Question

hi there a have an assignment on multiprocess programming and i’m facing with a problem when i try to read characters line by line (each line consists of 3 character and 1 integer). the problem is when i use fscanf it doesnt work properly and it reads the characters successfully just in every 2 steps. here you can see my code;

#include <stdio.h>     /* basic I/O routines.   */
#include <stdlib.h>
#include <unistd.h>    /* define fork(), etc.   */
#include <sys/types.h> /* define pid_t, etc.    */
#include <sys/wait.h>  /* define wait(), etc.   */
#include <signal.h>    /* define signal(), etc. */
#include <pthread.h>
#include <ctype.h>

void child_process(int);
void parent_process();
void function();
int counter=0;
int bond_number=0;

char* f_strand;

int main(int argc, char* argv[]) {

    counter = atoi(argv[1]);

    function();

    fflush(stdout);

    execl("/usr/bin/killall","killall","tail",(char *) 0);
    return 0;
}

void function(){

    int i,k;
    pid_t child_pid[counter];
    char array[counter];
    char array_opst[counter];
    srand ( time(NULL) );
    int temp;

    FILE* fptr;
    fptr = fopen("sample.txt","w");

    if(! (f_strand=(char*)malloc(counter*sizeof(char))) ){
        printf("Error\n");
        exit(1);
    }

    for(i=0; i<counter; i++){

        temp = rand()%4;

        if(temp==0){
            *(f_strand+i) = 'A';
            fprintf(fptr,"A\n");
        }
        else if(temp==1){
            *(f_strand+i) = 'C';
            fprintf(fptr,"C\n");
        }
        else if(temp==2){
            *(f_strand+i) = 'T';
            fprintf(fptr,"T\n");
        }
        else if(temp==3){
            *(f_strand+i) = 'G';
            fprintf(fptr,"G\n");
        }
    }

    fclose(fptr);

                    for (i = 0; i < counter; i++) {
                      if ( (child_pid[i] = fork() ) < 0) {

                        perror("fork");
                        abort();
                      } 
                      else if (child_pid[i] == 0) {

                        child_process(i);
                        exit(0);
                      }
                    }
    int status;
    pid_t pid;
    int num=counter;
        while (num > 0) {
          pid = wait(&status);
          //printf("Child with PID %ld exited with status 0x%x.\n", (long)pid, status);
          --num;  // Remove pid from the pids array.
        }


    parent_process();
    printf("\n");

}

void child_process(int index){

    int i;
    char str,str1,str2;

    FILE* fptr;
    fptr = fopen("sample.txt","r");

    fseek (fptr, 2*index, SEEK_SET);
    fscanf(fptr,"%c", &str);

    FILE* fptr2;
    fptr2 = fopen("sample_2.txt","a");

       if( str == 'A' ){
            str1='T';
            str2=';';
            fprintf(fptr2,"%c%c%c%d\n",str,str2,str1,index);
            }
       else if( str == 'T' ){
            str1='A';
            str2=';';
            fprintf(fptr2,"%c%c%c%d\n",str,str2,str1,index);
            }
       else if( str == 'G' ){
            str1='C';
            str2=':';
            fprintf(fptr2,"%c%c%c%d\n",str,str2,str1,index);
            }
       else if( str == 'C' ){
            str1='G';
            str2=':';
            fprintf(fptr2,"%c%c%c%d\n",str,str2,str1,index);
            }

    fflush(stdout);
    fclose(fptr);
    fclose(fptr2);
    sleep(1);

}

void parent_process(void){

    FILE* fptr;
    fptr = fopen("sample_2.txt","r");

    char str,str1,str2;
    int index,i,k;
    char array_opst[counter]; //  second strand
    char array_type[counter]; // bond type

    i=0;

    while( !feof(fptr) ){

        if(4==fscanf(fptr,"%c%c%c%d",&str,&str1,&str2,&index)){

            if(feof(fptr))
                break;

            printf("STR = %c\n, K = %d",str,k);

            array_type[index]=str1;
            array_opst[index]=str2;
            printf("WHILE\n");

            if(array_type[index] == ':')
                bond_number=bond_number+2;
            else if(array_type[index] == ';')
                bond_number=bond_number+3;
    }

        i++;
}
    fclose(fptr);

}

this is the output of sample2.txt

C:G5
G:C6
A;T4
C:G3
G:C7
C:G8
G:C2
A;T9
T;A1
C:G0

in the parent_process() it tries to read 3 characters and one integer line by line from sample2.txt file. but it does this process in every two lines. i mean it reads first third fifth lines and goes on. i will be appreciated if you can help and thanks anyway.

Answer 1

This is an interesting multifaceted issue. The first thing to note is that you’re using the c conversion. The C99 standard states as follows (§7.19.6.2¶8).

Input white-space characters (as specified by the isspace function) are skipped, unless
the specification includes a [, c, or n specifier.

Now, let’s take a look at the important line.

    if(4==fscanf(fptr,"%c%c%c%d",&str,&str1,&str2,&index)){

What occurs is that whitespace (the padding between the entries, the line terminator, etc.) is being consumed by fscanf. When this occurs, the read is misaligned such that it fails to read an integer when encountering the d conversion. As a result, fscanf does not evaluate to 4, the branch is not taken and thus the incorrectly read data isn’t printed.

Instead, it requires incorrectly reading every other line to ‘realign’ to the proper input and thus you only see these select lines printed. To evidence this, try replacing the condition with merely 0 < fscanf(...) and the incorrect reads will produce the output to demonstrate the intermediate attempts at reading every other line.