Thread: Can't get fgets to work properly

Hello.
I'm making a small application to make permutations of strings in order to find any anagrams. These permutations are then checked against a dictionary file using fgets. It compiles nicely but when it is run it just crashes.
Here is the entire source code but I suspect only the "checkDict" function is faulty because everything runs just fine when I comment it out.

Code:

#include <stdio.h>
#define VERSION 0.7
#define MAX_DICT_LEN 100
#define MAX_RESULTS 10

int maxResultsCounter = 0;
char *temp;
FILE *dict;

// check current permutation against entire dict.txt
void checkDict(char *str){
	while(fgets(temp, MAX_DICT_LEN, dict) != NULL)
		if(strcmp(str, temp) == 0){
			printf("%s\n", str);
			maxResultsCounter++;
		}
}

// swap *first and *second
void swap(char* first, char* second){
	char temp = *second;
	*second = *first;
	*first = temp;
}

// compute permutations
int perm(char* set, int begin, int end){
	int i, c = 0, range = end - begin;
	if(range == 1){
		
		// MAX_RESULTS invariant
		if(maxResultsCounter > MAX_RESULTS)
			return 0;
		//checkDict(set);
	}
	else
		for(i = 0; i < range; i++){
			swap(&set[begin], &set[begin+i]);	// first swap
			perm(set, begin+1, end);			// recursion
			swap(&set[begin], &set[begin+i]);	// swap back
		}
	return 0;
}

int main(int argc, char *argv[]){
	
	// test arguments
	if(argc == 1 || argc > 3){
		printf("AnagramAnalyzer (v. %2.1f)\n\n", VERSION);
		printf("syntax: \n\t* <anagram> <args>\n\n");
		printf("args: \n\t[-s] adds a space to the anagram\n");
		return 0;
	}		
	
	char *in = argv[1];
	
	// convert to lower case
	int i, len = strlen(in);
	for(i = 0; i < len; i++)
		if(in[i] > 64 && in[i] < 91)
			in[i] += 32;
		
	// check for additional arguments
	if(argc == 3 && strcmp(argv[2], "-s") == 0){
		in[len] = ' ';
		in[len+1] = '\0';
		len++;
	}
	
	// open file stream
	if((dict = fopen("dict.txt", "r")) == NULL){
		printf("Couldn't locate dict file!\n");
		return 0;
	}
	
	// permutate and check dictionary
	printf("Checking...\n");
	perm(in, 0, len);
	
	// stop and close
	fclose(dict);
	return 0;
}

Thanks a lot MK27. It runs without crashing now.

I still can't get it to work though. If I pass a permutation of a word I know is in the dict.txt file, or even the word itself, it doesn't print it.

I'm pretty new to C and I can't really understand why it doesn't work. I may be missing something major though

The string returned by fgets still has a '\n' appended.
That's why strcmp fails.
Kurt

Oh cool thanks Kurt.

I tried doing something along the lines of

Code:

// check current permutation against entire dict.txt
void checkDict(char *str){
	while(fgets(temp, MAX_DICT_LEN, dict) != NULL){
		int i;
		for(i = 0; i < strlen(temp); i++)
			if(temp[i] == '\n')
				temp[i] = "";
		if(strcmp(str, temp) == 0){
			printf("%s\n", str);
			maxResultsCounter++;
		}
	}
}

to remove the undesired '\n' but that won't compile. I tried with empty "" as well. That compiles but still doesn't work. It throws the warning: assignment makes integer from pointer without a cast.

try

Code:

if(temp[i] == '\n')
    temp[i] = 0;

Kurt

Thanks a lot to all of you. It finally works!

I don't know if the thread is now closed or if people would like to help do a bit of optimization. It seems awfully slow with words over 4 characters.

- lonkz

I didn't realise I was doing that. Could you point out where?

Originally Posted by lonkz

I didn't realise I was doing that. Could you point out where?

Code:

void checkDict(char *str){
	while(fgets(temp, MAX_DICT_LEN, dict) != NULL){
		int i;
		for(i = 0; i < strlen(temp); i++)
			if(temp[i] == '\n')
				temp[i] = "";
		if(strcmp(str, temp) == 0){
			printf("%s\n", str);
			maxResultsCounter++;
		}
	}
}

Oh yeah, nicely spotted. I seem to have been looking at the old code originally posted frantically trying to figure out where I made the mistake.

Well needless to say it has been corrected. Thanks for helping out

Many thanks to everyone who helped me. The application now works great and fast for smaller words. However it doesn't seem to like more than 5 characters. I get some pretty weird results.

Here is the new code:

Code:

#include <stdio.h>
#define VERSION 1.0
#define MAX_DICT_LEN 20
#define MAX_RESULTS 10

int maxResultsCounter = 0;
char temp[MAX_DICT_LEN + 1];
FILE *dict;

// check current permutation against entire dict.txt
void checkDict(char *str){
	
	// open file stream and check it
	if((dict = fopen("dict.txt", "r")) == NULL)
		printf("Couldn't locate dict file!\n");
	
	// continuously read line from dict
	int i;
	while(fgets(temp, MAX_DICT_LEN, dict) != NULL){
		int len = strlen(temp);
		for(i = 0; i < len; i++)
			if(temp[i] == '\n')
				temp[i] = '\0';
			
		// compare str to the line read
		if(strcmp(str, temp) == 0){
			printf("%s\n", str);
			maxResultsCounter++;
		}
	}
}

// swap *first and *second
void swap(char* first, char* second){
	char temp = *second;
	*second = *first;
	*first = temp;
}

// compute permutations
int perm(char* set, int begin, int end){
	int i, c = 0, range = end - begin;
	if(range == 1){
		
		// MAX_RESULTS invariant
		if(maxResultsCounter > MAX_RESULTS)
			return 0;
		checkDict(set);
	}
	else
		for(i = 0; i < range; i++){
			swap(&set[begin], &set[begin+i]);	// first swap
			perm(set, begin + 1, end);		// recursion
			swap(&set[begin], &set[begin+i]);	// swap back
		}
	return 0;
}

int main(int argc, char *argv[]){
	
	// test arguments
	if(argc == 1 || argc > 3){
		printf("AnagramAnalyzer (v. %2.1f)\n\n", VERSION);
		printf("syntax: \n\t* <anagram> <args>\n\n");
		printf("args: \n\t[-s] adds a space to the anagram\n");
		return 0;
	}		
	
	// initialize sequence
	char *in = argv[1];
	
	// convert to lower case
	int i, len = strlen(in);
	for(i = 0; i < len; i++)
		if(in[i] > 64 && in[i] < 91)
			in[i] += 32;
		
	// check for additional arguments
	if(argc == 3 && strcmp(argv[2], "-s") == 0){
		in[len] = ' ';
		in[len+1] = '\0';
		len++;
	}
	
	// permutate and check dictionary
	printf("Checking...\n");
	perm(in, 0, len);
	
	// stop and close
	fclose(dict);
	return 0;
}

I tried executing the application with "waters" as argument to see what I'd come up with. Here is a print out from my windows cmd:
Checking...
waters
waster
rawest
Couldn't locate dict file!

Now this seems odd to me. Why would it open a new file stream three times and then not be able to find the file on the fourth attempt?
Everyone have been most helpful so far so I hoped there would be more words of wisdom to collect.

How many times do you try to open the file? Why would you try to open the file three times?

I tried initialising the stream in the main method just once before calling any functions and that results in only one anagram being printed. This caused me to draw the assumption that every time I run the dictionary file through (ie. call the checkDict function) and either find something or not I need to reinitialise the file stream in order for it to begin from the very top again.
Am I wrong in believing this?