1 /*************************************************
2 * Perl-Compatible Regular Expressions *
3 *************************************************/
4
5 /* PCRE is a library of functions to support regular expressions whose syntax
6 and semantics are as close as possible to those of the Perl 5 language.
7
8 Written by Philip Hazel
9 Copyright (c) 1997-2013 University of Cambridge
10
11 -----------------------------------------------------------------------------
12 Redistribution and use in source and binary forms, with or without
13 modification, are permitted provided that the following conditions are met:
14
15 * Redistributions of source code must retain the above copyright notice,
16 this list of conditions and the following disclaimer.
17
18 * Redistributions in binary form must reproduce the above copyright
19 notice, this list of conditions and the following disclaimer in the
20 documentation and/or other materials provided with the distribution.
21
22 * Neither the name of the University of Cambridge nor the names of its
23 contributors may be used to endorse or promote products derived from
24 this software without specific prior written permission.
25
26 THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
27 AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
28 IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
29 ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE
30 LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
31 CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
32 SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
33 INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
34 CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
35 ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
36 POSSIBILITY OF SUCH DAMAGE.
37 -----------------------------------------------------------------------------
38 */
39
40
41 /* This module contains an internal function for validating UTF-8 character
42 strings. */
43
44
45 #include "config.h"
46
47 #include "pcre_internal.h"
48
49
50 /*************************************************
51 * Validate a UTF-8 string *
52 *************************************************/
53
54 /* This function is called (optionally) at the start of compile or match, to
55 check that a supposed UTF-8 string is actually valid. The early check means
56 that subsequent code can assume it is dealing with a valid string. The check
57 can be turned off for maximum performance, but the consequences of supplying an
58 invalid string are then undefined.
59
60 Originally, this function checked according to RFC 2279, allowing for values in
61 the range 0 to 0x7fffffff, up to 6 bytes long, but ensuring that they were in
62 the canonical format. Once somebody had pointed out RFC 3629 to me (it
63 obsoletes 2279), additional restrictions were applied. The values are now
64 limited to be between 0 and 0x0010ffff, no more than 4 bytes long, and the
65 subrange 0xd000 to 0xdfff is excluded. However, the format of 5-byte and 6-byte
66 characters is still checked.
67
68 From release 8.13 more information about the details of the error are passed
69 back in the returned value:
70
71 PCRE_UTF8_ERR0 No error
72 PCRE_UTF8_ERR1 Missing 1 byte at the end of the string
73 PCRE_UTF8_ERR2 Missing 2 bytes at the end of the string
