RE: use unsigned iterator i in generated source

> -----Original Message-----
> From: Lorenzo Bettini [mailto:[email protected]]
> Sent: Tuesday, June 27, 2006 5:38 PM
> To: Users list for gengetopt
> Subject: Re: [help-gengetopt] use unsigned iterator i in generated
> source
> 
> 
> Gyozo, PAPP (VBuster) wrote:
> > Hi,
> > 
> >> -----Original Message-----
> >> From: Lorenzo Bettini [mailto:[email protected]]
> >> Sent: Saturday, June 03, 2006 4:11 PM
> >> Gyozo, PAPP (VBuster) wrote:
> >>> Hi,
> >>>
> >>> I running the latest 2.17rc. This version generates the 
> >> getoptions_file_save() functions with a signed int i which 
> >> results a lots of signed/unsigned warning. Please modify the 
> >> corresponding .h_skel to use unsigned int.
> >> Hi
> >>
> >> I checked this matter, and the code is
> >>
> >> for (i = 0; i < args_info->@opt_var@_given; ++i)
> >>
> >> which should not generate warnings, since the _given field 
> is an int, 
> >> and not an unsigned.  I myself always compile the tests with 
> >> -Wall so I 
> >> should have received the warning too, which didn't happen...  what 
> >> warning do you get (in particular, which line of code 
> generates the 
> >> warning)?
> > 
> > Sorry for the inconvenience, this was my fault. I patched 
> my gengetopt and forgot it.
> > 
> 
> so the warning doesn't show up?

No, with the original tarball. This was the root of evil:

@@ -143,7 +141,7 @@
 @parser_name@_file_save(const char *filename, struct @args_info@ *args_info)
 {
   FILE *outfile;
-  int i = 0;
+  unsigned int i = 0;
 
   outfile = fopen(filename, "w");
 
Quite obvious. Once again, sorry to blame.