String reversing problem

[Albert]

Why doesn't:

#include <stdio.h>

void reverse(char[], int);

main()
{
char s[5];

s[0] = 'h';
s[1] = 'e';
s[2] = 'l';
s[3] = 'l';
s[4] = 'o';
reverse(s, 5);

for (int i=0; i<=4; i++)
putchar(s);
return 0;
}

void reverse(char s[], int num_elements)
{
int i, j;

for (i=0,j=num_elements-1; (i<=num_elements-1) && (j>=0); i++,j--)
s = s[j];
}

output:

olleh

?

 

[Artie Gold]

Why doesn't:

#include <stdio.h>

void reverse(char[], int);

main()
{
char s[5];

s[0] = 'h';
s[1] = 'e';
s[2] = 'l';
s[3] = 'l';
s[4] = 'o';
reverse(s, 5);

for (int i=0; i<=4; i++)
putchar(s);
return 0;
}

void reverse(char s[], int num_elements)
{
int i, j;

for (i=0,j=num_elements-1; (i<=num_elements-1) && (j>=0); i++,j--)

Think what happens when i==4 and j==1, for example. [Do they still call
it `desk checking'?]

s = s[j];
}

output:

olleh

?

HTH,
--ag

 

[Diptendra]

use this one
#include <stdio.h>


void reverse(char[], int);


main()
{
char s[5];
int i;

s[0] = 'h';
s[1] = 'e';
s[2] = 'l';
s[3] = 'l';
s[4] = 'o';
reverse(s, 5);


for ( i = 0; i<=4; i++)
putchar(s);
return 0;



}


void reverse(char s[], int num_elements)
{
int i, j;

for (i=0,j=num_elements-1; (i<=num_elements-1) && (j>=0); i++,j--)
s = s[j];



}

 

[Albert]

What do you mean by 'desk checking'?

 

[slebetman]

What do you mean by 'desk checking'?

I think he means checking by pen & paper. I'd call it "checking by pen
& paper" although I always use a whiteboard for it.

Computers can only do what you tell them to do. As they say: garbage
in, garbage out. Sometimes when the computer doesn't do what you want
it is worth checking if you told it to do what you thought you wanted.

 

[slebetman]

Why doesn't:

<snip>
s = s[j];

This is a good example of why desk-checking is good. Lets take a
"hello" string shall we? Note that in the "diagram" below I mark the
variable being "read" from with + and the variable being "written" to
with ^.

h e l l o // original string

o e l l o // doing s[0] = s[4]
^ +

o l l l o // doing s[1] = s[3]
^ +

o l l l o // doing s[2] = s[2]
^

o l l l o // doing s[3] = s[1]
+ ^

At this point I hope you see the problem with you code since you've
overwritten the original 'e' with an 'l'.

 

[pai]

hi ,
I think an extra variable is needed to store, bcoz u cant
interchange 2 variable as such.
or is there any way to do it ...
Pai

 

[Richard Heathfield]

pai said:

hi ,
I think an extra variable is needed to store, bcoz u cant
interchange 2 variable as such.
or is there any way to do it ...

There is a way to do this under certain conditions, but it's not a very
bright idea.

 

[Chuck F.]

I think an extra variable is needed to store, bcoz u cant
interchange 2 variable as such. or is there any way to do it

Include context, without which your message is meaningless. For
means on the broken google interface, see my sig below.

Try this, after #include <string.h>:

/* reverse string in place. Return length */
static size_t revstring(char *stg)
{
char *last, temp;
size_t lgh;

if ((lgh = strlen(stg)) > 1) {
last = stg + lgh; /* points to '\0' */
while (last-- > stg) {
temp = *stg; *stg++ = *last; *last = temp;
}
}
return lgh;
} /* revstring */

--
"If you want to post a followup via groups.google.com, don't use
the broken "Reply" link at the bottom of the article. Click on
"show options" at the top of the article, then click on the
"Reply" at the bottom of the article headers." - Keith Thompson
More details at: <http://cfaj.freeshell.org/google/>

 

[haroon]

Why doesn't: [...]
void reverse(char s[], int num_elements)
{
int i, j;

for (i=0,j=num_elements-1; (i<=num_elements-1) && (j>=0); i++,j--)
s = s[j];
}

try writing reverse(...) like this:

/***/
void reverse(char s[], int num_elements)
{
int i, j;
char t;

for (i=0,j=num_elements-1; (i<=(num_elements-1) / 2) && (j>=0);
i++,j--)
{
t = s;
s = s[j];
s[j] = t;
}
}

/***/

then analyze both to figure out whats the difference and what happend.