74 PCRE_UTF8_ERR3 Missing 3 bytes at the end of the string
75 PCRE_UTF8_ERR4 Missing 4 bytes at the end of the string
76 PCRE_UTF8_ERR5 Missing 5 bytes at the end of the string
77 PCRE_UTF8_ERR6 2nd-byte's two top bits are not 0x80
78 PCRE_UTF8_ERR7 3rd-byte's two top bits are not 0x80
79 PCRE_UTF8_ERR8 4th-byte's two top bits are not 0x80
80 PCRE_UTF8_ERR9 5th-byte's two top bits are not 0x80
81 PCRE_UTF8_ERR10 6th-byte's two top bits are not 0x80
82 PCRE_UTF8_ERR11 5-byte character is not permitted by RFC 3629
83 PCRE_UTF8_ERR12 6-byte character is not permitted by RFC 3629
84 PCRE_UTF8_ERR13 4-byte character with value > 0x10ffff is not permitted
85 PCRE_UTF8_ERR14 3-byte character with value 0xd000-0xdfff is not permitted
86 PCRE_UTF8_ERR15 Overlong 2-byte sequence
87 PCRE_UTF8_ERR16 Overlong 3-byte sequence
88 PCRE_UTF8_ERR17 Overlong 4-byte sequence
89 PCRE_UTF8_ERR18 Overlong 5-byte sequence (won't ever occur)
90 PCRE_UTF8_ERR19 Overlong 6-byte sequence (won't ever occur)
91 PCRE_UTF8_ERR20 Isolated 0x80 byte (not within UTF-8 character)
92 PCRE_UTF8_ERR21 Byte with the illegal value 0xfe or 0xff
93 PCRE_UTF8_ERR22 Unused (was non-character)
94
95 Arguments:
96 string points to the string
97 length length of string, or -1 if the string is zero-terminated
98 errp pointer to an error position offset variable
99
100 Returns: = 0 if the string is a valid UTF-8 string
101 > 0 otherwise, setting the offset of the bad character
102 */
103
104 int
105 PRIV(valid_utf)(PCRE_PUCHAR string, int length, int *erroroffset)
106 {
107 #ifdef SUPPORT_UTF
108 register PCRE_PUCHAR p;
109
110 if (length < 0)
111 {
112 for (p = string; *p != 0; p++);
113 length = (int)(p - string);
114 }
115
116 for (p = string; length-- > 0; p++)
117 {
118 register pcre_uchar ab, c, d;
119
120 c = *p;
121 if (c < 128) continue; /* ASCII character */
122
123 if (c < 0xc0) /* Isolated 10xx xxxx byte */
124 {
125 *erroroffset = (int)(p - string);
126 return PCRE_UTF8_ERR20;
127 }
128
129 if (c >= 0xfe) /* Invalid 0xfe or 0xff bytes */
130 {
131 *erroroffset = (int)(p - string);
132 return PCRE_UTF8_ERR21;
133 }
134
135 ab = PRIV(utf8_table4)[c & 0x3f]; /* Number of additional bytes */
136 if (length < ab)
137 {
138 *erroroffset = (int)(p - string); /* Missing bytes */
139 return ab - length; /* Codes ERR1 to ERR5 */
140 }
141 length -= ab; /* Length remaining */
142
143 /* Check top bits in the second byte */
144
145 if (((d = *(++p)) & 0xc0) != 0x80)
146 {
147 *erroroffset = (int)(p - string) - 1;
148 return PCRE_UTF8_ERR6;
149 }
150
151 /* For each length, check that the remaining bytes start with the 0x80 bit
152 set and not the 0x40 bit. Then check for an overlong sequence, and for the
153 excluded range 0xd800 to 0xdfff. */
154
155 switch (ab)
156 {
157 /* 2-byte character. No further bytes to check for 0x80. Check first byte
158 for for xx00 000x (overlong sequence). */
159
160 case 1: if ((c & 0x3e) == 0)
161 {
162 *erroroffset = (int)(p - string) - 1;
163 return PCRE_UTF8_ERR15;
164 }
165 break;
166
167 /* 3-byte character. Check third byte for 0x80. Then check first 2 bytes
168 for 1110 0000, xx0x xxxx (overlong sequence) or
169 1110 1101, 1010 xxxx (0xd800 - 0xdfff) */
170
171 case 2:
172 if ((*(++p) & 0xc0) != 0x80) /* Third byte */
173 {
174 *erroroffset = (int)(p - string) - 2;
