Home | History | Annotate | Line # | Download | only in fs
unicode.h revision 1.2 
      1 /* $NetBSD: unicode.h,v 1.2 2004/12/27 18:14:36 jdolecek Exp $ */
      2 
      3 /*-
      4  * Copyright (c) 2001, 2004 The NetBSD Foundation, Inc.
      5  * All rights reserved.
      6  *
      7  * Redistribution and use in source and binary forms, with or without
      8  * modification, are permitted provided that the following conditions
      9  * are met:
     10  * 1. Redistributions of source code must retain the above copyright
     11  *    notice, this list of conditions and the following disclaimer.
     12  * 2. Redistributions in binary form must reproduce the above copyright
     13  *    notice, this list of conditions and the following disclaimer in the
     14  *    documentation and/or other materials provided with the distribution.
     15  * 3. All advertising materials mentioning features or use of this software
     16  *    must display the following acknowledgement:
     17  *        This product includes software developed by the NetBSD
     18  *        Foundation, Inc. and its contributors.
     19  * 4. Neither the name of The NetBSD Foundation nor the names of its
     20  *    contributors may be used to endorse or promote products derived
     21  *    from this software without specific prior written permission.
     22  *
     23  * THIS SOFTWARE IS PROVIDED BY THE NETBSD FOUNDATION, INC. AND CONTRIBUTORS
     24  * ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED
     25  * TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
     26  * PURPOSE ARE DISCLAIMED.  IN NO EVENT SHALL THE FOUNDATION OR CONTRIBUTORS
     27  * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
     28  * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
     29  * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
     30  * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
     31  * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
     32  * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
     33  * POSSIBILITY OF SUCH DAMAGE.
     34  */
     35 
     36 /*
     37  * Routines for handling Unicode encoded in UTF-8 form, code derived from
     38  * src/lib/libc/locale/utf2.c.
     39  */
     40 
     41 /*
     42  * Read one UTF8-encoded character off the string, shift the string pointer
     43  * and return the character.
     44  */
     45 static u_int16_t
     46 wget_utf8(const char **str, size_t *sz)
     47 {
     48 	int c;
     49 	u_int16_t rune = 0;
     50 	const char *s = *str;
     51 	static const int _utf_count[16] = {
     52 		1, 1, 1, 1, 1, 1, 1, 1,
     53 		0, 0, 0, 0, 2, 2, 3, 0,
     54 	};
     55 
     56 	/* must be called with at least one byte remaining */
     57 	KASSERT(*sz > 0);
     58 
     59 	c = _utf_count[(s[0] & 0xf0) >> 4];
     60 	if (c == 0 || c > *sz) {
     61     decoding_error:
     62 		/*
     63 		 * The first character is in range 128-255 and doesn't
     64 		 * mark valid a valid UTF-8 sequence. There is not much
     65 		 * we can do with this, so handle by returning
     66 		 * the first character as if it would be a correctly
     67 		 * encoded ISO-8859-1 character.
     68 		 */
     69 		c = 1;
     70 	}
     71 
     72 	switch (c) {
     73 	case 1:
     74 		rune = s[0] & 0xff;
     75 		break;
     76 	case 2:
     77 		if ((s[1] & 0xc0) != 0x80)
     78 			goto decoding_error;
     79 		rune = ((s[0] & 0x1F) << 6) | (s[1] & 0x3F);
     80 		break;
     81 	case 3:
     82 		if ((s[1] & 0xC0) != 0x80 || (s[2] & 0xC0) != 0x80)
     83 			goto decoding_error;
     84 		rune = ((s[0] & 0x1F) << 12) | ((s[1] & 0x3F) << 6)
     85 		    | (s[2] & 0x3F);
     86 		break;
     87 	}
     88 
     89 	*str += c;
     90 	*sz -= c;
     91 	return rune;
     92 }
     93 
     94 /*
     95  * Encode wide character and write it to the string. 'n' specifies
     96  * how much buffer space remains in 's'. Returns number of bytes written
     97  * to the target string 's'.
     98  */
     99 static int
    100 wput_utf8(char *s, size_t n, u_int16_t wc)
    101 {
    102 	if (wc & 0xf800) {
    103 		if (n < 3) {
    104 			/* bound check failure */
    105 			return 0;
    106 		}
    107 
    108 		s[0] = 0xE0 | ((wc >> 12) & 0x0F);
    109 		s[1] = 0x80 | ((wc >> 6) & 0x3F);
    110 		s[2] = 0x80 | ((wc) & 0x3F);
    111 		return 3;
    112 	} else if (wc & 0x0780) {
    113 		if (n < 2) {
    114 			/* bound check failure */
    115 			return 0;
    116 		}
    117 
    118 		s[0] = 0xC0 | ((wc >> 6) & 0x1F);
    119 		s[1] = 0x80 | ((wc) & 0x3F);
    120 		return 2;
    121 	} else {
    122 		if (n < 1) {
    123 			/* bound check failure */
    124 			return 0;
    125 		}
    126 
    127 		s[0] = wc;
    128 		return 1;
    129 	}
    130 }
    131