sh?tpile of errors

[Richard]

Re: sh?tpile of errors

santosh <santosh.k83@gm ail.comwrites:

Richard wrote:

>"Bill Cunningham" <nospam@nspam.c omwrites:

>

>>Here's a reposting of the source.
>>>
>>#include <stdio.h>
>>#include <stdlib.h>
>>#include <ctype.h>
>>#define ex exit(EXIT_FAILU RE)
>>>
>>int main (int argc, char *argv[]) {
>> if (argc!=4) {
>> puts("print usage error");
>> ex;
>> }
>> if (isalpha(argv[1]) || isalpha(argv[2]) {
>> puts("print non alpha allowed");
>> ex;
>> }
>> FILE *fp;
>> double x,y;
>> x=strtod(argv[1],NULL);
>> y=strtod(argv[2],NULL);
>> if (fp=fopen(argv[3],"a"))==EOF) {
>> puts("fopen error");
>> }
>> fprintf(fp,"%.2 f",x,y);
>> if (fclose(fp))==E OF) {
>> puts("fclose error");
>> }
>>return 0;
>>}

>>
>I needed a good laugh so I complied the rubbish above. It took 10
>seconds to get it to compile with only warnings.

>
<snip>
>
Why republish horribly broken code when you aren't going to attempt to
correct it?
>
That it compiles is beside the point.
>

Don't be such a moron all your life Santosh. It compiling is the first
step to get Bill to fix his own bugs. So no, compiling it is NOT beside
the point.

[Bill Cunningham]

Re: sh?tpile of errors

Santosh or anyone,

isalpha takes as its parameter an int. If it is testing to see if something
is a char it doesn't make sense to me to take an int. What am I not seeing?

http://www.cppreference.com/stdstring/isalpha.html

This is a pretty good reference page. char **argv is the parameter I am
passing to main. isalpha is not going to take this. It wants an int. Am I
going to have to use atoi in here somewhere just to test if something is a
char or not?

Bill

[Keith Thompson]

Re: sh?tpile of errors

"Bill Cunningham" <nospam@nspam.c omwrites:

"Richard" <rgrdev@gmail.c omwrote in message
news:g77e50$52s $1@registered.m otzarella.org.. .
>
Non production code!
>
Yes that's right Richard. Non production meant to be used only by me or I
would use exit(EXIT_FAILU RE) everytime it is to be used and not the #define
ex macro.

In code meant to be used only by you, you can do anything you like.

If you want to post it here and expect to get help, I suggest that you
drop the "ex" macro and write "exit(EXIT_FILA URE)", so that it's
easier for us to read your code and offer advice.

--
Keith Thompson (The_Other_Keit h) kst-u@mib.org <http://www.ghoti.net/~kst>
Nokia
"We must do something. This is something. Therefore, we must do this."
-- Antony Jay and Jonathan Lynn, "Yes Minister"

[Keith Thompson]

Re: sh?tpile of errors

"Bill Cunningham" <nospam@nspam.c omwrites:

"Keith Thompson" <kst-u@mib.orgwrote in message
news:lnwsixbz3w .fsf@nuthaus.mi b.org...
>

>Yes. (*blush*)
>>

You left me kinda scratching my head there on that one Keith. I'd never
seen int *argv[] before. But what do I know.

"int *argv[]" was a mistake on my part. Please ignore it.

What I meant was "char *argv[]".

To restore some context and correct my own error:

argv can be declared either as char **argv or as char *argv[].

--
Keith Thompson (The_Other_Keit h) kst-u@mib.org <http://www.ghoti.net/~kst>
Nokia
"We must do something. This is something. Therefore, we must do this."
-- Antony Jay and Jonathan Lynn, "Yes Minister"

[santosh]

Re: sh?tpile of errors

Bill Cunningham wrote:

>
"santosh" <santosh.k83@gm ail.comwrote in message
news:g77fjv$97o $1@registered.m otzarella.org.. .
>

>Why republish horribly broken code when you aren't going to attempt
>to correct it?
>>
>That it compiles is beside the point.

>
This has been re-written Santosh. [ ... ]

My question was at Richard.

[santosh]

Re: sh?tpile of errors

Bill Cunningham wrote:

Santosh or anyone,
>
isalpha takes as its parameter an int. If it is testing to see if
something is a char it doesn't make sense to me to take an int. What
am I not seeing?

This question has been debated to death here. Just search the archives.

http://www.cppreference.com/stdstring/isalpha.html

>
This is a pretty good reference page. char **argv is the parameter I
am passing to main. isalpha is not going to take this.

Right.

It wants an
int. Am I going to have to use atoi in here somewhere just to test if
something is a char or not?

No. What is the type of argv[n][n]? Will that be compatible with the
type that isalpha takes?

