/* libmisc/utf8.c - UTF-8 routines
 *
 * Copyright (C) 2024-2025  Luke T. Shumaker <lukeshu@lukeshu.com>
 * SPDX-License-Identifier: AGPL-3.0-or-later
 */

#include <libmisc/utf8.h>

void utf8_decode_codepoint(const uint8_t *str, size_t len, uint32_t *ret_ch, uint8_t *ret_chlen) {
	uint32_t ch;
	uint8_t  chlen;
	if      ((str[0] & 0b10000000) == 0b00000000) { ch = str[0] & 0b01111111; chlen = 1; }
	else if ((str[0] & 0b11100000) == 0b11000000) { ch = str[0] & 0b00011111; chlen = 2; }
	else if ((str[0] & 0b11110000) == 0b11100000) { ch = str[0] & 0b00001111; chlen = 3; }
	else if ((str[0] & 0b11111000) == 0b11110000) { ch = str[0] & 0b00000111; chlen = 4; }
	else goto invalid;
	if ((ch == 0 && chlen != 1) || chlen > len) goto invalid;
	for (uint8_t i = 1; i < chlen; i++) {
		if ((str[i] & 0b11000000) != 0b10000000) goto invalid;
		ch = (ch << 6) | (str[i] & 0b00111111);
	}
	if (ch > 0x10FFFF) goto invalid;
	*ret_ch    = ch;
	*ret_chlen = chlen;
	return;
 invalid:
	*ret_chlen = 0;
}

bool _utf8_is_valid(const uint8_t *str, size_t len, bool forbid_nul) {
	for (size_t pos = 0; pos < len;) {
		uint32_t ch;
		uint8_t  chlen;
		utf8_decode_codepoint(&str[pos], len-pos, &ch, &chlen);
		if (chlen == 0 || (forbid_nul && ch == 0))
			return false;
		pos += chlen;
	}
	return true;
}