Set X to 5

[thagulf2017]

so i'm reading an exercise on structures in the C k&R book.

it says to modify getword routine in order to take in to consideration

underscores, comments etc.

i'm trying to start small (i.e check for begining of comment and set X to 5 if found) before writing the actual code but it seems hard apparently. heres the code:

#include <stdio.h>
#include <string.h>
#include <ctype.h>


#define MAXWORD 100

#define NKEYS (sizeof keytab/sizeof(keytab[0]))

#define CINDEX 12


#define BUFSIZE 100

char buf[BUFSIZE]; /* buffer for ungetch */
int bufp = 0; /* next free position in buf */
int getch(void) /* get a (possibly pushed-back) character */
{
    return (bufp > 0) ? buf[--bufp] : getchar();
}
void ungetch(int c) /* push character back on input */
{
    if (bufp >= BUFSIZE)
        printf("ungetch: too many characters\n");
    else
        buf[bufp++] = c;
}

struct key{
    char *word;
    int count;
} keytab[] = {
     {"auto", 0},
     {"break", 0},
     {"case", 0},
     {"char", 0},
     {"const", 0},
     {"continue", 0},
     {"default", 0},
     {"for", 0},
     {"unsigned", 0},
     {"signed", 0},
     {"volatile", 0},
     {"void", 0},
     {"comments", 0},
     {"while", 0}
};
int getword(char *, int);
int binsearch(char *, struct key *, int);
int main(){
    int n;
    char word[MAXWORD];

    while(getword(word, MAXWORD) != EOF)
        if( isalpha(word[0]))
            if ( (n = binsearch(word, keytab, NKEYS)) >= 0)
                keytab[n].count++;
    for( n = 0; n < NKEYS; n++ )
        printf("%4d  %s\n", keytab[n].count, keytab[n].word);

    return 0;
}

int getword(char *word, int lim){
    int c, getch(void);
    void ungetch(int c);

    char *w = word;
    while(isspace(c = getch()))
        ;
    if(c != EOF || c == '/'){
        if(c == '/'){
            c = getch();
            if( c == '*'){
                 int x = 5;
            }
        }
        *w++ = c;
    }
    if (!isalpha(c)){
        *w = c;
        return c;
    }

    for(; --lim > 0; w++)
    if(!isalnum(*w = getch())){
        ungetch(*w);
        break;
    }
    *w = '\0';
    return w[0];
}

int binsearch(char *word, struct key keytab[], int n){
    int cond;
    int low, mid, high;

    low = 0;
    high = n - 1;
    while(low <= high){
        mid = (low+high)/2;

        if( (cond = strcmp(word, keytab[mid].word)) < 0)
            high = mid - 1;
        else if ( cond > 0)
            low = mid + 1;
        else
            return mid;
    }
    return -1;

}

when i step through the code in a debugger, i first input '/' then '*' without quotes and i assume this code

if(c == '/'){
            c = getch();
            if( c == '*'){
                 int x = 5;
            }
        }

should handle it, how ever, execution goes thus;

]if(c == '/'){ // evaluates as TRUE so rest code executes
            c = getch(); // debugger just skips this, i don't get to input '*'
            if( c == '*'){ // thus, this is is skipped
                 int x = 5;
            }
        }

what i'm i doing wrong please?

[Salem]

It seems fine.

$ gdb -q ./a.out
Reading symbols from ./a.out...done.
(gdb) b 71
Breakpoint 1 at 0x40088c: file foo.c, line 71.
(gdb) run
Starting program: /home/sc/Documents/a.out 
/* foo */

Breakpoint 1, getword (word=0x7fffffffdda0 "", lim=100) at foo.c:71
71	    if(c != EOF || c == '/'){
(gdb) p/c c
$1 = 47 '/'
(gdb) n
72	        if(c == '/'){
(gdb) 
73	            c = getch();
(gdb) 
74	            if( c == '*'){
(gdb) p/c c
$2 = 42 '*'
(gdb) n
75	                 int x = 5;

The problem is, x is locally declared within that block, and you don't use x for anything. So it's all too easy for the compiler to remove it as an unused statement.

