Reading all .txt files from a folder?

Question

I am trying to read all .txt files which is in the directory named "dataset". All text files has name like 1.txt, 2.txt, 3.txt... And then saving the contents of the files to a structure named FILES.

I used the dirent.h library and readdir( ) function as I saw in some sources. But the file name which program read from directory does not return correctly. Here is my related code:

#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <dirent.h>

typedef struct FILES{
    char **words;
    int wordCount;
}FILES;

void readFiles();

FILES *files;
int fileCount;

int main(){
    readFiles();

    return 0;
}

void readFiles(){
    FILE *file;
    DIR *directory;
    struct dirent *filesInDirectory;
    int counter;

    fileCount = 1;
    files = (FILES *)malloc(sizeof(FILES));
    directory = opendir("dataset");
    if(directory == NULL){
        printf("Warning: The directory name that is given in the code is not 
valid ..!");
        return;
    }else{
        while((filesInDirectory = readdir(directory)) != NULL){
            printf("%s\n", filesInDirectory->d_name);
            file = fopen(filesInDirectory->d_name, "r+");
            if(file == NULL){
                printf("Warning: The file named %s could not open ..!", 
filesInDirectory->d_name);
                return;
            }
            files[fileCount-1].wordCount = 1;
            files[fileCount-1].words = (char **)malloc(files[fileCount-
1].wordCount * sizeof(char *));
            counter = 0;

            while(!feof(file)){
                files[fileCount-1].words[counter] = (char *)malloc(20 * 
sizeof(char));
                fscanf(file, "%s", files[fileCount-1].words[counter]);
                files[fileCount-1].wordCount++;
                files[fileCount-1].words = (char **)realloc(files[fileCount-
1].words, files[fileCount-1].wordCount * sizeof(char *));
                counter++;
            }
            fileCount++;            
            fclose(file);
        }
    }

}

The file name that I printed in here "printf("%s\n", filesInDirectory->d_name);" is ".". Where am I doing wrong?

`while(!feof(file)){` -- This is wrong - see https://stackoverflow.com/questions/5431941/why-is-while-feof-file-always-wrong — Ed Heal, Dec 25 '17 at 16:59
@EdHeal Yes, thank you for this. I will keep that in my mind. But do you have any idea why the file name does not returns correctly? — JollyRoger, Dec 25 '17 at 17:01
`readdir` will return directories - First one being `.` and the second `..` — Ed Heal, Dec 25 '17 at 17:03
Also casting `malloc` is bad - See https://stackoverflow.com/questions/605845/do-i-cast-the-result-of-malloc — Ed Heal, Dec 25 '17 at 17:04
@EdHeal I could not notice it. What can I use instead readdir to read .txt files or at least taking the names of its. — JollyRoger, Dec 25 '17 at 17:05
Check the returned filenames and check that they end in `.txt` — Ed Heal, Dec 25 '17 at 17:06
@EdHeal I am sorry, could not understand. Returned from which function? I did not know how to read files or at least file names in a directory. — JollyRoger, Dec 25 '17 at 17:10
Check that `filesInDirectory->d_name` ends in `.txt`! Use `strlen` and `strcmp` — Ed Heal, Dec 25 '17 at 17:12

score 0 · Accepted Answer · answered Dec 25 '17 at 17:21

THIS IS NOT AN ANSWER BUT TOO BIG FOR A COMMENT

Here are your problems:

1.

 files = (FILES *)malloc(sizeof(FILES));

This is enough size of one FILES. This is will not be enough

After

while((filesInDirectory = readdir(directory)) != NULL){

Put something like

size_t len = strlen(filesInDirectory->d_name);
if (len < 5 || strcmp(filesInDirectory->d_name + len - 4, ".txt") != 0) {
  continue;
}

This will check to ensure files end in .txt

Consider using fstat to ensure that the file is a text file
Remove the casts on malloc
Perhaps use realloc for files
while ... feof is bad - see link above
fscanf(file, "%s", - Buffer overruns is possible. Do something about this. Read the manual page

Reading all .txt files from a folder?

1 Answers1