cvs.gedasymbols.org/archives/browse.cgi   search  
Mail Archives: djgpp/1999/08/12/18:18:03

Message-Id: <199908120623.JAA20842@ankara.Foo.COM>
From: "S. M. Halloran" <mitch AT duzen DOT com DOT tr>
Organization: User RFC 822- and 1123-compliant
To: djgpp AT delorie DOT com
Date: Thu, 12 Aug 1999 10:28:52 +0200
MIME-Version: 1.0
Subject: Re: funny errors from RHIDE
In-reply-to: <Ybcs3.18$sM4.36@dfiatx1-snr1.gtei.net>
X-mailer: Pegasus Mail for Win32 (v3.11)
Reply-To: djgpp AT delorie DOT com
X-Mailing-List: djgpp AT delorie DOT com
X-Unsubscribes-To: listserv AT delorie DOT com

It really shouldn't be too tough for you to copy/import/whatever short code 
into an email message now, rather than attach it (?).  Anyway....

On 11 Aug 99, difman was found to have commented thusly:

> I tried to compile my program using djgpp in RHIDE, but it gives errors
> like Error: 'FILE' undeclared (first use in function) and 'x' undeclared
> (first use in function) even though i have declared them in the function.
> it also gives Error: 'EOF' undeclared whats the deal?
> 

Your code which you attached:

======================================

#include <stdlib.h>

main(void)
{
      FILE *text;
      char filename[10];
      char mytext[1000];
      int x = 0;
      int y = 0;

      printf("Enter name of file: ");
      gets(filename, 80);

      text = fopen(filename, "r");

      while(mytext[x] != EOF) {
                 x++;
                 }

      while(y != x/2) {
                mytext[y] = fgetc(filename);
                putchar(mytext[y]);
                y++;
                }

      fclose(*text);

      return 0;
}
=================================
This is the list of errors and warnings the compiler produced within RHide 
using my particular settings:

(5) Warning:  return-type defaults to `int'
(6) Error: `FILE' undeclared (first use in this function)
(6) Error: (Each undeclared identifier is reported only once
(6) Error: for each function it appears in.)
(6) Error: `text' undeclared (first use in this function)
(6) Warning:  statement with no effect
(7) Error: parse error before `char'
(12) Warning:  implicit declaration of function `printf'
(13) Warning:  implicit declaration of function `gets'
(13) Error: `filename' undeclared (first use in this function)
(14) Warning:  implicit declaration of function `fopen'
(15) Error: `mytext' undeclared (first use in this function)
(15) Error: `x' undeclared (first use in this function)
(15) Error: `EOF' undeclared (first use in this function)
(18) Error: `y' undeclared (first use in this function)
(19) Warning:  implicit declaration of function `fgetc'
(20) Warning:  implicit declaration of function `putchar'
(23) Warning:  implicit declaration of function `fclose'

But there are even more serious flaws that are not revealed by the compiler, 
namely serious errors in logic such as the while { } loop for an uninitialized 
array in which the conditional test will always fail.  Probably you want to 
visit the C language group first and get questions regarding use of C answered 
there.  They can tell you things about the reserved words, as well as the 
functions of the standard C library, and the headers included for certain 
functions.

This is also probably a good time to make sure that you not only deal with 
compiler error messages, but also compiler *warning* messages, because lots of 
programs crash or have unintended/undefined effects and they wouldn't if the 
programmer dealt with the 98% of the warning messages that the compiler 
produces.  Indeed, you have the option of treating warnings as errors if you 
check 'pedantic-errors' below.

Below is my own list of compiler warning messages I like to see in general.  If 
you go to Options->Compiler->Warnings in RHide, and make sure the following 
warnings are checked at the very least, you can probably eliminate most 
aggravating crashes that occur at run-time.  By the way, you need to check the
-O option in Options->Compiler->Optimizations if you want to get warnings about 
uninitialized variables, a big source of problems.  The only warnings I am 
ignoring routinely are the "passing arg # as signed/unsigned due to prototype"; 
all the others should really be corrected and eliminated. (Btw, I welcome 
comments from others as to whether my warning list is too little or too much.)

RHide warning message options:

 [ ] -pedantic
 [ ] -pedantic-errors
 [ ] -w
 [x] -W
 [ ] -Waggregate-return
 [x] -Wall
 [x] -Wbad-function-cast
 [x] -Wcast-align
 [x] -Wcast-qual
 [ ] -Wchar-subscripts
 [x] -Wcomment
 [x] -Wconversion
 [ ] -Wenum-clash
 [ ] -Werror
 [x] -Wformat
 [ ] -Wid-clash-
 [ ] -Wimplicit
 [ ] -Winline
 [ ] -Wlarger-than-
 [x] -Wmissing-declarations
 [x] -Wmissing-prototypes
 [x] -Wnested-externs
 [ ] -Wno-import
 [ ] -Woverloaded-virtual
 [x] -Wparentheses
 [x] -Wpointer-arith
 [x] -Wredundant-decls
 [ ] -Wreorder
 [x] -Wreturn-type
 [x] -Wshadow
 [x] -Wstrict-prototypes
 [ ] -Wswitch
 [ ] -Wsynth
 [ ] -Wtemplate-debugging
 [x] -Wtraditional
 [ ] -Wtrigraphs
 [x] -Wuninitialized
 [x] -Wunused
 [x] -Wwrite-strings


Mitch Halloran
Research (Bio)chemist
Duzen Laboratories Group
Ankara       TURKEY

- Raw text -


  webmaster     delorie software   privacy  
  Copyright © 2019   by DJ Delorie     Updated Jul 2019