int elh_iterate_words__xany__UTF8(const char *buf, size_t bufsize, lh_word_check_t *check, lh_word_callback_t *callback, void *userdata)
{
unichar check_buf[WORD_MAX_CHARS];
int prop;
const char *curr = buf;
const char *buf_end = buf+bufsize;
const char *word_begin = curr;
const char *word_end;
unichar uchr;
size_t word_length;
while (curr < buf_end)
{
word_begin = curr;
uchr = eh_decode_char__UTF8 (&curr, buf_end);
prop = unichar_getprops (uchr);
if (prop & UCP_ALPHA)
{
check_buf[0] = uchr;
word_length = 1;
for(;;)
{
word_end = curr;
uchr = eh_decode_char__UTF8 (&curr, buf_end);
if (uchr < 0)
{
if ((UNICHAR_NO_DATA == uchr) || (UNICHAR_BAD_ENCODING == uchr))
return uchr;
if (UNICHAR_EOD == uchr)
break;
}
prop = unichar_getprops (uchr);
if (!(prop & UCP_ALPHA))
break;
if (WORD_MAX_CHARS > word_length)
check_buf[word_length] = uchr;
word_length++;
}
if (WORD_MAX_CHARS < word_length)
goto done_word;
if (NULL!=check && 0 == check (check_buf, word_length))
goto done_word;
callback ((utf8char *)(word_begin), word_end-word_begin, userdata);
done_word:
if (prop & UCP_IDEO)
goto proc_ideo;
continue;
}
if (prop & UCP_IDEO)
{
proc_ideo:
check_buf[0] = uchr;
if (NULL!=check && 0 == check (check_buf, 1))
continue;
callback ((utf8char *)(word_begin), curr-word_begin, userdata);
continue;
}
if ((uchr < 0) && ((UNICHAR_NO_DATA == uchr) || (UNICHAR_BAD_ENCODING == uchr)))
return uchr;
}
return 0;
}
int elh_iterate_patched_words__xany__UTF8(const char *buf, size_t bufsize, lh_word_check_t *check, lh_word_patch_t *patch, lh_word_callback_t *callback, void *userdata)
{
unichar check_buf[WORD_MAX_CHARS];
int prop;
const char *curr = buf;
const char *buf_end = buf+bufsize;
const char *word_begin = curr;
const char *word_end = NULL;
unichar uchr;
size_t word_length;
unichar patch_buf[WORD_MAX_CHARS];
const unichar *arg_begin;
size_t arg_length;
char word_buf[BUFSIZEOF__UTF8_WORD];
char *hugeword_buf = NULL;
size_t hugeword_buf_size = 0;
while (curr < buf_end)
{
word_begin = curr;
uchr = eh_decode_char__UTF8 (&curr, buf_end);
prop = unichar_getprops (uchr);
if (prop & UCP_ALPHA)
{
check_buf[0] = uchr;
word_length = 1;
for(;;)
{
word_end = curr;
uchr = eh_decode_char__UTF8 (&curr, buf_end);
if (uchr < 0)
{
if ((UNICHAR_NO_DATA == uchr) || (UNICHAR_BAD_ENCODING == uchr))
return uchr;
if (UNICHAR_EOD == uchr)
break;
}
prop = unichar_getprops (uchr);
if (!(prop & UCP_ALPHA))
break;
if (WORD_MAX_CHARS > word_length)
check_buf[word_length] = uchr;
word_length++;
}
if (WORD_MAX_CHARS < word_length)
goto done_word;
if (NULL!=check && 0 == check (check_buf, word_length))
goto done_word;
if (NULL != patch)
{
if (0 == patch (check_buf, word_length, patch_buf, &arg_length))
goto done_word;
arg_begin = patch_buf;
}
else
{
callback ((utf8char *) word_begin, word_end-word_begin, userdata);
goto done_word;
}
word_end = eh_encode_buffer__UTF8 (arg_begin, arg_begin+arg_length, word_buf, word_buf+BUFSIZEOF__UTF8_WORD);
if (NULL != word_end)
{
callback ((utf8char *)(word_buf), word_end-word_buf, userdata);
goto done_word;
}
if (hugeword_buf_size<(word_length*MAX_UTF8_CHAR))
{
if (hugeword_buf_size)
dk_free (hugeword_buf, hugeword_buf_size);
hugeword_buf_size = word_length*MAX_UTF8_CHAR;
hugeword_buf = (char *) dk_alloc (hugeword_buf_size);
}
word_end = eh_encode_buffer__UTF8 (arg_begin, arg_begin+arg_length, hugeword_buf, hugeword_buf+hugeword_buf_size);
callback ((utf8char *)(hugeword_buf), word_end-hugeword_buf, userdata);
done_word:
if (prop & UCP_IDEO)
goto proc_ideo;
continue;
}
if (prop & UCP_IDEO)
{
proc_ideo:
check_buf[0] = uchr;
if (NULL!=check && 0 == check (check_buf, 1))
continue;
if (NULL != patch)
{
if (0 == patch (check_buf, 1, patch_buf, &arg_length))
continue;
arg_begin = patch_buf;
}
else
{
callback ((utf8char *) word_begin, curr-word_begin, userdata);
continue;
}
word_end = eh_encode_buffer__UTF8 (arg_begin, arg_begin+arg_length, word_buf, word_buf+BUFSIZEOF__UTF8_WORD);
callback ((utf8char *)(word_buf), word_end-word_buf, userdata);
continue;
}
if ((uchr < 0) && ((UNICHAR_NO_DATA == uchr) || (UNICHAR_BAD_ENCODING == uchr)))
goto cleanup; /* see below */
}
uchr = 0;
cleanup:
if (hugeword_buf_size)
dk_free (hugeword_buf, hugeword_buf_size);
return uchr;
}
int elh_count_words__xViAny__UTF8(const char *buf, size_t bufsize, lh_word_check_t *check)
{
unichar check_buf[WORD_MAX_CHARS];
int res = 0;
int prop;
const char *curr = buf;
const char *buf_end = buf+bufsize;
const char *word_begin = curr;
const char *word_end = NULL;
unichar uchr;
size_t word_length;
while (curr < buf_end)
{
word_begin = curr;
uchr = eh_decode_char__UTF8 (&curr, buf_end);
prop = unichar_getprops (uchr);
if (prop & UCP_ALPHA)
{
check_buf[0] = uchr;
word_length = 1;
for(;;)
{
word_end = curr;
uchr = eh_decode_char__UTF8 (&curr, buf_end);
if (uchr < 0)
{
if ((UNICHAR_NO_DATA == uchr) || (UNICHAR_BAD_ENCODING == uchr))
return uchr;
if (UNICHAR_EOD == uchr)
break;
}
prop = unichar_getprops (uchr);
if (!(prop & UCP_ALPHA) && !IS_CONNECTIVE)
break;
if (WORD_MAX_CHARS > word_length)
check_buf[word_length] = uchr;
word_length++;
}
if (WORD_MAX_CHARS < word_length)
goto done_word;
if (NULL!=check && 0 == check(check_buf, word_length))
goto done_word;
res++;
done_word:
if (prop & UCP_IDEO)
goto proc_ideo;
continue;
}
if (prop & UCP_IDEO)
{
proc_ideo:
check_buf[0] = uchr;
if (NULL!=check && 0 == check(check_buf, 1))
continue;
res++;
continue;
}
if ((uchr < 0) && ((UNICHAR_NO_DATA == uchr) || (UNICHAR_BAD_ENCODING == uchr)))
return uchr;
}
return res;
}
