function declaration nested?

[Phil Reardon]

Ive been away from programming for a few years and am having difficulty
accessing a function from a math/engineering library that I want to use .
I thought that double foo(double); inserted in the calling routine would
let me say x=foo(y); But I get a warning from the compilation that says
"warning: nested extern declaration of 'foo'. Thanks for any suggestions

[Micah Cowan]

Re: function declaration nested?

"Phil Reardon" <pcr1@pcrt.co m> writes:
[color=blue]
> Ive been away from programming for a few years and am having difficulty
> accessing a function from a math/engineering library that I want to use .
> I thought that double foo(double); inserted in the calling routine would
> let me say x=foo(y); But I get a warning from the compilation that says
> "warning: nested extern declaration of 'foo'. Thanks for any suggestions[/color]

Please post the minimum compileable (or compile-attemptable) code
that exhibits the problem you are experiencing; until you do that
we can only talk about our suspicions, not point out what we know
to be the problem. "Inserted in the calling routine?"

If you mean you did something like:

void my_name_is_of_n o_consequence(v oid)
{
double foo(double);

x=foo(y);
}

And you got that warning; well, an implementation is within its
rights to complain about anything it wants to. It may be that the
implementation "thinks" that you might have meant to declare
foo() at file-scope, I dunno. As long as it compiles, there isn't
really a problem in this case.

HTH,
Micah

[osmium]

Re: function declaration nested?

Phil Reardon writes:
[color=blue]
> Ive been away from programming for a few years and am having difficulty
> accessing a function from a math/engineering library that I want to use .
> I thought that double foo(double); inserted in the calling routine would
> let me say x=foo(y); But I get a warning from the compilation that says
> "warning: nested extern declaration of 'foo'. Thanks for any suggestions[/color]

C does not allow nesting of functions, a la Pascal. It may be that what you
want is a *pointer to* a function. That comes up commonly in mathematics.
and C has provisions for that.

[Daniel Schüle]

Re: function declaration nested?

[...]
[color=blue]
> C does not allow nesting of functions, a la Pascal. It may be that what[/color]
you[color=blue]
> want is a *pointer to* a function. That comes up commonly in mathematics.
> and C has provisions for that.[/color]

http://gcc.gnu.org/onlinedocs/gcc-3.0.4/gcc_5.html

see nested functions

i couldnt figure out whether this feature refers to gcc-extensions or
to C99 extentions ..

can you say for sure it's not in C99?

thx

--
Daniel

[osmium]

Re: function declaration nested?

Daniel Schüle writes:
[color=blue][color=green]
> > C does not allow nesting of functions, a la Pascal. It may be that what[/color]
> you[color=green]
> > want is a *pointer to* a function. That comes up commonly in[/color][/color]
mathematics.[color=blue][color=green]
> > and C has provisions for that.[/color]
>
> http://gcc.gnu.org/onlinedocs/gcc-3.0.4/gcc_5.html
>
> see nested functions
>
> i couldnt figure out whether this feature refers to gcc-extensions or
> to C99 extentions ..
>
> can you say for sure it's not in C99?[/color]

No, I can't say for sure, I am not a language maven. But the start of the
paragraph you refer to says:


"GNU C provides several language features not found in ISO standard
C."

Which tells me that *this* is a GNU C extension. I also think it is highly
unlikely that they would start nesting at this point in time.
---
Just scanning the list of GNU extensions supports the pedantic set's
insistence on limiting the newsgroup to standard C. There's an *awful lot*
of stuff there! So much that if I were using GNU I might be tempted to
abandon this newsgroup in favor of a GNU newsgroup.

[August Derleth]

Re: function declaration nested?

"osmium" <r124c4u102@com cast.net> wrote in
news:bm976a$k75 6k$1@ID-179017.news.uni-berlin.de on Sat 11 Oct 2003
11:22:46a:

[color=blue]
> So much that if I were using GNU I might be
> tempted to abandon this newsgroup in favor of a GNU newsgroup.[/color]

Please, do. At least as long as you wish to discuss nonstandard extensions
to C.

[Irrwahn Grausewitz]

Re: function declaration nested?

August Derleth <libertarian232 003**@onewest.n et> wrote:
[color=blue]
>"osmium" <r124c4u102@com cast.net> wrote in
>news:bm976a$k7 56k$1@ID-179017.news.uni-berlin.de on Sat 11 Oct 2003
>11:22:46a:
>
>[color=green]
>> So much that if I were using GNU I might be
>> tempted to abandon this newsgroup in favor of a GNU newsgroup.[/color]
>
>Please, do. At least as long as you wish to discuss nonstandard extensions
>to C.[/color]

C'mon, he said /if/, and he didn't discuss nonstandard extensions; the
question was, if a certain feature was (topical) ISO-C or (OT) GNU.

Regards
--
Irrwahn
(irrwahn33@free net.de)

[Dan Pop]

Re: function declaration nested?

