1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
|
#include <stdio.h>
#include <stdlib.h>
/* ICU4C */
#include <unicode/utypes.h>
#include <unicode/ustring.h>
#include <unicode/ucnv.h>
#include <unicode/unorm2.h>
#include "util.h"
int main(int argc, char **argv)
{
int i;
UErrorCode err;
UConverter *uc = ucnv_open("UTF8", &err);
if (U_FAILURE(err)) return EXIT_FAILURE;
const UNormalizer2 *NFKC = unorm2_getNFKCInstance(&err);
if (U_FAILURE(err)) return EXIT_FAILURE;
for (i = 1; i < argc; ++i) {
if (argv[i][0] == '-') {
fprintf(stderr, "unrecognized option: %s\n", argv[i]);
return EXIT_FAILURE;
}
size_t len;
uint8_t *src = readfile(argv[i], &len);
if (!src) {
fprintf(stderr, "error reading %s\n", argv[i]);
return EXIT_FAILURE;
}
/* convert UTF8 data to ICU's UTF16 */
UChar *usrc = (UChar*) malloc(2*len * sizeof(UChar));
ucnv_toUChars(uc, usrc, 2*len, (char*) src, len, &err);
if (U_FAILURE(err)) return EXIT_FAILURE;
size_t ulen = u_strlen(usrc);
/* ICU's insane normalization API requires you to
know the size of the destination buffer in advance,
or alternatively to repeatly try normalizing and
double the buffer size until it succeeds. Here, I just
allocate a huge destination buffer to avoid the issue. */
UChar *udest = (UChar*) malloc(10*ulen * sizeof(UChar));
mytime start = gettime();
for (int i = 0; i < 100; ++i) {
unorm2_normalize(NFKC, usrc, ulen, udest, 10*ulen, &err);
if (U_FAILURE(err)) return EXIT_FAILURE;
}
printf("%s: %g\n", argv[i], elapsed(gettime(), start) / 100);
free(udest);
free(usrc);
free(src);
}
return EXIT_SUCCESS;
}
|
