Re: Possible errors in call to output_operand_lossage?

[Philipp Thomas <pthomas at suse dot de>]

On Sat, 2 Feb 2002 01:50:39 +0000 (GMT), you wrote:

>Shouldn't we fix gettext so particular functions can be marked as
>non-C-format functions, so it doesn't then mark any of the msgid
>arguments to those functions as c-format, rather than marking every call?

I'd have to pass that to Bruno Haible, one of the gettext maintainers.
Bruno, would that be possible?

>Anyway, %A is a valid printf format, albeit not being used as a printf
>format here.

Yes, it is. And it's probably the reason why xgettext did mark it as
being printf style. One would have to call xgettext with --debug to
see why it decides that a message is printf style.

>If it can't handle valid C99 formats, what does it think of
>the printf extensions used in GCC's diagnostics?

I haven't checked what xgettext does make of them yet.


Philipp