175 return PCRE_UTF8_ERR7;
176 }
177 if (c == 0xe0 && (d & 0x20) == 0)
178 {
179 *erroroffset = (int)(p - string) - 2;
180 return PCRE_UTF8_ERR16;
181 }
182 if (c == 0xed && d >= 0xa0)
183 {
184 *erroroffset = (int)(p - string) - 2;
185 return PCRE_UTF8_ERR14;
186 }
187 break;
188
189 /* 4-byte character. Check 3rd and 4th bytes for 0x80. Then check first 2
190 bytes for for 1111 0000, xx00 xxxx (overlong sequence), then check for a
191 character greater than 0x0010ffff (f4 8f bf bf) */
192
193 case 3:
194 if ((*(++p) & 0xc0) != 0x80) /* Third byte */
195 {
196 *erroroffset = (int)(p - string) - 2;
197 return PCRE_UTF8_ERR7;
198 }
199 if ((*(++p) & 0xc0) != 0x80) /* Fourth byte */
200 {
201 *erroroffset = (int)(p - string) - 3;
202 return PCRE_UTF8_ERR8;
203 }
204 if (c == 0xf0 && (d & 0x30) == 0)
205 {
206 *erroroffset = (int)(p - string) - 3;
207 return PCRE_UTF8_ERR17;
208 }
209 if (c > 0xf4 || (c == 0xf4 && d > 0x8f))
210 {
211 *erroroffset = (int)(p - string) - 3;
212 return PCRE_UTF8_ERR13;
213 }
214 break;
215
216 /* 5-byte and 6-byte characters are not allowed by RFC 3629, and will be
217 rejected by the length test below. However, we do the appropriate tests
218 here so that overlong sequences get diagnosed, and also in case there is
219 ever an option for handling these larger code points. */
220
221 /* 5-byte character. Check 3rd, 4th, and 5th bytes for 0x80. Then check for
222 1111 1000, xx00 0xxx */
223
224 case 4:
225 if ((*(++p) & 0xc0) != 0x80) /* Third byte */
226 {
227 *erroroffset = (int)(p - string) - 2;
228 return PCRE_UTF8_ERR7;
229 }
230 if ((*(++p) & 0xc0) != 0x80) /* Fourth byte */
231 {
232 *erroroffset = (int)(p - string) - 3;
233 return PCRE_UTF8_ERR8;
234 }
235 if ((*(++p) & 0xc0) != 0x80) /* Fifth byte */
236 {
237 *erroroffset = (int)(p - string) - 4;
238 return PCRE_UTF8_ERR9;
239 }
240 if (c == 0xf8 && (d & 0x38) == 0)
241 {
242 *erroroffset = (int)(p - string) - 4;
243 return PCRE_UTF8_ERR18;
244 }
245 break;
246
247 /* 6-byte character. Check 3rd-6th bytes for 0x80. Then check for
248 1111 1100, xx00 00xx. */
249
250 case 5:
251 if ((*(++p) & 0xc0) != 0x80) /* Third byte */
252 {
253 *erroroffset = (int)(p - string) - 2;
254 return PCRE_UTF8_ERR7;
255 }
256 if ((*(++p) & 0xc0) != 0x80) /* Fourth byte */
257 {
258 *erroroffset = (int)(p - string) - 3;
259 return PCRE_UTF8_ERR8;
260 }
261 if ((*(++p) & 0xc0) != 0x80) /* Fifth byte */
262 {
263 *erroroffset = (int)(p - string) - 4;
264 return PCRE_UTF8_ERR9;
265 }
266 if ((*(++p) & 0xc0) != 0x80) /* Sixth byte */
267 {
268 *erroroffset = (int)(p - string) - 5;
269 return PCRE_UTF8_ERR10;
270 }
271 if (c == 0xfc && (d & 0x3c) == 0)
272 {
273 *erroroffset = (int)(p - string) - 5;
274 return PCRE_UTF8_ERR19;
275 }
276 break;
277 }
278
279 /* Character is valid under RFC 2279, but 4-byte and 5-byte characters are
280 excluded by RFC 3629. The pointer p is currently at the last byte of the
281 character. */
282
283 if (ab > 3)
284 {
285 *erroroffset = (int)(p - string) - ab;
286 return (ab == 4)? PCRE_UTF8_ERR11 : PCRE_UTF8_ERR12;
287 }
288 }
289
290 #else /* Not SUPPORT_UTF */
291 (void)(string); /* Keep picky compilers happy */
292 (void)(length);
293 (void)(erroroffset);
294 #endif
295
296 return PCRE_UTF8_ERR0; /* This indicates success */
297 }
298
299 /* End of pcre_valid_utf8.c */