Substitute an instruction depending on a condition - c++

Question

I have two for loops that I want to write in a function as one. The problem is that it differ only in one instruction
for (int i = 1; i <= fin_cabecera - 1 ; i++ ){
buffer[i] &= 0xfe;
if (bitsLetraRestantes < 0) {
bitsLetraRestantes = 7;
mask = 0x80;
letra = sms[++indiceLetra]; //*differs here*
}
char c = (letra & mask) >> bitsLetraRestantes--;
mask >>= 1;
buffer[i] ^= c;
}
And the other
for (int i = datos_fichero; i <= tamanio_en_bits + datos_fichero; i++){
buffer[i] &= 0xfe;
if (bitsLetraRestantes < 0) {
bitsLetraRestantes = 7;
mask = 0x80;
f.read(&letra, 1); //*differs here*
}
char c = (letra & mask) >> bitsLetraRestantes--;
mask >>= 1;
buffer[i] ^= c;
}
I thought in something like this:
void write_bit_by_bit(unsigned char buffer[], int from, int to, bool type) {
for (int i = to; i <= from; i++) {
buffer[i] &= 0xfe;
if (bitsLetraRestantes < 0) {
bitsLetraRestantes = 7;
mask = 0x80;
type ? (letra = sms[++indiceLetra]) : f.read(&letra, 1);
}
char c = (letra & mask) >> bitsLetraRestantes--;
mask >>= 1;
buffer[i] ^= c;
}
}
But I think there has to be a better method.
Context:
I will give more context (I will try explain it as better as I can within my language limitations). I have to read one byte each time because The Buffer variable represents a image pixel. sms is a message that have to be hidden within the image, and letra is a single char of that message. In order to not modify the aspect of the image, each bit of each character have to be written in the last bit of each pixel. Let me give you and example.
letra = 'H' // 01001000 in binary
buffer[0] = 255 // white pixel 11111111
In order to hide the H char, I will need 8 pixel:
The result will be like:
buffer[0] //11111110,
buffer[1] //11111111
buffer[2] //11111110
buffer[3] //11111110
buffer[4] //11111111
buffer[5] //11111110
buffer[6]//11111110
buffer[7]//11111110
The H is hidden in the last bit of the image. I hope I explained well.
[Solution]
Thanks to #anatolyg I've rewrited the code and now works just as I wanted. Here is how it looks:
void write_bit_by_bit(unsigned char buffer[], ifstream& f,int from, int to, char sms[], bool type){
unsigned short int indiceLetra = 0;
short int bitsLetraRestantes = 7;
unsigned char mask = 0x80; //Empezamos por el bit más significativo (10000000)
char* file_buffer;
if(type){ //Write file data
int number_of_bytes_to_read = get_file_size(f);
file_buffer = new char[number_of_bytes_to_read];
f.read(file_buffer, number_of_bytes_to_read);
}
const char* place_to_get_stuff_from = type ? file_buffer : sms;
char letra = place_to_get_stuff_from[0];
for (int i = from; i <= to; i++) {
buffer[i] &= 0xfe; //hacemos 0 último bit con máscara 11111110
//TODO: Hacer con dos for
if (bitsLetraRestantes < 0) {
bitsLetraRestantes = 7;
mask = 0x80;
letra = place_to_get_stuff_from[++indiceLetra];//letra = sms[++indiceLetra];
}
char c = (letra & mask) >> bitsLetraRestantes--;
mask >>= 1;
buffer[i] ^= c; //Almacenamos en el ultimo bit del pixel el valor del caracter
}
}
int ocultar(unsigned char buffer[],int tamImage, char sms[], int tamSms){
ifstream f(sms);
if (f) {
strcpy(sms,basename(sms));
buffer[0] = 0xff;
int fin_cabecera = strlen(sms)*8 + 1;
buffer[fin_cabecera] = 0xff;
write_bit_by_bit(buffer, f, 1, fin_cabecera -1, sms, WRITE_FILE_NAME);
int tamanio_en_bits = get_file_size(f) * 8;
int datos_fichero = fin_cabecera + 1;
write_bit_by_bit(buffer, f, datos_fichero, tamanio_en_bits + datos_fichero, sms, WRITE_FILE_DATA);
unsigned char fin_contenido = 0xff;
short int bitsLetraRestantes = 7;
unsigned char mask = 0x80;
for (int i = tamanio_en_bits + datos_fichero + 1;
i < tamanio_en_bits + datos_fichero + 1 + 8; i++) {
buffer[i] &= 0xfe;
char c = (fin_contenido & mask) >> bitsLetraRestantes--;
mask >>= 1;
buffer[i] ^= c;
}
}
return 0;
}

Answer 1

Since you are talking about optimization here, consider performing the read outside the loop. This will be a major optimization (reading 10 bytes at once must be quicker than reading 1 byte 10 times). This will require an additional buffer for (the file?) f.
if (!type)
{
char f_buffer[ENOUGH_SPACE];
number = calc_number_of_bytes_to_read();
f.read(f_buffer, number);
}
for (...) {
// your code
}
After you have done this, your original question is easy to answer:
const char* place_to_get_stuff_from = type ? sms : f_buffer;
for (...) {
...
letra = place_to_get_stuff_from[++indiceLetra];
...
}