Usenet 1994 January/usenetsourcesnewsgroupsinfomagicjanuary1994.iso/sources/std_unix/volume.29/text0072.txt

Text File | 1992-12-26 | 1.9 KB | 43 lines

Submitted-by: [e-mail] (Norman Diamond)

In article <[e-mail]> [e-mail] (Ronald F. Guilmette) writes:
>Certain of the functions which must be declared (preferably with prototypes)
>in the <stdio.h> file (e.g. vprintf) are defined (by the ANSI C standard) to
>take arguments of type `va_list'.

If a user's program declares printf instead of doing #include <stdio.h> then
the user's program must include a prototype with ellipsis ", ...".
If you're talking about the internals of the implementation's <stdio.h> then
the implementation is allowed to encode the declarations however it likes,
not necessarily encoded in the C language, not necessarily using prototypes
-- the implementation can choose any method that works for itself.

>But careful implementors know that if `va_list' is declared whever <stdio.h>
>is included, this will cause an small but unnecessary (and possibly illegal)
>pollution of the user's namespace.
>My first question is simply this.  Is the practice of avoiding definition
>of a va_list type in <stdio.h> strictly required by the ANSI C standard?

Yes.  s/possibly//  And carelessness yields non-conforming implementations.

>In what other cases are such "hidden built-in type hacks" required as a result
>of other requirements in the ANSI C standard (or in POSIX 1003.1-1990)?

It's probably easiest to encode hidden built-in type hacks for EVERYTHING and
avoid exactly this worry (assuming that you're an implementor).

[Examples deleted:  size_t in <time.h> and time_t in <sys/stat.h>]

ANSI C section 4.12.1, page 171 line 12, says that <time.h> declares size_t.
I don't have the POSIX standard.

--
Norman Diamond       [e-mail]
If this were the company's opinion, I wouldn't be allowed to post it.
"It's been a lovely recession."

[ Note followup-to: line, please.  Thanks -- mod ]

Volume-Number: Volume 29, Number 73