Simple spell checker

I have a function in my spell checker codes to load words from a text file and store them in an array of strings.. Problem is, I can't figure out how to correct it.

For instance, the 'dictionary' text file has:

line1
line2
line3

and my code goes:

Code:

---

// assuming I only have 10 words, and done all other necessary declarations..
char *words[10];

for (i=0; fgets(s, 9, dict) != NULL; i++)
        words[i] = s;

---

How do I correct it so it doesn't store "line3" in every single element of my array?

you made array of pointers - you should change it to aray of strings

char words[10][MAX_LEN];
if you know the maximum length

and read into it like
fgets(words[i], sizeof words[i], dict)

If you do not know the maximum string size in advance - you will need a dynamic allocatoin and strcpy to move strings

It works! Thanks for your help!! :D

Quote:

---

Originally Posted by vart

you made array of pointers - you should change it to aray of strings

char words[10][MAX_LEN];
if you know the maximum length

and read into it like
fgets(words[i], sizeof words[i], dict)

If you do not know the maximum string size in advance - you will need a dynamic allocatoin and strcpy to move strings

---

Now it comes to another problem.. How do I remove the newline character (\n) from the words that was copied from fgets??

there are several approaches. I prefer:

Code:

---

char *p = strchr(buffer,'\n');
if(p) *p = '\0';

---

Hey buddy

can you please be more specific about what you want to do in your program so

that i can help you.

Also if you are reading text from TEXT file use fscanf() function to retrieve values

so that you can get values without new line char "\n"

and store this values directly in STRUCTURE so that comparison is easy.

Quote:

---

Originally Posted by sndpchikane

Also if you are reading text from TEXT file use fscanf() function to retrieve values

so that you can get values without new line char "\n"

---

Not necessary. Fgets works just as well. removing the newline is easy. Fscanf isn't "better", it's just another method of doing it.

Quote:

---

Originally Posted by sndpchikane

and store this values directly in STRUCTURE so that comparison is easy.

---

That depends on what you want to do. If there's no good reason to do it, you're just going to get unneccesary overhead.

Well, it really is a simple spell checker.. Read a text file containing 'dictionary' words, read another text file containing paragraphs of words, and print out if any of the words are not in the 'dictionary'.

I've done a spell checker in Java some time ago and it was slightly more complex than this (using data structures, and there were spelling suggestion feature as well), but I'm trying to improve on my C - which I only know the very basics.

Anyway, the way I did it was, I store the words of the 'dictionary' in an array, and then as I read each word in the file, i compare them to the elements in the array.

So far (with vart's help), I've got the dictionary loading done. Now I'm doing the strcmp part, which is giving me another headache as strcmp(token, words[i]) doesn't seem to give the correct results.

I tried printing out both token and words[i] out, and it seemed the same to me - but strcmp doesn't agree. I suspect it has something to do with the \0 character?

I tried using the debugger, and it seems that token and words[i] have really odd values.. like one being \"line3\\000\\217\\031\\177\" and the other \"line3\\r\\000\\324\\340\\310\" when both of them prints "line3".

Do you know that strcmp returns 0 when two strings are equal?

Code:

---

    char *a = "gorilla";
    if( strcmp(a,"gorilla") == 0 )
        printf("KING KONG vs GODZILLA\nFIGHT!\n");

---

The above actually prints a Mortal Monster Combat intro message... :)

Do you have a documentation? Docs explicitly specify that strcmp returns 0 when strings match, as vart points out.
As for the strings... well, I'm guessing you never initialized your array, hence it contains "garbage." But all strings are expected to end when a '\0' (or 0) is encountered. Thus it never reads any of that junk beyond the 0, so you're fine. Nothing wrong with it.

Oh dear, I'm really sorry I tend to skip a lot of details when I post my question..

My actual part of the code that does (or is supposed to do) the string compare is:

Code:

---

/* this whole chunk of codes are in a while-loop that reads each token in the file */

for (i=0; i<strlen(*words); i++) {
    if (strcmp(token, words[i]) == 0) {
        /* just an int variable to keep track on whether the word is found or not,
            i initialized it to 1 */
        notFound = 0;
        printf("Here!");
        break;
    }
}
           
if (notFound) {
    printf("%s\n", token);
}

---

I've put printf's around to print the values of token and words[i] in each loop and it printed out as expected (eg. token0, word[0].. token0, word[1].. token0, word[2]..... token1, word[0].. and so on), but the strcmp(token, words[i]) never gave me a 0 despite, say, token is "string1" and words[i] is also "string1".

