/[pcre]/code/trunk/pcre_valid_utf8.c
ViewVC logotype

Contents of /code/trunk/pcre_valid_utf8.c

Parent Directory Parent Directory | Revision Log Revision Log


Revision 654 - (hide annotations) (download)
Tue Aug 2 11:00:40 2011 UTC (3 years, 2 months ago) by ph10
File MIME type: text/plain
File size: 9961 byte(s)
Documentation and general text tidies in preparation for test release.

1 nigel 77 /*************************************************
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 ph10 409 Copyright (c) 1997-2009 University of Cambridge
10 nigel 77
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 ph10 200 #ifdef HAVE_CONFIG_H
46 ph10 236 #include "config.h"
47 ph10 200 #endif
48 ph10 199
49 nigel 77 #include "pcre_internal.h"
50    
51    
52     /*************************************************
53     * Validate a UTF-8 string *
54     *************************************************/
55    
56     /* This function is called (optionally) at the start of compile or match, to
57 ph10 598 check that a supposed UTF-8 string is actually valid. The early check means
58 nigel 77 that subsequent code can assume it is dealing with a valid string. The check
59 ph10 598 can be turned off for maximum performance, but the consequences of supplying an
60     invalid string are then undefined.
61 nigel 77
62 ph10 211 Originally, this function checked according to RFC 2279, allowing for values in
63     the range 0 to 0x7fffffff, up to 6 bytes long, but ensuring that they were in
64     the canonical format. Once somebody had pointed out RFC 3629 to me (it
65 ph10 293 obsoletes 2279), additional restrictions were applied. The values are now
66 ph10 211 limited to be between 0 and 0x0010ffff, no more than 4 bytes long, and the
67 ph10 654 subrange 0xd000 to 0xdfff is excluded. However, the format of 5-byte and 6-byte
68 ph10 598 characters is still checked.
69 ph10 211
70 ph10 654 From release 8.13 more information about the details of the error are passed
71 ph10 606 back in the returned value:
72 ph10 598
73     PCRE_UTF8_ERR0 No error
74     PCRE_UTF8_ERR1 Missing 1 byte at the end of the string
75     PCRE_UTF8_ERR2 Missing 2 bytes at the end of the string
76     PCRE_UTF8_ERR3 Missing 3 bytes at the end of the string
77     PCRE_UTF8_ERR4 Missing 4 bytes at the end of the string
78     PCRE_UTF8_ERR5 Missing 5 bytes at the end of the string
79     PCRE_UTF8_ERR6 2nd-byte's two top bits are not 0x80
80     PCRE_UTF8_ERR7 3rd-byte's two top bits are not 0x80
81     PCRE_UTF8_ERR8 4th-byte's two top bits are not 0x80
82     PCRE_UTF8_ERR9 5th-byte's two top bits are not 0x80
83     PCRE_UTF8_ERR10 6th-byte's two top bits are not 0x80
84     PCRE_UTF8_ERR11 5-byte character is not permitted by RFC 3629
85     PCRE_UTF8_ERR12 6-byte character is not permitted by RFC 3629
86     PCRE_UTF8_ERR13 4-byte character with value > 0x10ffff is not permitted
87     PCRE_UTF8_ERR14 3-byte character with value 0xd000-0xdfff is not permitted
88     PCRE_UTF8_ERR15 Overlong 2-byte sequence
89     PCRE_UTF8_ERR16 Overlong 3-byte sequence
90     PCRE_UTF8_ERR17 Overlong 4-byte sequence
91     PCRE_UTF8_ERR18 Overlong 5-byte sequence (won't ever occur)
92     PCRE_UTF8_ERR19 Overlong 6-byte sequence (won't ever occur)
93     PCRE_UTF8_ERR20 Isolated 0x80 byte (not within UTF-8 character)
94     PCRE_UTF8_ERR21 Byte with the illegal value 0xfe or 0xff
95    
96 nigel 77 Arguments:
97     string points to the string
98     length length of string, or -1 if the string is zero-terminated
99 ph10 654 errp pointer to an error position offset variable
100 nigel 77
101 ph10 606 Returns: = 0 if the string is a valid UTF-8 string
102     > 0 otherwise, setting the offset of the bad character
103 nigel 77 */
104    
105 nigel 87 int
106 ph10 606 _pcre_valid_utf8(USPTR string, int length, int *erroroffset)
107 nigel 77 {
108 ph10 110 #ifdef SUPPORT_UTF8
109 ph10 409 register USPTR p;
110 nigel 77
111     if (length < 0)
112     {
113     for (p = string; *p != 0; p++);
114     length = p - string;
115     }
116    
117     for (p = string; length-- > 0; p++)
118     {
119 ph10 598 register int ab, c, d;
120 ph10 654
121 ph10 598 c = *p;
122     if (c < 128) continue; /* ASCII character */
123 ph10 654
124 ph10 598 if (c < 0xc0) /* Isolated 10xx xxxx byte */