Eg.

$ gcc -Wall -g foo.c
foo.c: In function ‘getword’:
foo.c:75:22: warning: unused variable ‘x’ [-Wunused-variable]
                  int x = 5;
                      ^

Try with something a little more obvious, say

            c = getch();
            if( c == '*'){
                 fprintf(stderr,"DEBUG: comment start\n");
            }

$ gcc -Wall -g foo.c
$ ./a.out 
auto
auto
/* comment */
DEBUG: comment start
   2  auto
   0  break
   0  case
   0  char
   0  const
   0  continue
   0  default
   0  for
   0  unsigned
   0  signed
   0  volatile
   0  void
   0  comments
   0  while

[thagulf2017]

thanks very much, that has been resolved but now i don't know why ret which is to store the return value of getword in order to decide whether its a comment or normal keyword is always returning value 1.

please help and sorry for my noobish questions @salem

#include <stdio.h>
#include <string.h>
#include <ctype.h>


#define MAXWORD 100

#define NKEYS (sizeof keytab/sizeof(keytab[0]))

#define CINDEX 12


#define BUFSIZE 100

char buf[BUFSIZE]; /* buffer for ungetch */
int bufp = 0; /* next free position in buf */
int getch(void) /* get a (possibly pushed-back) character */
{
    return (bufp > 0) ? buf[--bufp] : getchar();
}
void ungetch(int c) /* push character back on input */
{
    if (bufp >= BUFSIZE)
        printf("ungetch: too many characters\n");
    else
        buf[bufp++] = c;
}

struct key{
    char *word;
    int count;
} keytab[] = {
     {"auto", 0},
     {"break", 0},
     {"case", 0},
     {"char", 0},
     {"const", 0},
     {"continue", 0},
     {"default", 0},
     {"for", 0},
     {"unsigned", 0},
     {"signed", 0},
     {"volatile", 0},
     {"void", 0},
     {"comments", 0},
     {"while", 0}
};
int getword(char *, int);
int binsearch(char *, struct key *, int);
int main(){
    int n;
    char word[MAXWORD];
    int ret;
    while(ret = (getword(word, MAXWORD) != EOF)){
        if (ret = CINDEX)
            keytab[CINDEX].count++;
        else {
        if( isalpha(word[0]))
            if ( (n = binsearch(word, keytab, NKEYS)) >= 0)
                keytab[n].count++;
        }
    }
    for( n = 0; n < NKEYS; n++ )
        printf("%4d  %s\n", keytab[n].count, keytab[n].word);

    return 0;
}

int getword(char *word, int lim){
    int c, getch(void);
    void ungetch(int c);

    char *w = word;
    while(isspace(c = getch()))
        ;
    if(c != EOF || c == '/'){
        if(c == '/'){
            c = getch();
            if( c == '*'){
                lim -= 2;
                for(; (c = getch()) != '*' && lim > 2; lim--)
                    ;
                if(c == '*'){
                    c = getch();
                    if(c == '/')
                        return CINDEX;
                }

            }
        }
        else
            *w++ = c;
    }
    if (!isalpha(c)){
        *w = c;
        return c;
    }

    for(; --lim > 0; w++)
    if(!isalnum(*w = getch())){
        ungetch(*w);
        break;
    }
    *w = '\0';
    return w[0];
}

int binsearch(char *word, struct key keytab[], int n){
    int cond;
    int low, mid, high;

    low = 0;
    high = n - 1;
    while(low <= high){
        mid = (low+high)/2;

        if( (cond = strcmp(word, keytab[mid].word)) < 0)
            high = mid - 1;
        else if ( cond > 0)
            low = mid + 1;
        else
            return mid;
    }
    return -1;

}

[Salem]

You absolutely must start using -Wall in your compilations to filter out all the stupid mistakes like using = when you mean ==

$ gcc -Wall foo.c
foo.c: In function ‘main’:
foo.c:54:5: warning: suggest parentheses around assignment used as truth value [-Wparentheses]
     while(ret = (getword(word, MAXWORD) != EOF)){
     ^
foo.c:55:9: warning: suggest parentheses around assignment used as truth value [-Wparentheses]
         if (ret = CINDEX)
         ^

Also, instead of returning some magic number, why don't you just store the word "comments" in the word parameter, then main works just the same as before without any need for special cases, or forgetting to change CINDEX when you add more words to your keytab.