Did you remove the newline character as vart suggested?

Quote:

---

Originally Posted by xuftugulus

Did you remove the newline character as vart suggested?

---

Yup, I did, and the 'dictionary' array (i.e. words[]) prints out fine when i do a simple loop to print out all the elements..

Code:

---

for (i=0; i<strlen(*words); i++) {
    if (strcmp(token, words[i]) == 0) {
        /* just an int variable to keep track on whether the word is found or not,
            i initialized it to 1 */
        notFound = 0;
        printf("Here!");
        break;
    }
}

---

I'm pretty sure strlen(*words) is incorrect here. What do you intend that to do?

--
Mats

Quote:

---

Originally Posted by matsp

I'm pretty sure strlen(*words) is incorrect here. What do you intend that to do?

--
Mats

---

Umm, to specify the boundary for the for-loop.. The size of the 'dictionary', i.e, char word[][]. Initially I put for (i=0; i<strlen(words); i++) (same thing, only without the asterisk), but the compiler gave me an error without the asterisk. The for-loop works fine though (as far as I know), I'm only having problem with the string compare.

Deosn't work that way. You want the size of the array, not the length of the string, so you can't rely on strlen here.

Quote:

---

Originally Posted by Elysia

Deosn't work that way. You want the size of the array, not the length of the string, so you can't rely on strlen here.

---

I declared the array as char words[i][j]; (i and j are arbitrary numbers for me to test files that contain only a few lines - here I put char words[10][10];)

I tried using sizeof(words), but it gave me a value of 100..

To get the size of an array, you can usually do:
sizeof(array) / sizeof(array[0]);

Quote:

---

To get the size of an array, you can usually do:
sizeof(array) / sizeof(array[0]);

---

But you have to do it in the function where the array is declared.
In any other funtion - you will need to pass an array size as a parameter

Thanks, I'll try to get the array size problem fixed.

But back to my other question, why would printf("%s, %s\n", token, words[i]); output "string1, string1", but when I use strcmp(token, words[i]) == 0 it won't give a true?

print each of the words[i] characters as decimal or hex to see if there is any invisible characters that "get in the way".

--
Mats

Quote:

---

Originally Posted by matsp

print each of the words[i] characters as decimal or hex to see if there is any invisible characters that "get in the way".

--
Mats

---

do you mind explaining how it's supposed to be done? i'm really lost at this. not typecasting is it?

Well since this won't help you get any marks (if this is an assignment)

Code:

---

void print_string_as_hex(const char * s)
{
    size_t  len = strlen(s),
            i = 0;
   
    for(i = 0; i < len; i++)
    {
        printf("%X", s[i]);
    }
   
    return;
}

---

I didn't test it, hope it works :)

Read it and understand it before using it.

In my opinion you're jumping onto the 'code train' too early, design it first -- for example:

• open the dictionary file
• read each line (word), trim the newline character if nessisary -- adding the word to a linked list or array (sizing the array as nessisary with realloc() :))
• close the file
• do whatever with the array or linked list of words

Quote:

---

would printf("&#37;s, %s\n", token, words[i]); output "string1, string1",

---

try something like

Code:

---

printf("\"%s\", \"%s\"\n", token, words[i]);

---

Quote:

---

Originally Posted by zacs7

Well since this won't help you get any marks (if this is an assignment)

Code:

---

void print_string_as_hex(const char * s)
{
    size_t  len = strlen(s),
            i = 0;
   
    for(i = 0; i < len; i++)
    {
        printf("%X", s[i]);
    }
   
    return;
}

---

I didn't test it, hope it works :)

Read it and understand it before using it.

In my opinion you're jumping onto the 'code train' too early, design it first -- for example:

• open the dictionary file
• read each line (word), trim the newline character if nessisary -- adding the word to a linked list or array (sizing the array as nessisary with realloc() :))
• close the file
• do whatever with the array or linked list of words

---

It is an assignment, but I'm not looking for the solution - i just need help with particular parts that I seem to have problems with. :)

About the design, I do have a sketch of the flowchart before I started coding, and I'm currently at the "do whatever with the array or linked list of words" - which is where I'm now stuck.

