tb_char_t const* tb_regex_replace_done(tb_char_t const* pattern, tb_size_t mode, tb_char_t const* cstr, tb_size_t size, tb_size_t start, tb_char_t const* replace_cstr, tb_size_t replace_size, tb_size_t* plength)
{
// clear length first
if (plength) *plength = 0;
// init regex
tb_char_t* result = tb_null;
tb_regex_ref_t regex = tb_regex_init(pattern, mode);
if (regex)
{
// replace regex
tb_size_t result_size = 0;
tb_char_t const* result_cstr = tb_regex_replace(regex, cstr, size, start, replace_cstr, replace_size, &result_size);
if (result_cstr && result_size)
{
// save result
result = tb_strndup(result_cstr, result_size);
if (result)
{
// save length
if (plength) *plength = result_size;
}
}
// exit regex
tb_regex_exit(regex);
}
// ok?
return result;
}
tb_long_t tb_regex_match(tb_regex_ref_t self, tb_char_t const* cstr, tb_size_t size, tb_size_t start, tb_size_t* plength, tb_vector_ref_t* presults)
{
// check
tb_regex_t* regex = (tb_regex_t*)self;
tb_assert_and_check_return_val(regex && regex->code && regex->match_data && cstr, -1);
// done
tb_long_t ok = -1;
do
{
// clear length first
if (plength) *plength = 0;
// end?
tb_check_break(start < size);
// init options
#ifdef __tb_debug__
tb_uint32_t options = 0;
#else
tb_uint32_t options = PCRE2_NO_UTF_CHECK;
#endif
// match it
tb_long_t count = pcre2_match(regex->code, (PCRE2_SPTR)cstr, (PCRE2_SIZE)size, (PCRE2_SIZE)start, options, regex->match_data, tb_null);
if (count < 0)
{
// no match?
tb_check_break(count != PCRE2_ERROR_NOMATCH);
#if defined(__tb_debug__) && !defined(TB_CONFIG_OS_WINDOWS)
// get error info
PCRE2_UCHAR info[256];
pcre2_get_error_message(count, info, sizeof(info));
// trace
tb_trace_d("match failed at offset %lu: error: %ld, %s\n", start, count, info);
#endif
// end
break;
}
// check
tb_assertf_and_check_break(count, "ovector has not enough space!");
// get output vector
PCRE2_SIZE* ovector = pcre2_get_ovector_pointer(regex->match_data);
tb_assert_and_check_break(ovector);
// get the match offset and length
tb_size_t offset = (tb_size_t)ovector[0];
tb_size_t length = (tb_size_t)ovector[1] - ovector[0];
tb_assert_and_check_break(offset + length <= size);
// trace
tb_trace_d("matched count: %lu, offset: %lu, length: %lu", count, offset, length);
// save results
if (presults)
{
// init results if not exists
tb_vector_ref_t results = *presults;
if (!results)
{
// init it
if (!regex->results) regex->results = tb_vector_init(16, tb_element_mem(sizeof(tb_regex_match_t), tb_regex_match_exit, tb_null));
// save it
*presults = results = regex->results;
}
tb_assert_and_check_break(results);
// clear it first
tb_vector_clear(results);
// done
tb_long_t i = 0;
tb_regex_match_t entry;
for (i = 0; i < count; i++)
{
// get substring offset and length
tb_size_t substr_offset = ovector[i << 1];
tb_size_t substr_length = ovector[(i << 1) + 1] - ovector[i << 1];
tb_assert_and_check_break(substr_offset + substr_length <= size);
// make match entry
entry.cstr = tb_strndup(cstr + substr_offset, substr_length);
entry.size = substr_length;
entry.start = substr_offset;
tb_assert_and_check_break(entry.cstr);
// trace
tb_trace_d(" matched: [%lu, %lu]: %s", entry.start, entry.size, entry.cstr);
// append it
tb_vector_insert_tail(results, &entry);
}
tb_assert_and_check_break(i == count);
}
// save length
if (plength) *plength = length;
// ok
ok = offset;
} while (0);
// ok?
return ok;
}