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 ANY
11  * REPRESENTATION OR WARRANTY OF ANY KIND CONCERNING THE MERCHANTABILITY
12  * OF THIS SOFTWARE OR ITS FITNESS FOR ANY PARTICULAR PURPOSE.
13  */
14 #include <stdarg.h>
15 #include <string.h>
16 #include "utf.h"
17 #include "utfdef.h"
18 
19 enum
20 {
21 	Bit1	= 7,
22 	Bitx	= 6,
23 	Bit2	= 5,
24 	Bit3	= 4,
25 	Bit4	= 3,
26 	Bit5	= 2,
27 
28 	T1	= ((1<<(Bit1+1))-1) ^ 0xFF,	/* 0000 0000 */
29 	Tx	= ((1<<(Bitx+1))-1) ^ 0xFF,	/* 1000 0000 */
30 	T2	= ((1<<(Bit2+1))-1) ^ 0xFF,	/* 1100 0000 */
31 	T3	= ((1<<(Bit3+1))-1) ^ 0xFF,	/* 1110 0000 */
32 	T4	= ((1<<(Bit4+1))-1) ^ 0xFF,	/* 1111 0000 */
33 	T5	= ((1<<(Bit5+1))-1) ^ 0xFF,	/* 1111 1000 */
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 	Rune4	= (1<<(Bit4+3*Bitx))-1,
39                                         /* 0001 1111 1111 1111 1111 1111 */
40 
41 	Maskx	= (1<<Bitx)-1,			/* 0011 1111 */
42 	Testx	= Maskx ^ 0xFF,			/* 1100 0000 */
43 
44 	Bad	= Runeerror,
45 };
46 
47 /*
48  * Modified by Wei-Hwa Huang, Google Inc., on 2004-09-24
49  * This is a slower but "safe" version of the old chartorune
50  * that works on strings that are not necessarily null-terminated.
51  *
52  * If you know for sure that your string is null-terminated,
53  * chartorune will be a bit faster.
54  *
55  * It is guaranteed not to attempt to access "length"
56  * past the incoming pointer.  This is to avoid
57  * possible access violations.  If the string appears to be
58  * well-formed but incomplete (i.e., to get the whole Rune
59  * we'd need to read past str+length) then we'll set the Rune
60  * to Bad and return 0.
61  *
62  * Note that if we have decoding problems for other
63  * reasons, we return 1 instead of 0.
64  */
65 int
charntorune(Rune * rune,const char * str,int length)66 charntorune(Rune *rune, const char *str, int length)
67 {
68 	int c, c1, c2, c3;
69 	long l;
70 
71 	/* When we're not allowed to read anything */
72 	if(length <= 0) {
73 		goto badlen;
74 	}
75 
76 	/*
77 	 * one character sequence (7-bit value)
78 	 *	00000-0007F => T1
79 	 */
80 	c = *(uchar*)str;
81 	if(c < Tx) {
82 		*rune = c;
83 		return 1;
84 	}
85 
86 	// If we can't read more than one character we must stop
87 	if(length <= 1) {
88 		goto badlen;
89 	}
90 
91 	/*
92 	 * two character sequence (11-bit value)
93 	 *	0080-07FF => T2 Tx
94 	 */
95 	c1 = *(uchar*)(str+1) ^ Tx;
96 	if(c1 & Testx)
97 		goto bad;
98 	if(c < T3) {
99 		if(c < T2)
100 			goto bad;
101 		l = ((c << Bitx) | c1) & Rune2;
102 		if(l <= Rune1)
103 			goto bad;
104 		*rune = l;
105 		return 2;
106 	}
107 
108 	// If we can't read more than two characters we must stop
109 	if(length <= 2) {
110 		goto badlen;
111 	}
112 
113 	/*
114 	 * three character sequence (16-bit value)
115 	 *	0800-FFFF => T3 Tx Tx
116 	 */
117 	c2 = *(uchar*)(str+2) ^ Tx;
118 	if(c2 & Testx)
119 		goto bad;
120 	if(c < T4) {
121 		l = ((((c << Bitx) | c1) << Bitx) | c2) & Rune3;
122 		if(l <= Rune2)
123 			goto bad;
124 		*rune = l;
125 		return 3;
126 	}
127 
128 	if (length <= 3)
129 		goto badlen;
130 
131 	/*
132 	 * four character sequence (21-bit value)
133 	 *	10000-1FFFFF => T4 Tx Tx Tx
134 	 */
135 	c3 = *(uchar*)(str+3) ^ Tx;
136 	if (c3 & Testx)
137 		goto bad;
138 	if (c < T5) {
139 		l = ((((((c << Bitx) | c1) << Bitx) | c2) << Bitx) | c3) & Rune4;
140 		if (l <= Rune3)
141 			goto bad;
142 		*rune = l;
143 		return 4;
144 	}
145 
146 	// Support for 5-byte or longer UTF-8 would go here, but
147 	// since we don't have that, we'll just fall through to bad.
148 
149 	/*
150 	 * bad decoding
151 	 */
152 bad:
153 	*rune = Bad;
154 	return 1;
155 badlen:
156 	*rune = Bad;
157 	return 0;
158 
159 }
160 
161 
162 /*
163  * This is the older "unsafe" version, which works fine on
164  * null-terminated strings.