In addition, your keytab list MUST be sorted into alphabetic order for binsearch to work properly.

[thagulf2017]

You absolutely must start using -Wall in your compilations to filter out all the stupid mistakes like using = when you mean ==

$ gcc -Wall foo.c
foo.c: In function ‘main’:
foo.c:54:5: warning: suggest parentheses around assignment used as truth value [-Wparentheses]
     while(ret = (getword(word, MAXWORD) != EOF)){
     ^
foo.c:55:9: warning: suggest parentheses around assignment used as truth value [-Wparentheses]
         if (ret = CINDEX)
         ^

Also, instead of returning some magic number, why don't you just store the word "comments" in the word parameter, then main works just the same as before without any need for special cases, or forgetting to change CINDEX when you add more words to your keytab.

In addition, your keytab list MUST be sorted into alphabetic order for binsearch to work properly.

i use codeblocks, how can i compile with -Wall flag?

also, i believe comments are different from the other keywords in that they can be anything or any length as long as they fall
within /* and */ but the others are fixed keywords so no way for strcmp to compare - this ofcourse is my own level of understanding. Please shed more light sir.

[Salem]

I don't mean return the content of the comment, I mean return the literal string "comments".

Like
strcpy(word,"comments");

Also, this is how you enable -Wall
Project->Build Options...

[jimblumberg]

You should also check the -Wextra and -pedantic boxes and select one of the more modern C/C++ standards (C11/C99) as well.

Jim

[thagulf2017]

I don't mean return the content of the comment, I mean return the literal string "comments".

Like
strcpy(word,"comments");

Also, this is how you enable -Wall
Project->Build Options...

applied strcpy() as suggested, code is much neater and works too, thanks salem. :

i have another problem tho, i was trying to implement binsearch using pointers but before changing the actual function i am playing with normal pointer arithmetic. see this

#include <stdio.h>


int main(){

int x[] = {1,2,3};
int *y = x+2;
int *z = x;
int *mid;

mid = (z+y)/2;

 printf("%d \t\t %p \n", *mid, mid);

}

when i compile i get error;

C:\Users\xx\Documents\tc.c|11|error: invalid operands to binary + (have 'int *' and 'int *')|

whats wrong? in k&R they used

struct key *low = &tab[0];
struct key *high = &tab[n];
struct key *mid;
while (low < high) {
mid = low + (high-low) / 2;

and it works

[laserlight]

You can subtract pointers (i.e., to obtain the difference in positions between the elements that the pointers point to), but you cannot add them.

[thagulf2017]

You can subtract pointers (i.e., to obtain the difference in positions between the elements that the pointers point to), but you cannot add them.

im confused right now, isn't both *low' and *high' pointers in the context used in the k & R?

[laserlight]

im confused right now, isn't both *low' and *high' pointers in the context used in the k & R?

Yes, but let's look at the statement from K&R:

mid = low + (high-low) / 2;

Let's replace them with the types:

int* = int* + (int* - int*) / int;

Now, let's resolve the expression on the right hand of the assignment, considering only the types:

int* = int* + (int) / int;
int* = int* + int;
int* = int*;

I simplified by using int instead of ptrdiff_t. As you can see, at no point are pointers added together. Now let's contrast with yours:

mid = (z+y)/2;

considering the types only:

int* = (int* + int*) / int;

here we can immediately see that the expression does not make sense: what does it mean to add an int* to an int*? Therefore, it is an error.

[thagulf2017]

i think i understand this now (*high - *low) returns int not a pointer so it makes sense now, thanks salem