Notes

#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <stdbool.h>
#include <math.h>


const char matrix_bigram_strings[10][3] ={"th", "he", "in", "er", "an", "en", "ch", "de", "ei", "te"};

const char matrix_trigram_strings[10][4]={"the", "and", "ing", "ent", "ion", "der", "sch", "ich", "nde", "die"};

const char languages[2][8]={"english", "german"};

//Ýngilizce dilinin frekans deðerleri
const float frequency_eng[20]={ 2.71, 2.33, 2.03, 1.78, 1.61, 1.13, 0.01, 0.01, 0.01, 0.01, 1.81, 0.73, 0.72, 0.42, 0.42, 0.01, 0.01, 0.01, 0.01, 0.01 };
//Almanca dilinin frekans deðerleri
const float frequency_germ[20]={ 0.01, 0.89, 1.71, 3.90, 1.07, 3.61, 2.36, 2.31, 1.98, 1.98, 0.01, 0.01, 0.01, 0.01, 0.01, 1.04, 0.76, 0.75, 0.72, 0.62 };


float calculated_frequencies[20];
float distances [2]={0,0};


void filter_str(char str[]);
void calculate_frequencies_bi(char str[]);
void calculate_frequencies_tri(char str[]);
void calculate_distances();
void detect_lang();

int main()
{
char metin[100];
printf("Metin giriniz: ");
gets(metin);

filter_str(metin);
calculate_frequencies_bi(metin);
calculate_frequencies_tri(metin);
calculate_distances();
detect_lang();

return 0;
}

void filter_str(char str[]){

/* Metnin içerisindeki her karakter taranýr ve ASCII (decimal) numarasý
[A(65)-Z(90)], [a(97)-z(122)] ve 32 (boþluk karakteri) olmayan karakterler
boþluk karakterine çevrilir. */
for (int i = 0; i<strlen(str); i++){
if ((str[i] >= 65 && str[i] <= 90) || (str[i] >= 97 && str[i] <= 122) || (str[i] == 32))
continue;
else
str[i] = " ";
}
}

void calculate_frequencies_bi(char str[]){

int a = 0;
float harf_sayisi = 0;
int bigram_sayisi[10] = {0, 0, 0, 0, 0, 0, 0, 0, 0, 0};

while (a < 10) {
for (int i = 0; i < strlen(str); i++) {
if ((str[i] == matrix_bigram_strings[a][0]) && (str[++i] == matrix_bigram_strings[a][1]))
bigram_sayisi[a]++;
else
continue;
}

a++;

}

for (int i = 0; i < strlen(str); i++){
if (str[i] != 32)
harf_sayisi++;
else
continue;
}

for (int i = 0; i < 10; i++){
calculated_frequencies[i] = (bigram_sayisi[i] / (harf_sayisi / 2)) * 100;
}

printf("Bigram frekanslari: ");
for (int i = 0; i < 10; i++) {
printf("%s(%.2f) ", matrix_bigram_strings[i], calculated_frequencies[i]);
}
}

void calculate_frequencies_tri(char str[]){

int a = 0;
float harf_sayisi = 0;
int trigram_sayisi[10] = {0, 0, 0, 0, 0, 0, 0, 0, 0, 0};

while (a < 10) {
for (int i = 0; i < strlen(str); i++) {
if ((str[i] == matrix_trigram_strings[a][0]) && (str[i+1] == matrix_trigram_strings[a][1]) && (str[i+2] == matrix_trigram_strings[a][2]))
trigram_sayisi[a]++;
else
continue;

}

a++;

}

for (int i = 0; i < strlen(str); i++){
if (str[i] != 32)
harf_sayisi++;
else
continue;
}

for (int i = 10; i < 20; i++){
calculated_frequencies[i] = (trigram_sayisi[i-10] / (harf_sayisi / 3)) * 100;
}

printf("nTrigram frekanslari: ");
for (int i = 0; i < 10; i++) {
printf("%s(%.2f) ", matrix_trigram_strings[i], calculated_frequencies[i+10]);
}
}

void calculate_distances(){

int oklit_eng, oklit_germ = 0;

for (int i = 0; i < 20; i++) {
oklit_eng += pow(calculated_frequencies[i] - frequency_eng[i], 2);
oklit_germ += pow(calculated_frequencies[i] - frequency_germ[i], 2);
}

oklit_eng = sqrt(oklit_eng);
oklit_germ = sqrt(oklit_germ);

distances[0] = oklit_eng;
distances[1] = oklit_germ;
}

void detect_lang(){

if (distances[0] < distances[1])
printf("nGirilen metin Ingilizce'dir.(%s)", languages[0]);
else
printf("nGirilen metin Almanca'dir.(%s)", languages[1]);

}