How to access a char array and change lower case letters to upper case, and vice versa

I'm currently working on a class project for Structured Computer Organization using an x86 processor. The value that I am accessing is an 1 byte char, but I do not know how to compare it to an uppercase. They said to use an ASCII table of the hex format, but I'm not sure how to even compare the two.

void changeCase (char char_array[], int array_size ) {
    __asm {
            // BEGIN YOUR CODE HERE
 
        mov eax, char_array;        //eax is base image
        mov edi, 0;
        
    readArray:
        cmp edi, array_size;
        jge  exit;
        mov ebx, edi;           //using ebx as offset
        shl ebx, 2;
        mov cl, [eax + ebx];    //using ecx to be the storage register
    
    check:
        //working on it
        cmp cl, 0x41;       //check if cl is <= than ASCII value 65 (A)
        jl next_indx;
        cmp cl, 0x7A;       //check if cl is >= than ASCII value 122 (z)
        jg next_indx;
        cmp cl, 'a';
        jl convert_down;
        jge convert_up;
        

    convert_down:
        or cl, 0x20;        //make it lowercase
        jmp write;

    convert_up:
        and cl, 0x20;       //make it uppercase
        jmp write;

    write:
        mov byte ptr [eax + ebx], cl    //slight funky town issue here,

    next_indx:
        inc edi;

    exit:
        cmp edi, array_size;
        jl readArray;

    mov char_array, eax;
            // END YOUR CODE HERE
    }
}

Anything helps at this point. Thank you in advance for the help!

edit 1:

Thanks for all the suggestion and points of clarity, edited my code to reflect change. Some problem with access violation now.

edit 2 (+):

Thanks for the helpful eyes people. I'm still getting to translating all letters now.

Solution

For clarity's sake, I'll just use pure assembly and assume that...

char_array is a 32-bit pointer at [ebp+8].
array_size is a two's complement 32-bit number at [ebp+12].
For your platform (it is this way for most anyway), char's encoding is ASCII.

You should be able to deduce this yourself into inline assembly. Now, if you look at the table everyone is supposed to remember but barely anyone does, you'll notice some important details...

Uppercase letters A through Z map into codes 0x41 through 0x5A, respectively.
Lowercase letters a through z map into codes 0x61 through 0x7A, respectively.
Everything else is not a letter, and thus need no case conversion.
If you look at the binary representation of the upper and lowercase letter ranges, you'll notice that they are exactly the same, with the sole exception that uppercase letters have bit 6 cleared, and lowercase ones have it set.

As a result, the algorithm would be...

while array_size != 0
    byte = *char_array
    if byte >= 0x41 and byte <= 0x5A
        *char_array |= 0x20 // Turn it lowercase
    else if byte >= 0x61 and byte <= 0x7A
        *char_array &= 0xDF // Turn it uppercase
    array_size -= 1
    char_array += 1

Now, let's translate this into assembly...

mov eax, [ebp+8]      # char *eax = char_array
mov ecx, [ebp+12]     # int ecx = array_size

.loop:
    or ecx, ecx       # Compare ecx against itself
    jz .end_loop      # If ecx (array_size) is zero, we're done
    mov dl, [eax]     # Otherwise, store the byte at *eax (*char_array) into `char dl`
    cmp dl, 'A'       # Compare dl (*char_array) against 'A' (lower bound of uppercase letters)
    jb .continue      # If dl` (*char_array) is lesser than `A`, continue the loop
    cmp dl, 'Z'       # Compare dl (*char_array) against 'Z' (upper bound of uppercase letters)
    jbe .is_uppercase # If dl (*char_array) is lesser or equal to 'Z', then jump to .is_uppercase
    cmp dl, 'a'       # Compare dl (*char_array) against 'a' (lower bound of lowercase letters)
    jb .continue      # If dl (*char_array) is lesser than 'a', continue the loop
    cmp dl, 'z'       # Compare dl (*char_array) against 'z' (upper bound of lowercase letters)
    jbe .is_lowercase # If dl (*char_array) is lesser or equal to 'z', then jump to .is_lowercase
    jmp .continue     # All tests failed, so continue the loop

    .is_uppercase:
        or dl, 20h    # Set the 6th bit
        mov [eax], dl # Send the byte back to where it came from
        jmp .continue # Continue the loop

    .is_lowercase:
        and dl, DFh   # Clear the 6th bit
        mov [eax], dl # Send the byte back to where it came from
        jmp .continue # Continue the loop

    .continue:
        inc eax       # Increment `eax` (`char_array`), much of like a pointer increment
        dec ecx       # Decrement `ecx` (`array_size`), so as to match the previous pointer increment
        jmp .loop     # Continue

.end_loop:

Once code reaches .end_loop, you're done.

I hope this has led a light on you!