«

Hi.

I've isntalled Bloodshed Dev-C++ 4.9.9 for Windows. It comes with MinGW
3.4.2. I'm trying to use getopt_long() in my console application to parse
multiple command line arguments syntaxes. I do that by looping through many
arrays of struct options. It looks like getopt_long() only processes the
first array of options; it returns -1 immediately on the second array
onwards. Here's the code I used:

#include <cstdlib>
#include <iostream>
#include <getopt.h>

using namespace std;

int main(int argc, char *argv[])
{
static option instArray[] = {
{ "install", required_argument, NULL, 'f' },
{ "name", required_argument, NULL, 'l' },
{ "description", required_argument, NULL, 'n' },
{ "title", required_argument, NULL, 'd' },
{ }
};
static option helpArray[] = {
{ "help", no_argument, NULL, 'h' },
{ "version", no_argument, NULL, 'v' },
{ }
};
static struct optItem {
char* name;
char* shortOpts;
option* opts;
} optArray[] = {
{ "Help", "-f:l:n:d:", helpArray },
{ "Install", "hv", instArray }
};
const unsigned nArrayLength = sizeof(optArray) / sizeof(optArray[0]);

// Loop twice with long option arrays
for ( int i = 0, option,; i < nArrayLength; i++ )
{
int optIndex = 0;
cout << "Arg table: " << optArray[i].name << endl;
while ( (option = getopt_long(argc, argv,
optArray[i].shortOpts, optArray[i].opts,
&optIndex)) != -1
)
if ( option == 1 )
cout << " Argument " << optarg << " found." << endl;
}
}


Here's the output:

Arg table: Help
Argument install found.
D:\Projects\Examples\Getopt\getopt.exe: unknown option -- v
D:\Projects\Examples\Getopt\getopt.exe: unknown option -- b
Arg table: Install


I expected

Arg table: Help
Argument install found.
D:\Projects\Examples\Getopt\getopt.exe: unknown option -- v
D:\Projects\Examples\Getopt\getopt.exe: unknown option -- b
Arg table: Install
Argument install found.
D:\Projects\Examples\Getopt\getopt.exe: unknown option -- v
D:\Projects\Examples\Getopt\getopt.exe: unknown option -- b

Did I do something wrong or is it a bug?

Thanks in advance,
Vince C.

Vince C. wrote:
> Hi.
>
> I've isntalled Bloodshed Dev-C++ 4.9.9 for Windows. It comes with MinGW
> 3.4.2. I'm trying to use getopt_long() in my console application to parse
> multiple command line arguments syntaxes. I do that by looping through many
> arrays of struct options. It looks like getopt_long() only processes the
> first array of options; it returns -1 immediately on the second array
> onwards. Here's the code I used:
>

Unfortunately, Vince, getopt_long() is not part of the C++ standard
(ISO/IEC 14882:1998 or 14882:2003), and as such, is off-topic here.
However, I would recommend you ask in gnu.g++.help, where it would be
on-topic.

> [code redacted]

red floyd wrote:

> Unfortunately, Vince, getopt_long() is not part of the C++ standard
> (ISO/IEC 14882:1998 or 14882:2003), and as such, is off-topic here.
> However, I would recommend you ask in gnu.g++.help, where it would be
> on-topic.
>
>> [code redacted]

Sorry, I thought comp.std.c++ was about C++ standards and comp.lang.c++
about everything else. I have no descriptions about the forums in my NNTP
client. I also thought gnu.g++ was about compiler g++ only...

It's a bit tricky to find the appropriate group, I'm sorry :s ...

Thanks for the info. I'll post over gnu.g++.

--

Vince C.

red floyd wrote:

> Unfortunately, Vince, getopt_long() is not part of the C++ standard
> (ISO/IEC 14882:1998 or 14882:2003), and as such, is off-topic here.
> However, I would recommend you ask in gnu.g++.help, where it would be
> on-topic.
>
>> [code redacted]

Too bad... I see only *one* post ("Spammed by FAX") in gnu.g++ . I think
that forum is dead... I just hope there is a chance I get a response
somewhere...

--

Vince C.

On Tue, 17 Jul 2007 11:48:45 +0200, Vince C. wrote:

> red floyd wrote:
>
>> Unfortunately, Vince, getopt_long() is not part of the C++ standard
>> (ISO/IEC 14882:1998 or 14882:2003), and as such, is off-topic here.
>> However, I would recommend you ask in gnu.g++.help, where it would be
>> on-topic.
>>
>>> [code redacted]
>
> Too bad... I see only *one* post ("Spammed by FAX") in gnu.g++

Is there actually such a ng? Try gnu.g++.help.

--
Lionel B

Lionel B wrote:

>> Too bad... I see only *one* post ("Spammed by FAX") in gnu.g++
>
> Is there actually such a ng? Try gnu.g++.help.

Yes, there is one. But gnu.g++.help looks a little bit busier so I'll post
there.

Thanks.

--

Vince C.

On Jul 17, 11:45 am, "Vince C." <n...@hotmail.com> wrote:
> red floyd wrote:
> > Unfortunately, Vince, getopt_long() is not part of the C++ standard
> > (ISO/IEC 14882:1998 or 14882:2003), and as such, is off-topic here.
> > However, I would recommend you ask in gnu.g++.help, where it would be
> > on-topic.

> >> [code redacted]

> Sorry, I thought comp.std.c++ was about C++ standards and comp.lang.c++
> about everything else.

comp.std.c++ is about the C++ standard itself. comp.lang.c++ is
about everything that concerns C++ in general, but excludes
platform and implementation specific issues. The distinction
isn't always very clear. Discussion about threading in C++ is
certainly relevant, but basic threading algorithms aren't, nor
are details of a platform-specific threading API. The classic
"test" is that if the answer would be more or less the same for
most platforms, but different for a different language, the post
belongs here, but if the answer would be more or less the same
for different languages, but different for different platforms,
it's off topic. But of course, that test supposes that you know
the answer before posing the question.

In this case, getopts_long is very much 1) a GNUism, and 2)
available in C and (probably) in other languages as well.
Questions about how best to encapsulate it in a C++ class (or
whether there exist C++ alternatives) would probably be on
topic, but that's about it.

--
James Kanze (GABI Software) email:
Conseils en informatique orientée objet/
Beratung in objektorientierter Datenverarbeitung
9 place Sémard, 78210 St.-Cyr-l'École, France, +33 (0)1 30 23 00 34