MayRiv MayRiv - 3 months ago 32
Linux Question

Using iconv translit to convert from UTF-8 to CP1251

I'am trying to convert string "

" from UTF-8 to CP1251 via C++ library
iconv.h
using TRANSLIT and as a result I get string "
a?
", when expecting "
aU
".

When I use php script
<?php echo iconv("UTF-8", "Windows-1251//TRANSLIT", "Ü");>
on this computer, I get "
aU
" string as result.

Here's the code:

#include <cstdlib>
#include <iconv.h>
#include <locale.h>
#include <stdio.h>
#include <string.h>
#include <errno.h>
using namespace std;

int IConvert(char *buf,char *outbuf, size_t len, const char *from, const char *to)
{
iconv_t iconv_cd;
if ((iconv_cd = iconv_open(to, from)) == (iconv_t) -1) {
printf("Cannot open iconv from %s to %s\n", from, to);
return 0;
}

char *inbuf = buf;
size_t inlen = len;
size_t outlen = len;
size_t res = 0;

while (inlen > 0 && outlen > 0) {
res = iconv(iconv_cd, &inbuf, &inlen, &outbuf, &outlen);
if (res == 0)
break;

if (res == (size_t) (-1)) {
if (errno != EILSEQ && errno != EINVAL) {
iconv_close(iconv_cd);
*outbuf = '\0';
printf("Erorr %s (%s)\n", strerror(errno), from);

return 0;
} else if (inbuf < outbuf) {
iconv_close(iconv_cd);
*outbuf = '\0';
printf("Erorr %s (inbuf < outbuf)\n", strerror(errno));

return 0;
}
}
if (inlen > 0 && outlen > 0) {
*outbuf++ = *inbuf++;
inlen--;
outlen--;
}
}
iconv_close(iconv_cd);
*outbuf = '\0';
return 1;
}
int main(int argc, char** argv) {
char* line = "\u00C0a\u00DC";
char* from = (char*) malloc(strlen(from)+1);
char* to = (char*) malloc(strlen(from)+1);
strcpy(from, line);
printf("%s\n", from);
IConvert(from, to, strlen(from)+1, "UTF-8", "CP1251//TRANSLIT");
printf("%s\n", to);
return 0;
}


Any idea what problem could be?

Answer

The solution is

setlocale(LC_ALL, "");

at the beginning of your program. Yes, the locale influences the transliteration. In a German locale Ü would be transliterated to UE not U.