165  */
166 int
chartorune(Rune * rune,const char * str)167 chartorune(Rune *rune, const char *str)
168 {
169 	int c, c1, c2, c3;
170 	long l;
171 
172 	/*
173 	 * one character sequence
174 	 *	00000-0007F => T1
175 	 */
176 	c = *(uchar*)str;
177 	if(c < Tx) {
178 		*rune = c;
179 		return 1;
180 	}
181 
182 	/*
183 	 * two character sequence
184 	 *	0080-07FF => T2 Tx
185 	 */
186 	c1 = *(uchar*)(str+1) ^ Tx;
187 	if(c1 & Testx)
188 		goto bad;
189 	if(c < T3) {
190 		if(c < T2)
191 			goto bad;
192 		l = ((c << Bitx) | c1) & Rune2;
193 		if(l <= Rune1)
194 			goto bad;
195 		*rune = l;
196 		return 2;
197 	}
198 
199 	/*
200 	 * three character sequence
201 	 *	0800-FFFF => T3 Tx Tx
202 	 */
203 	c2 = *(uchar*)(str+2) ^ Tx;
204 	if(c2 & Testx)
205 		goto bad;
206 	if(c < T4) {
207 		l = ((((c << Bitx) | c1) << Bitx) | c2) & Rune3;
208 		if(l <= Rune2)
209 			goto bad;
210 		*rune = l;
211 		return 3;
212 	}
213 
214 	/*
215 	 * four character sequence (21-bit value)
216 	 *	10000-1FFFFF => T4 Tx Tx Tx
217 	 */
218 	c3 = *(uchar*)(str+3) ^ Tx;
219 	if (c3 & Testx)
220 		goto bad;
221 	if (c < T5) {
222 		l = ((((((c << Bitx) | c1) << Bitx) | c2) << Bitx) | c3) & Rune4;
223 		if (l <= Rune3)
224 			goto bad;
225 		*rune = l;
226 		return 4;
227 	}
228 
229 	/*
230 	 * Support for 5-byte or longer UTF-8 would go here, but
231 	 * since we don't have that, we'll just fall through to bad.
232 	 */
233 
234 	/*
235 	 * bad decoding
236 	 */
237 bad:
238 	*rune = Bad;
239 	return 1;
240 }
241 
242 int
isvalidcharntorune(const char * str,int length,Rune * rune,int * consumed)243 isvalidcharntorune(const char* str, int length, Rune* rune, int* consumed) {
244 	*consumed = charntorune(rune, str, length);
245 	return *rune != Runeerror || *consumed == 3;
246 }
247 
248 int
runetochar(char * str,const Rune * rune)249 runetochar(char *str, const Rune *rune)
250 {
251 	/* Runes are signed, so convert to unsigned for range check. */
252 	unsigned long c;
253 
254 	/*
255 	 * one character sequence
256 	 *	00000-0007F => 00-7F
257 	 */
258 	c = *rune;
259 	if(c <= Rune1) {
260 		str[0] = c;
261 		return 1;
262 	}
263 
264 	/*
265 	 * two character sequence
266 	 *	0080-07FF => T2 Tx
267 	 */
268 	if(c <= Rune2) {
269 		str[0] = T2 | (c >> 1*Bitx);
270 		str[1] = Tx | (c & Maskx);
271 		return 2;
272 	}
273 
274 	/*
275 	 * If the Rune is out of range, convert it to the error rune.
276 	 * Do this test here because the error rune encodes to three bytes.
277 	 * Doing it earlier would duplicate work, since an out of range
278 	 * Rune wouldn't have fit in one or two bytes.
279 	 */
280 	if (c > Runemax)
281 		c = Runeerror;
282 
283 	/*
284 	 * three character sequence
285 	 *	0800-FFFF => T3 Tx Tx
286 	 */
287 	if (c <= Rune3) {
288 		str[0] = T3 |  (c >> 2*Bitx);
289 		str[1] = Tx | ((c >> 1*Bitx) & Maskx);
290 		str[2] = Tx |  (c & Maskx);
291 		return 3;
292 	}
293 
294 	/*
295 	 * four character sequence (21-bit value)
296 	 *     10000-1FFFFF => T4 Tx Tx Tx
297 	 */
298 	str[0] = T4 | (c >> 3*Bitx);
299 	str[1] = Tx | ((c >> 2*Bitx) & Maskx);
300 	str[2] = Tx | ((c >> 1*Bitx) & Maskx);
301 	str[3] = Tx | (c & Maskx);
302 	return 4;
303 }
304 
305 int
runelen(Rune rune)306 runelen(Rune rune)
307 {
308 	char str[10];
309 
310 	return runetochar(str, &rune);
311 }
312 
313 int
runenlen(const Rune * r,int nrune)314 runenlen(const Rune *r, int nrune)
315 {
316 	int nb, c;
317 
318 	nb = 0;
319 	while(nrune--) {
320 		c = *r++;
321 		if (c <= Rune1)
322 			nb++;
323 		else if (c <= Rune2)
324 			nb += 2;
325 		else if (c <= Rune3)
326 			nb += 3;
327 		else /* assert(c <= Rune4) */
328 			nb += 4;
329 	}
330 	return nb;
331 }
332 
333 int
fullrune(const char * str,int n)334 fullrune(const char *str, int n)
335 {
336 	if (n > 0) {
337 		int c = *(uchar*)str;
338 		if (c < Tx)
339 			return 1;
340 		if (n > 1) {
341 			if (c < T3)
342 				return 1;
343 			if (n > 2) {
344 				if (c < T4 || n > 3)
345 					return 1;
346 			}
347 		}
348 	}
349 	return 0;
350 }
351