icu-test.c

/** gcc -g -O0 -fno-inline -o test test.c `pkg-config --libs --cflags icu-uc icu-io` */

// We will be using UTF8
#define U_CHARSET_IS_UTF8 1

#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include "unicode/uchar.h"
#include "unicode/ucnv.h"
#include "unicode/utypes.h"
#include "unicode/utf8.h"
#include "unicode/ustdio.h"
#include "unicode/unorm.h"

char *test = "I - Björk Guðmundsdóttir - abcçdefgğhıijklmnoöprsştuüvyz - \xd1\x82\xe0\xb8\x81\xe0\xb8\xa7";
UConverter *converter;
UFILE *out;


/**
 *
 */
void utf8_convert_from_char(char *src, UChar **dst, uint32_t *dst_len) {
    UErrorCode status = U_ZERO_ERROR;

    if (converter == NULL) {
        printf("Default Converter name: %s\n", ucnv_getDefaultName());
        converter = ucnv_open("UTF-8", &status);        // Use default converter
        if (U_FAILURE(status)) {
            printf("Cannot create converter: %d\n", status);
            exit(1);
        }
    }

    *dst = (UChar *)malloc((strlen(src)+1) * sizeof(UChar));
    *dst_len = ucnv_toUChars(converter, *dst, strlen(src)+1, src, strlen(src), &status);
}


/**
 *
 */
int utf8_tolower(UChar *src, int32_t src_len, UChar **dst, int32_t *dst_len, const char *locale) {
    UErrorCode status = U_ZERO_ERROR;
    *dst = (UChar *)malloc((src_len+1) * sizeof(UChar));
    *dst_len= u_strToLower(*dst, src_len+1, src, src_len, locale, &status);
    return 0;
}


/**
 *
 */
int utf8_toupper(UChar *src, int32_t src_len, UChar **dst, int32_t *dst_len, const char *locale) {
    UErrorCode status = U_ZERO_ERROR;
    *dst = (UChar *)malloc((src_len+1) * sizeof(UChar));
    *dst_len= u_strToUpper(*dst, src_len+1, src, src_len, locale, &status);
    return 0;
}


/**
 *
 */
int utf8_reverse(UChar *src, int32_t src_len, UChar **dst, int32_t *dst_len) {
    *dst = (UChar *)malloc((src_len+1) * sizeof(UChar));
    *dst_len = src_len;

    int i;
    for (i=0; i!=src_len; i++) {
        (*dst)[i] = src[src_len - i - 1];
    }
    return 0;
}


/**
 *
 */
int32_t utf8_strcasecmp(UChar *src, uint32_t src_len, UChar *dst, uint32_t dst_len) {
    UErrorCode status = U_ZERO_ERROR;
    int32_t i = u_strCaseCompare(src, src_len, dst, dst_len, 0, &status);
    return i;
}


/**
 *
 */
int32_t utf8_strcmp(UChar *src, uint32_t src_len, UChar *dst, uint32_t dst_len) {
    UErrorCode status = U_ZERO_ERROR;
    return u_strCompare(src, src_len, dst, dst_len, 0, &status);
}


/**
 *
 */
void main(int argc, char *argv[]) {
    out = u_finit(stdout, NULL, NULL);

    UChar *dst;
    int32_t dst_len;
    utf8_convert_from_char(test, &dst, &dst_len);
    u_fprintf(out, "(%d) %S\n", dst_len, dst);

    UChar *dst2;
    int32_t dst2_len;
    utf8_tolower(dst, dst_len, &dst2, &dst2_len, NULL);
    u_fprintf(out, "(%d) %S\n", dst2_len, dst2);

    UChar *dst3;
    int32_t dst3_len;
    utf8_tolower(dst, dst_len, &dst3, &dst3_len, "tr_TR");
    u_fprintf(out, "(%d) %S\n", dst3_len, dst3);

    UChar *dst7;
    int32_t dst7_len;
    utf8_toupper(dst, dst_len, &dst7, &dst7_len, NULL);
    u_fprintf(out, "(%d) %S\n", dst7_len, dst7);


    UChar *dst4;
    int32_t dst4_len;
    utf8_reverse(dst, dst_len, &dst4, &dst4_len);
    u_fprintf(out, "(%d) %S\n", dst4_len, dst4);

    UChar *dst5, *dst6;
    int32_t dst5_len, dst6_len;
    utf8_convert_from_char("Björk Guðmundsdóttir", &dst5, &dst5_len);
    utf8_convert_from_char("BJÖRK GUÐMUNDSDÓTTIR", &dst6, &dst6_len);

    u_fprintf(out, "(%d) %S\n", dst5_len, dst5);
    u_fprintf(out, "(%d) %S\n", dst6_len, dst6);

    int32_t i;
    i = utf8_strcasecmp(dst6, dst6_len, dst5, dst5_len);
    printf("I1: %d\n", i);
    i = utf8_strcasecmp(dst5, dst5_len, dst6, dst6_len);
    printf("I2: %d\n", i);

    i = utf8_strcmp(dst6, dst6_len, dst5, dst5_len);
    printf("I3: %d\n", i);
    i = utf8_strcmp(dst5, dst5_len, dst6, dst6_len);
    printf("I4: %d\n", i);

}

Output:

Default Converter name: UTF-8
(62) I - Björk Guðmundsdóttir - abcçdefgğhıijklmnoöprsştuüvyz - тกว
(62) i - björk guðmundsdóttir - abcçdefgğhıijklmnoöprsştuüvyz - тกว
(62) ı - björk guðmundsdóttir - abcçdefgğhıijklmnoöprsştuüvyz - тกว
(62) I - BJÖRK GUÐMUNDSDÓTTIR - ABCÇDEFGĞHIIJKLMNOÖPRSŞTUÜVYZ - Тกว
(62) วกт - zyvüutşsrpöonmlkjiıhğgfedçcba - rittódsdnumðuG kröjB - I
(20) Björk Guðmundsdóttir
(20) BJÖRK GUÐMUNDSDÓTTIR
I1: 0
I2: 0
I3: -32
I4: 32