In <m3vfqw9uci.fsf @localhost.loca ldomain> Micah Cowan <micah@cowan.na me> writes:
[color=blue]
>"Phil Reardon" <pcr1@pcrt.co m> writes:
>[color=green]
>> Ive been away from programming for a few years and am having difficulty
>> accessing a function from a math/engineering library that I want to use .
>> I thought that double foo(double); inserted in the calling routine would
>> let me say x=foo(y); But I get a warning from the compilation that says
>> "warning: nested extern declaration of 'foo'. Thanks for any suggestions[/color]
>
>Please post the minimum compileable (or compile-attemptable) code
>that exhibits the problem you are experiencing; until you do that
>we can only talk about our suspicions, not point out what we know
>to be the problem. "Inserted in the calling routine?"
>
>If you mean you did something like:
>
> void my_name_is_of_n o_consequence(v oid)
> {
> double foo(double);
>
> x=foo(y);
> }
>
>And you got that warning; well, an implementation is within its
>rights to complain about anything it wants to. It may be that the
>implementati on "thinks" that you might have meant to declare
>foo() at file-scope, I dunno. As long as it compiles, there isn't
>really a problem in this case.[/color]

I suspect he's including a header that already provides a declaration
for foo().

Dan
--
Dan Pop
DESY Zeuthen, RZ group
Email: Dan.Pop@ifh.de

[Dan Pop]

Re: function declaration nested?

In <bm8tc7$jq7lf$1 @ID-179017.news.uni-berlin.de> "osmium" <r124c4u102@com cast.net> writes:
[color=blue]
>Phil Reardon writes:
>[color=green]
>> Ive been away from programming for a few years and am having difficulty
>> accessing a function from a math/engineering library that I want to use .
>> I thought that double foo(double); inserted in the calling routine would
>> let me say x=foo(y); But I get a warning from the compilation that says
>> "warning: nested extern declaration of 'foo'. Thanks for any suggestions[/color]
>
>C does not allow nesting of functions, a la Pascal.[/color]

The diagnostic is about a nested function declaration, not about a nested
function definition. Probably something like:

#include <math.h>

int main()
{
double sin(double);
return 0;
}

would trigger the same diagnostic. If this is the case, indeed, all the
OP has to do is to drop the declaration of foo inside the calling function
because foo is already declared. He may also want to check that the
declaration that is already in scope is the one he expects.

Dan
--
Dan Pop
DESY Zeuthen, RZ group
Email: Dan.Pop@ifh.de

[Micah Cowan]

Re: function declaration nested?

Dan.Pop@cern.ch (Dan Pop) writes:
[color=blue]
> In <m3vfqw9uci.fsf @localhost.loca ldomain> Micah Cowan <micah@cowan.na me> writes:
>[color=green]
> >"Phil Reardon" <pcr1@pcrt.co m> writes:
> >[color=darkred]
> >> Ive been away from programming for a few years and am having difficulty
> >> accessing a function from a math/engineering library that I want to use .
> >> I thought that double foo(double); inserted in the calling routine would
> >> let me say x=foo(y); But I get a warning from the compilation that says
> >> "warning: nested extern declaration of 'foo'. Thanks for any suggestions[/color]
> >
> >Please post the minimum compileable (or compile-attemptable) code
> >that exhibits the problem you are experiencing; until you do that
> >we can only talk about our suspicions, not point out what we know
> >to be the problem. "Inserted in the calling routine?"
> >
> >If you mean you did something like:
> >
> > void my_name_is_of_n o_consequence(v oid)
> > {
> > double foo(double);
> >
> > x=foo(y);
> > }
> >
> >And you got that warning; well, an implementation is within its
> >rights to complain about anything it wants to. It may be that the
> >implementati on "thinks" that you might have meant to declare
> >foo() at file-scope, I dunno. As long as it compiles, there isn't
> >really a problem in this case.[/color]
>
> I suspect he's including a header that already provides a declaration
> for foo().[/color]

But then what did he mean by "inserted in the calling routine?"

-Micah

[Dan Pop]

Re: function declaration nested?

In <m3zng555px.fsf @localhost.loca ldomain> Micah Cowan <micah@cowan.na me> writes:
[color=blue]
>Dan.Pop@cern.c h (Dan Pop) writes:
>[color=green]
>> In <m3vfqw9uci.fsf @localhost.loca ldomain> Micah Cowan <micah@cowan.na me> writes:
>>[color=darkred]
>> >"Phil Reardon" <pcr1@pcrt.co m> writes:
>> >
>> >> Ive been away from programming for a few years and am having difficulty
>> >> accessing a function from a math/engineering library that I want to use .
>> >> I thought that double foo(double); inserted in the calling routine would
>> >> let me say x=foo(y); But I get a warning from the compilation that says
>> >> "warning: nested extern declaration of 'foo'. Thanks for any suggestions
>> >
>> >Please post the minimum compileable (or compile-attemptable) code
>> >that exhibits the problem you are experiencing; until you do that
>> >we can only talk about our suspicions, not point out what we know
>> >to be the problem. "Inserted in the calling routine?"
>> >
>> >If you mean you did something like:
>> >
>> > void my_name_is_of_n o_consequence(v oid)
>> > {
>> > double foo(double);
>> >
>> > x=foo(y);
>> > }
>> >
>> >And you got that warning; well, an implementation is within its
>> >rights to complain about anything it wants to. It may be that the
>> >implementati on "thinks" that you might have meant to declare
>> >foo() at file-scope, I dunno. As long as it compiles, there isn't
>> >really a problem in this case.[/color]
>>
>> I suspect he's including a header that already provides a declaration
>> for foo().[/color]
>
>But then what did he mean by "inserted in the calling routine?"[/color]

The very thing you have illustrated above, but there is already a
declaration for foo() in scope, coming from some included header.

Dan
--
Dan Pop
DESY Zeuthen, RZ group
Email: Dan.Pop@ifh.de