 

[tmp123]

Include context, without which your message is meaningless. For
means on the broken google interface, see my sig below.

Try this, after #include <string.h>:

/* reverse string in place. Return length */
static size_t revstring(char *stg)
{
char *last, temp;
size_t lgh;

if ((lgh = strlen(stg)) > 1) {
last = stg + lgh; /* points to '\0' */
while (last-- > stg) {
temp = *stg; *stg++ = *last; *last = temp;
}
}
return lgh;
} /* revstring */

--
"If you want to post a followup via groups.google.com, don't use
the broken "Reply" link at the bottom of the article. Click on
"show options" at the top of the article, then click on the
"Reply" at the bottom of the article headers." - Keith Thompson
More details at: <http://cfaj.freeshell.org/google/>

Hi,

I agree on the previous, and, of course, if some day I need to code
something similar I will write more or less the same (specially in an
answer to a beginner).

But, just for fun, and taken into account is the third time this
question has been posted, another version:

int revstring ( char *s )
{
char *e;
int r;
for( e=s+(r=strlen(s))-1; s<e; *s^=*e^=*s^=*e, s++, e--);
return r;
}


Kind regards.

(hope google doesn't heats indentation).

 

[William J. Leary Jr.]

What do you mean by 'desk checking'?

I'd have called it "paper check," or, when I'm feeling whimsical "let's play
computer."

Whatever way it's expressed, it means get out some paper and a pen(cil) and
perform, yourself, the steps the computer will take to execute your program.
Sometimes it's better (or quicker) than a debugger. Sometimes it's the only
way to debug (for very limited platforms, for example). Well worth practicing.

- Bill

 

[Chuck F.]

Chuck F. wrote:
.... snip ...

.... snip ...

But, just for fun, and taken into account is the third time this
question has been posted, another version:

int revstring ( char *s )
{
char *e;
int r;
for( e=s+(r=strlen(s))-1; s<e; *s^=*e^=*s^=*e, s++, e--);
return r;
}

FYI your version invokes undefined behaviour. Try it with a string
of zero chars. My length test is not there for fun. I also have
evil suspicions about the xor operations.

--
"If you want to post a followup via groups.google.com, don't use
the broken "Reply" link at the bottom of the article. Click on
"show options" at the top of the article, then click on the
"Reply" at the bottom of the article headers." - Keith Thompson
More details at: <http://cfaj.freeshell.org/google/>

 

[tmp123]

FYI your version invokes undefined behaviour. Try it with a string
of zero chars. My length test is not there for fun. I also have
evil suspicions about the xor operations.

Hi,

If length is 0, then e=s-1, thus s<e is false exiting loop.

However, if the usage of pointer comparation and pointer substraction
is not welcome, another version:

void revstring ( char *s )
{
char *e;
for( e=s+strlen(s); s!=e-- && s!=e; *s^=*e^=*s++^=*e);
}

Kind regards.

PS: some days ago, someone posted about if "C is easy". I do not known,
but it is tricky.

 

[Keith Thompson]

void revstring ( char *s )
{
char *e;
for( e=s+strlen(s); s!=e-- && s!=e; *s^=*e^=*s++^=*e);
}

Undefined behavior.

PS: some days ago, someone posted about if "C is easy". I do not known,
but it is tricky.

It can be if you go out of your way to make it tricky.

 

[tmp123]

Undefined behavior.

Could be... but, could you prove your statement?

 

[Tim Rentsch]

Could be... but, could you prove your statement?

Even the first half of the control expression (namely, 's!=e--')
can yield undefined behavior if strlen(s) == 0.

 

[Keith Thompson]

Could be... but, could you prove your statement?

The expression
*s^=*e^=*s++^=*e
modifies both *s and *e twice between sequence points.

 

[Christian Bau]

Could be... but, could you prove your statement?

It's kind of obvious. Even if it wasn't, whoever wrote that kind of code
should be slapped silly. Immediate removal from any programming team
that I am involved with.

 

[Chuck F.]

If length is 0, then e=s-1, thus s<e is false exiting loop.

However, if the usage of pointer comparation and pointer
substraction is not welcome, another version:

void revstring ( char *s )
{
char *e;
for( e=s+strlen(s); s!=e-- && s!=e; *s^=*e^=*s++^=*e);
}

Same problem and undefined behaviour. You are not allowed to
generate a pointer that points before s (although you are allowed
to generate one that points just after s).

--
"If you want to post a followup via groups.google.com, don't use
the broken "Reply" link at the bottom of the article. Click on
"show options" at the top of the article, then click on the
"Reply" at the bottom of the article headers." - Keith Thompson
More details at: <http://cfaj.freeshell.org/google/>