[Keith Thompson]

Re: sh?tpile of errors

"Bill Cunningham" <nospam@nspam.c omwrites:

"William Pursell" <bill.pursell@g mail.comwrote in message
news:d32ed3d0-4e12-4b31-8cbf-c5b3608030ff@79 g2000hsk.google groups.com...
>
Assuming gcc & bash, try:
$ gcc foo.c 2>&1 | less

Bill, if you can't get OE Quotefix to work for you, then please pay
attention when you post a followup. If the article to which you're
replying was posted through Google Groups, please do *something* to
indicate the division between the quoted material and your new text.

Here's a reposting of the source.
>
#include <stdio.h>
#include <stdlib.h>
#include <ctype.h>
#define ex exit(EXIT_FAILU RE)
>
int main (int argc, char *argv[]) {
if (argc!=4) {
puts("print usage error");
ex;
}
if (isalpha(argv[1]) || isalpha(argv[2]) {
puts("print non alpha allowed");
ex;
}
FILE *fp;
double x,y;
x=strtod(argv[1],NULL);
y=strtod(argv[2],NULL);
if (fp=fopen(argv[3],"a"))==EOF) {
puts("fopen error");
}
fprintf(fp,"%.2 f",x,y);
if (fclose(fp))==E OF) {
puts("fclose error");
}
return 0;
}

This still has syntax errors. Look at the line with the isalpha
calls. Count the parentheses.

Here's your next assignment. Do not post code that produces an error
message containing the phrase "syntax error" or "parse error". In
most cases, that will be the very first error reported by the
compiler. Fix that error and recompile. Repeat until you have no
syntax errors. *Then* you can post your code and ask for further
help.

--
Keith Thompson (The_Other_Keit h) kst-u@mib.org <http://www.ghoti.net/~kst>
Nokia
"We must do something. This is something. Therefore, we must do this."
-- Antony Jay and Jonathan Lynn, "Yes Minister"

[Bill Cunningham]

Re: sh?tpile of errors


"santosh" <santosh.k83@gm ail.comwrote in message
news:g77hvk$fpe $1@registered.m otzarella.org.. .

> This has been re-written Santosh. [ ... ]

>
My question was at Richard.

My mistake.

Bill

[Richard]

Re: sh?tpile of errors

santosh <santosh.k83@gm ail.comwrites:

Bill Cunningham wrote:
>

> Santosh or anyone,
>>
>isalpha takes as its parameter an int. If it is testing to see if
>something is a char it doesn't make sense to me to take an int. What
>am I not seeing?

>
This question has been debated to death here. Just search the
archives.

So has every other question asked here. If you are not willing to
explain, why are you here? Will you refer posters to other posts about
casting the return of malloc, the return type of main etc? I doubt it.

>

>http://www.cppreference.com/stdstring/isalpha.html
>>
>This is a pretty good reference page. char **argv is the parameter I
>am passing to main. isalpha is not going to take this.

>
Right.
>

>It wants an
>int. Am I going to have to use atoi in here somewhere just to test if
>something is a char or not?

>
No. What is the type of argv[n][n]? Will that be compatible with the
type that isalpha takes?
>

You are wrong. He might well have to use it if the characters are there
as part of a command line integer parameter. He needs to convert the
character sequence to that integer.

--

[Bill Cunningham]

Re: sh?tpile of errors


"santosh" <santosh.k83@gm ail.comwrote in message
news:g77i4s$ahg $1@registered.m otzarella.org.. .

No. What is the type of argv[n][n]? Will that be compatible with the
type that isalpha takes?

Oh I see. I've never used argv[][]'s before. I see so much clearly now.
Thank you Thank you. I am going to try this code again.

Bill

[santosh]

Re: sh?tpile of errors

Richard wrote:

santosh <santosh.k83@gm ail.comwrites:

<snip>

>Why republish horribly broken code when you aren't going to attempt
>to correct it?
>>
>That it compiles is beside the point.
>>

>
Don't be such a moron all your life Santosh. It compiling is the first
step to get Bill to fix his own bugs.

No. Understanding C is the first step.

A typical C compiler will gladly accept highly bastardised C and will
attempt to generate an executable. Fixing errors reported by a compiler
*after* having learnt the language is one thing, but depending on
compiler diagnostics to learn the rules and semantics of C is not very
bright, IMO.

So no, compiling it is NOT beside the point.

Unless invoked in conforming mode an implementation is allowed to
silently accept anything and produce an executable. Even under
conforming mode, only diagnostics are guaranteed. Undefined behaviour
that is not a constraint violation might still be silently accepted and
typically you do *not* want that.

It's better to learn the language first and then use the compiler,
rather than depending on it to learn the language.