Just so no one gets the impression that I'm expected the entire solution here, here's what I've completed so far in the coding:

• check whether the command line arguments are valid; exit properly if not
• check if the provided dictionary filename exists
• check if the provided filename (of the file to be checked) exists
• open the dictionary
• read the dictionary words into an array
• close the dictionary file
• open the file to be checked
• read the each words in the file
• compare the words to the 'dictionary' array <<-- problem here
• print out the words that are not found in the dictionary
• close the file

I really appreciate all the help I get from you guys here, but yeah, I'm trying to do my own homework. :D

Quote:

---

Originally Posted by vart

try something like

Code:

---

printf("\"%s\", \"%s\"\n", token, words[i]);

---

---

ehh.. i tried that and it printed out "string1", "string1 (without the last double quote). where did it go?? is that the problem? :confused:

Quote:

---

Originally Posted by purplechirin

ehh.. i tried that and it printed out "string1", "string1 (without the last double quote). where did it go?? is that the problem? :confused:

---

I suppose it goes on the next line indicating that the second string contains \n at the end

Quote:

---

Originally Posted by vart

I suppose it goes on the next line indicating that the second string contains \n at the end

---

the double quote somewhat disappeared.. there's no extra " on the second line.

edit: i changed it a little, and instead of a double quote i put different symbols:

Code:

---

printf("!%s@, #%s$\n", token, words[i]);

---

and instead of printing out !string1@, #string2$, it came out:

Code:

---

$string1@, #string1
$string1@, #string2
$string1@, #string3
...

---

:confused:

That indicates that there is a '\r' [carriage return] at the end of the line - can you post:
1. Code that opens the "valid word list".
2. Code that reads the word.
3. Code that removes newline.

I suspect you are reading the file in binary mode, but it could be other things.

Using the "print in hex" variation will show you that it's got a 0D character at the end of the string, I suspect.

--
Mats

Quote:

---

Originally Posted by matsp

That indicates that there is a '\r' [carriage return] at the end of the line - can you post:
1. Code that opens the "valid word list".
2. Code that reads the word.
3. Code that removes newline.

I suspect you are reading the file in binary mode, but it could be other things.

Using the "print in hex" variation will show you that it's got a 0D character at the end of the string, I suspect.

--
Mats

---

While i do the printing in hex.. here are the codes:

Code that opens the 'valid word list':

Code:

---

FILE *dict;
dict = fopen(filename, "r");

---

Code that reads the word:

Code:

---

for (i=0; fgets(words[i], sizeof(words[i]), dict) != NULL; i++);

---

Code that removes newline:

Code:

---

/* remove the newline characters from each word */
for (i=0; i<sizeof(words)/sizeof(words[0]); i++) {
    char *p = strchr(words[i],'\n');
    if(p) {
        *p = '\0';
    }
}

---

EDIT: I did the printing in hex for the words[], and 0D was at the end of the last element in the array -

6C696E6531D <-- word[0]
6C696E6532D <-- word[1]
6C696E6533D <-- word[2]
...
6C696E653130D <-- word[9]

(in my 'dictionary' of 10 words)

Strange. Are you by any chance using a word-list generated by a Windows / DOS program on a Linux/Unix machine? That would explain the newline/carriage return problem.

If so, you should use "dos2unix wordlist" to make sure the newlines are converted from CR+LF to LF only.

--
Mats

Quote:

---

Originally Posted by matsp

Strange. Are you by any chance using a word-list generated by a Windows / DOS program on a Linux/Unix machine? That would explain the newline/carriage return problem.

If so, you should use "dos2unix wordlist" to make sure the newlines are converted from CR+LF to LF only.

--
Mats

---

Well, on my first attempt I was using Netbeans to create an empty text file which i added 10 words in for testing, and then on the second attempt I used the word list generated on unix.. Same results for both.. :(

Sorry, but I don't quite follow. Are you on a linux/unix system? If so, you could use "od -x wordlist" to list it in hex format. If you have ANY "0d" characters in the file, then you have a file with DOS/Windows style newlines, which is not what linux/unix expects, so you will need to strip them.

Modifying your strip functionality like this would do:

Code:

---

for (i=0; i<sizeof(words)/sizeof(words[0]); i++) {
    char *p = strchr(words[i],'\n');
    char *q = strchr(words[i],'\r');
    if (q && q < p) p = q;
    if(p) {
        *p = '\0';
    }
}

---

--
Mats