1 /*
2 * The authors of this software are Rob Pike and Ken Thompson.
3 * Copyright (c) 2002 by Lucent Technologies.
4 * Permission to use, copy, modify, and distribute this software for any
5 * purpose without fee is hereby granted, provided that this entire notice
6 * is included in all copies of any software which is or includes a copy
7 * or modification of this software and in all copies of the supporting
8 * documentation for such software.
9 * THIS SOFTWARE IS BEING PROVIDED "AS IS", WITHOUT ANY EXPRESS OR IMPLIED
10 * WARRANTY. IN PARTICULAR, NEITHER THE AUTHORS NOR LUCENT TECHNOLOGIES MAKE
11 * ANY REPRESENTATION OR WARRANTY OF ANY KIND CONCERNING THE MERCHANTABILITY
12 * OF THIS SOFTWARE OR ITS FITNESS FOR ANY PARTICULAR PURPOSE.
13 */
14 #include <stdlib.h>
15 #include <string.h>
16
17 #include "utf.h"
18
19 typedef unsigned char uchar;
20
21 enum
22 {
23 Bit1 = 7,
24 Bitx = 6,
25 Bit2 = 5,
26 Bit3 = 4,
27 Bit4 = 3,
28
29 T1 = ((1<<(Bit1+1))-1) ^ 0xFF, /* 0000 0000 */
30 Tx = ((1<<(Bitx+1))-1) ^ 0xFF, /* 1000 0000 */
31 T2 = ((1<<(Bit2+1))-1) ^ 0xFF, /* 1100 0000 */
32 T3 = ((1<<(Bit3+1))-1) ^ 0xFF, /* 1110 0000 */
33 T4 = ((1<<(Bit4+1))-1) ^ 0xFF, /* 1111 0000 */
34
35 Rune1 = (1<<(Bit1+0*Bitx))-1, /* 0000 0000 0111 1111 */
36 Rune2 = (1<<(Bit2+1*Bitx))-1, /* 0000 0111 1111 1111 */
37 Rune3 = (1<<(Bit3+2*Bitx))-1, /* 1111 1111 1111 1111 */
38
39 Maskx = (1<<Bitx)-1, /* 0011 1111 */
40 Testx = Maskx ^ 0xFF, /* 1100 0000 */
41
42 Bad = Runeerror,
43 };
44
45 int
chartorune(Rune * rune,const char * str)46 chartorune(Rune *rune, const char *str)
47 {
48 int c, c1, c2;
49 int l;
50
51 /*
52 * one character sequence
53 * 00000-0007F => T1
54 */
55 c = *(uchar*)str;
56 if(c < Tx) {
57 *rune = c;
58 return 1;
59 }
60
61 /*
62 * two character sequence
63 * 0080-07FF => T2 Tx
64 */
65 c1 = *(uchar*)(str+1) ^ Tx;
66 if(c1 & Testx)
67 goto bad;
68 if(c < T3) {
69 if(c < T2)
70 goto bad;
71 l = ((c << Bitx) | c1) & Rune2;
72 if(l <= Rune1)
73 goto bad;
74 *rune = l;
75 return 2;
76 }
77
78 /*
79 * three character sequence
80 * 0800-FFFF => T3 Tx Tx
81 */
82 c2 = *(uchar*)(str+2) ^ Tx;
83 if(c2 & Testx)
84 goto bad;
85 if(c < T4) {
86 l = ((((c << Bitx) | c1) << Bitx) | c2) & Rune3;
87 if(l <= Rune2)
88 goto bad;
89 *rune = l;
90 return 3;
91 }
92
93 /*
94 * bad decoding
95 */
96 bad:
97 *rune = Bad;
98 return 1;
99 }
100
101 int
runetochar(char * str,const Rune * rune)102 runetochar(char *str, const Rune *rune)
103 {
104 int c;
105
106 /*
107 * one character sequence
108 * 00000-0007F => 00-7F
109 */
110 c = *rune;
111 if(c <= Rune1) {
112 str[0] = c;
113 return 1;
114 }
115
116 /*
117 * two character sequence
118 * 0080-07FF => T2 Tx
119 */
120 if(c <= Rune2) {
121 str[0] = T2 | (c >> 1*Bitx);
122 str[1] = Tx | (c & Maskx);
123 return 2;
124 }
125
126 /*
127 * three character sequence
128 * 0800-FFFF => T3 Tx Tx
129 */
130 str[0] = T3 | (c >> 2*Bitx);
131 str[1] = Tx | ((c >> 1*Bitx) & Maskx);
132 str[2] = Tx | (c & Maskx);
133 return 3;
134 }
135
136 int
runelen(int c)137 runelen(int c)
138 {
139 Rune rune;
140 char str[10];
141
142 rune = c;
143 return runetochar(str, &rune);
144 }
145
146 int
utflen(const char * s)147 utflen(const char *s)
148 {
149 int c;
150 int n;
151 Rune rune;
152
153 n = 0;
154 for(;;) {
155 c = *(uchar*)s;
156 if(c < Runeself) {
157 if(c == 0)
158 return n;
159 s++;
160 } else
161 s += chartorune(&rune, s);
162 n++;
163 }
164 }
165