after startup. I don't actually understand what the
InitOthers() and InitMailEditMenu() methods actually do.
http://bonsai.mozilla.org/cvsblame.cgi?file=mozilla/xpfe/components/intl/nsCharsetMenu.cpp&rev=1.40&mark=442,443#415

Should Firefox be sending "browser" instead? This would make it call
"InitBrowserMenu", which would let it get one step further, but it would still
fail at:
http://bonsai.mozilla.org/cvsblame.cgi?file=mozilla/xpfe/components/intl/nsCharsetMenu.cpp&rev=1.40&mark=898#865
since firefox.js doesn't include kBrowserCachePrefKey
("intl.charsetmenu.browser.cache").

Maybe these asserts should just be replaced with rv checks?

Comment 3

[Jungshik Shin]

*** Bug 323714 has been marked as a duplicate of this bug. ***

Comment 4

[:Gavin Sharp [email: [e-mail]]]

(In reply to comment #2)
> Should Firefox be sending "browser" instead? This would make it call
> "InitBrowserMenu", which would let it get one step further, but it would still
> fail at:
> http://bonsai.mozilla.org/cvsblame.cgi?file=mozilla/xpfe/components/intl/nsCharsetMenu.cpp&rev=1.40&mark=898#865
> since firefox.js doesn't include kBrowserCachePrefKey
> ("intl.charsetmenu.browser.cache").

I was wrong about this, it won't fail, since "intl.charsetmenu.browser.cache" is in all.js.

Comment 5

[:Gavin Sharp [email: [e-mail]]]

(though I'm still not sure that that's the right thing to do)

Comment 6

[Henri Sivonen (:hsivonen)]

Marking as a duplicate of a newer bug, because the newer one has a more appropriate patch for m-c.