Bill should re-open K&R and start with the first page. I understand that
he has not progressed beyond the first few pages after several years,
but he really has no other alternative. Merely messing around with
programs written with guesswork and compiled with luck isn't going to
get him anywhere.

[santosh]

Re: sh?tpile of errors

Richard wrote:

santosh <santosh.k83@gm ail.comwrites:

>Bill Cunningham wrote:

>>char **argv is the parameter I am passing to main. isalpha is not
>>going to take this. It wants an int. Am I going to have to use atoi
>>in here somewhere just to test if something is a char or not?

>>
>No. What is the type of argv[n][n]? Will that be compatible with the
>type that isalpha takes?

>
You are wrong. He might well have to use it if the characters are
there as part of a command line integer parameter. He needs to convert
the character sequence to that integer.

Don't you find it tiresome to disagree just for the sake of it?

My response was directed at the specific complaint by Bill that
*isalpha* won't like argv as an argument. Clearly he wants to run
through the strings to check whether any of them contain alphabetic
characters (which would be error).

So what you say above makes no sense at all.

[Richard]

Re: sh?tpile of errors

santosh <santosh.k83@gm ail.comwrites:

Richard wrote:

>santosh <santosh.k83@gm ail.comwrites:

>
<snip>
>

>>Why republish horribly broken code when you aren't going to attempt
>>to correct it?
>>>
>>That it compiles is beside the point.
>>>

>>
>Don't be such a moron all your life Santosh. It compiling is the first
>step to get Bill to fix his own bugs.

>
No. Understanding C is the first step.

Sigh. Obviously.

>
A typical C compiler will gladly accept highly bastardised C and will
attempt to generate an executable. Fixing errors reported by a compiler
*after* having learnt the language is one thing, but depending on
compiler diagnostics to learn the rules and semantics of C is not very
bright, IMO.

Learning by doing Santosh. its why I recommend using a debugger. Clearly
Bill is incapable of reading code.

The point, which even you in your most "clc reg" mode must see that
fixing the bugs so he can COMPILE the code can only be benefifical. He
could not even match brackets for goodness sake.

[Richard]

Re: sh?tpile of errors

santosh <santosh.k83@gm ail.comwrites:

Richard wrote:

>santosh <santosh.k83@gm ail.comwrites:

>>Bill Cunningham wrote:

>

>>>char **argv is the parameter I am passing to main. isalpha is not
>>>going to take this. It wants an int. Am I going to have to use atoi
>>>in here somewhere just to test if something is a char or not?
>>>
>>No. What is the type of argv[n][n]? Will that be compatible with the
>>type that isalpha takes?

>>
>You are wrong. He might well have to use it if the characters are
>there as part of a command line integer parameter. He needs to convert
>the character sequence to that integer.

>
Don't you find it tiresome to disagree just for the sake of it?

I would if that were the reason. But you were wrong. Or appeared to be
wrong.

He asked if he could use atoi in there somewhere. And yes, he can
indeed. Remember we are talking about Bill. You talked about
isalpha. However he still needs the integer values from the chars passed
in.

[santosh]

Re: sh?tpile of errors

Richard wrote:

santosh <santosh.k83@gm ail.comwrites:
>

>Richard wrote:

>>santosh <santosh.k83@gm ail.comwrites:

>>
><snip>
>>

>>>Why republish horribly broken code when you aren't going to attempt
>>>to correct it?
>>>>
>>>That it compiles is beside the point.
>>>>
>>>
>>Don't be such a moron all your life Santosh. It compiling is the
>>first step to get Bill to fix his own bugs.

>>
>No. Understanding C is the first step.

>
Sigh. Obviously.
>

>>
>A typical C compiler will gladly accept highly bastardised C and will
>attempt to generate an executable. Fixing errors reported by a
>compiler *after* having learnt the language is one thing, but
>depending on compiler diagnostics to learn the rules and semantics of
>C is not very bright, IMO.

>
Learning by doing Santosh. its why I recommend using a debugger.
Clearly Bill is incapable of reading code.
>
The point, which even you in your most "clc reg" mode must see that
fixing the bugs so he can COMPILE the code can only be benefifical. He
could not even match brackets for goodness sake.

The very reason he started this thread was because he *could* *not*
understand the errors and warnings his compiler gave him. And that is a
direct result of the fact that he has only correctly learnt a minuscule
fraction of C (though he incorrectly uses a rather larger portion of
it.)

You need to know the language to understand all but the most simple of
typical compiler messages. Us telling him how to fix the code to shut
the compiler up is no good, since he doesn't understand *why* those
fixes are needed. There really is no alternative for him but to keep
trying from the very beginning of a good tutorial, if he really is
serious about this.

I'm surprised that you advocate this "learning by doing" philosophy for
Bill of all the people. Someone who can't match parenthesis isn't going
to get very far with gdb or gcc. It's better to understand than to
learn.