nuclear@10: /*
nuclear@10: libdrawtext - a simple library for fast text rendering in OpenGL
nuclear@10: Copyright (C) 2011  John Tsiombikas <nuclear@member.fsf.org>
nuclear@10: 
nuclear@10: This program is free software: you can redistribute it and/or modify
nuclear@10: it under the terms of the GNU Lesser General Public License as published by
nuclear@10: the Free Software Foundation, either version 3 of the License, or
nuclear@10: (at your option) any later version.
nuclear@10: 
nuclear@10: This program is distributed in the hope that it will be useful,
nuclear@10: but WITHOUT ANY WARRANTY; without even the implied warranty of
nuclear@10: MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
nuclear@10: GNU Lesser General Public License for more details.
nuclear@10: 
nuclear@10: You should have received a copy of the GNU Lesser General Public License
nuclear@10: along with this program.  If not, see <http://www.gnu.org/licenses/>.
nuclear@10: */
nuclear@10: #include "drawtext.h"
nuclear@10: 
nuclear@10: #define	U8_IS_FIRST(x)		(((((x) >> 7) & 1) == 0) || ((((x) >> 6) & 3) == 3))
nuclear@10: 
nuclear@10: static const char first_mask[] = {
nuclear@10: 	0,
nuclear@10: 	0x7f,	/* single byte, 7 bits valid */
nuclear@10: 	0x1f,	/* two-bytes, 5 bits valid */
nuclear@10: 	0xf,	/* three-bytes, 4 bits valid */
nuclear@10: 	0x7		/* four-bytes, 3 bits valid */
nuclear@10: };
nuclear@10: static const char first_shift[] = { 0, 7, 5, 4, 3 };	/* see above */
nuclear@10: 
nuclear@10: #define CONT_PREFIX	0x80
nuclear@10: #define CONT_MASK	0x3f
nuclear@10: #define CONT_SHIFT	6
nuclear@10: 
nuclear@10: /* last charcodes for 1, 2, 3 or 4-byte utf8 chars */
nuclear@10: static const int utf8_lastcode[] = { 0x7f, 0x7ff, 0xfff, 0x1fffff };
nuclear@10: 
nuclear@10: #define prefix_mask(x)	(~first_mask[x])
nuclear@10: #define prefix(x)		((prefix_mask(x) << 1) & 0xff)
nuclear@10: 
nuclear@10: 
nuclear@10: char *dtx_utf8_next_char(char *str)
nuclear@10: {
nuclear@10: 	return str + dtx_utf8_nbytes(str);
nuclear@10: }
nuclear@10: 
nuclear@10: int dtx_utf8_char_code(const char *str)
nuclear@10: {
nuclear@10: 	int i, nbytes, shift, code = 0;
nuclear@10: 	int mask;
nuclear@10: 
nuclear@10: 	if(!U8_IS_FIRST(*str)) {
nuclear@10: 		return -1;
nuclear@10: 	}
nuclear@10: 
nuclear@10: 	nbytes = dtx_utf8_nbytes(str);
nuclear@10: 	mask = first_mask[nbytes];
nuclear@10: 	shift = 0;
nuclear@10: 
nuclear@10: 	for(i=0; i<nbytes; i++) {
nuclear@10: 		if(!*str) {
nuclear@10: 			break;
nuclear@10: 		}
nuclear@10: 
nuclear@10: 		code = (code << shift) | (*str++ & mask);
nuclear@10: 		mask = 0x3f;
nuclear@10: 		shift = 6;
nuclear@10: 	}
nuclear@10: 	return code;
nuclear@10: }
nuclear@10: 
nuclear@10: int dtx_utf8_nbytes(const char *str)
nuclear@10: {
nuclear@10: 	int i, numset = 0;
nuclear@10: 	int c = *str;
nuclear@10: 
nuclear@10: 	if(!U8_IS_FIRST(c)) {
nuclear@10: 		for(i=0; !U8_IS_FIRST(str[i]); i++);
nuclear@10: 		return i;
nuclear@10: 	}
nuclear@10: 
nuclear@10: 	/* count the leading 1s */
nuclear@10: 	for(i=0; i<4; i++) {
nuclear@10: 		if(((c >> (7 - i)) & 1) == 0) {
nuclear@10: 			break;
nuclear@10: 		}
nuclear@10: 		numset++;
nuclear@10: 	}
nuclear@10: 
nuclear@10: 	if(!numset) {
nuclear@10: 		return 1;
nuclear@10: 	}
nuclear@10: 	return numset;
nuclear@10: }
nuclear@10: 
nuclear@10: int dtx_utf8_char_count(const char *str)
nuclear@10: {
nuclear@10: 	int n = 0;
nuclear@10: 
nuclear@10: 	while(*str) {
nuclear@10: 		n++;
nuclear@10: 		str = dtx_utf8_next_char((char*)str);
nuclear@10: 	}
nuclear@10: 	return n;
nuclear@10: }
nuclear@10: 
nuclear@10: size_t dtx_utf8_from_char_code(int code, char *buf)
nuclear@10: {
nuclear@10: 	size_t nbytes = 0;
nuclear@10: 	int i;
nuclear@10: 
nuclear@10: 	for(i=0; i<4; i++) {
nuclear@10: 		if(code <= utf8_lastcode[i]) {
nuclear@10: 			nbytes = i + 1;
nuclear@10: 			break;
nuclear@10: 		}
nuclear@10: 	}
nuclear@10: 
nuclear@10: 	if(!nbytes && buf) {
nuclear@10: 		for(i=0; i<(int)nbytes; i++) {
nuclear@10: 			int idx = nbytes - i - 1;
nuclear@10: 			int mask, shift, prefix;
nuclear@10: 
nuclear@10: 			if(idx > 0) {
nuclear@10: 				mask = CONT_MASK;
nuclear@10: 				shift = CONT_SHIFT;
nuclear@10: 				prefix = CONT_PREFIX;
nuclear@10: 			} else {
nuclear@10: 				mask = first_mask[nbytes];
nuclear@10: 				shift = first_shift[nbytes];
nuclear@10: 				prefix = prefix(nbytes);
nuclear@10: 			}
nuclear@10: 
nuclear@10: 			buf[idx] = (code & mask) | (prefix & ~mask);
nuclear@10: 			code >>= shift;
nuclear@10: 		}
nuclear@10: 	}
nuclear@10: 	return nbytes;
nuclear@10: }
nuclear@10: 
nuclear@10: size_t dtx_utf8_from_string(const wchar_t *str, char *buf)
nuclear@10: {
nuclear@10: 	size_t nbytes = 0;
nuclear@10: 	char *ptr = buf;
nuclear@10: 
nuclear@10: 	while(*str) {
nuclear@10: 		int cbytes = dtx_utf8_from_char_code(*str++, ptr);
nuclear@10: 		if(ptr) {
nuclear@10: 			ptr += cbytes;
nuclear@10: 		}
nuclear@10: 		nbytes += cbytes;
nuclear@10: 	}
nuclear@10: 	return nbytes;
nuclear@10: }