125     {
126 ph10 606 *erroroffset = p - string;
127 ph10 654 return PCRE_UTF8_ERR20;
128     }
129 ph10 598
130     if (c >= 0xfe) /* Invalid 0xfe or 0xff bytes */
131     {
132 ph10 606 *erroroffset = p - string;
133 ph10 654 return PCRE_UTF8_ERR21;
134     }
135    
136 ph10 211 ab = _pcre_utf8_table4[c & 0x3f]; /* Number of additional bytes */
137 ph10 654 if (length < ab)
138 ph10 598 {
139 ph10 606 *erroroffset = p - string; /* Missing bytes */
140     return ab - length; /* Codes ERR1 to ERR5 */
141 ph10 654 }
142 ph10 598 length -= ab; /* Length remaining */
143 nigel 77
144     /* Check top bits in the second byte */
145 ph10 654
146     if (((d = *(++p)) & 0xc0) != 0x80)
147 ph10 598 {
148 ph10 606 *erroroffset = p - string - 1;
149 ph10 654 return PCRE_UTF8_ERR6;
150     }
151 nigel 77
152 ph10 654 /* For each length, check that the remaining bytes start with the 0x80 bit
153 ph10 598 set and not the 0x40 bit. Then check for an overlong sequence, and for the
154     excluded range 0xd800 to 0xdfff. */
155 ph10 211
156 nigel 77 switch (ab)
157     {
158 ph10 598 /* 2-byte character. No further bytes to check for 0x80. Check first byte
159     for for xx00 000x (overlong sequence). */
160 ph10 654
161     case 1: if ((c & 0x3e) == 0)
162 ph10 598 {
163 ph10 654 *erroroffset = p - string - 1;
164     return PCRE_UTF8_ERR15;
165     }
166     break;
167 ph10 211
168 ph10 654 /* 3-byte character. Check third byte for 0x80. Then check first 2 bytes
169 ph10 598 for 1110 0000, xx0x xxxx (overlong sequence) or
170     1110 1101, 1010 xxxx (0xd800 - 0xdfff) */
171 ph10 654
172 nigel 77 case 2:
173 ph10 598 if ((*(++p) & 0xc0) != 0x80) /* Third byte */
174     {
175 ph10 654 *erroroffset = p - string - 2;
176 ph10 606 return PCRE_UTF8_ERR7;
177 ph10 654 }
178 ph10 598 if (c == 0xe0 && (d & 0x20) == 0)
179     {
180 ph10 606 *erroroffset = p - string - 2;
181 ph10 654 return PCRE_UTF8_ERR16;
182     }
183 ph10 598 if (c == 0xed && d >= 0xa0)
184     {
185 ph10 606 *erroroffset = p - string - 2;
186 ph10 654 return PCRE_UTF8_ERR14;
187     }
188 nigel 77 break;
189    
190 ph10 598 /* 4-byte character. Check 3rd and 4th bytes for 0x80. Then check first 2
191     bytes for for 1111 0000, xx00 xxxx (overlong sequence), then check for a
192     character greater than 0x0010ffff (f4 8f bf bf) */
193 ph10 654
194 nigel 77 case 3:
195 ph10 598 if ((*(++p) & 0xc0) != 0x80) /* Third byte */
196     {
197 ph10 654 *erroroffset = p - string - 2;
198 ph10 606 return PCRE_UTF8_ERR7;
199 ph10 654 }
200 ph10 598 if ((*(++p) & 0xc0) != 0x80) /* Fourth byte */
201     {
202 ph10 654 *erroroffset = p - string - 3;
203 ph10 606 return PCRE_UTF8_ERR8;
204 ph10 654 }
205 ph10 598 if (c == 0xf0 && (d & 0x30) == 0)
206     {
207 ph10 606 *erroroffset = p - string - 3;
208 ph10 654 return PCRE_UTF8_ERR17;
209     }
210 ph10 598 if (c > 0xf4 || (c == 0xf4 && d > 0x8f))
211     {
212 ph10 606 *erroroffset = p - string - 3;
213 ph10 654 return PCRE_UTF8_ERR13;
214 ph10 598 }
215 nigel 77 break;
216    
217 ph10 598 /* 5-byte and 6-byte characters are not allowed by RFC 3629, and will be
218 ph10 654 rejected by the length test below. However, we do the appropriate tests
219 ph10 598 here so that overlong sequences get diagnosed, and also in case there is
220 ph10 654 ever an option for handling these larger code points. */
221 ph10 211
222 ph10 598 /* 5-byte character. Check 3rd, 4th, and 5th bytes for 0x80. Then check for
223 ph10 654 1111 1000, xx00 0xxx */
224    
225     case 4:
226 ph10 598 if ((*(++p) & 0xc0) != 0x80) /* Third byte */
227     {
228 ph10 654 *erroroffset = p - string - 2;
229 ph10 606 return PCRE_UTF8_ERR7;
230 ph10 654 }
231 ph10 598 if ((*(++p) & 0xc0) != 0x80) /* Fourth byte */
232     {
233 ph10 654 *erroroffset = p - string - 3;
234 ph10 606 return PCRE_UTF8_ERR8;
235 ph10 654 }
236 ph10 598 if ((*(++p) & 0xc0) != 0x80) /* Fifth byte */
237     {
238 ph10 654 *erroroffset = p - string - 4;
239 ph10 606 return PCRE_UTF8_ERR9;
240 ph10 654 }
241     if (c == 0xf8 && (d & 0x38) == 0)
242 ph10 598 {
243 ph10 606 *erroroffset = p - string - 4;
244 ph10 654 return PCRE_UTF8_ERR18;
245     }
246 nigel 77 break;
247    
248 ph10 598 /* 6-byte character. Check 3rd-6th bytes for 0x80. Then check for
249     1111 1100, xx00 00xx. */
250    
251 nigel 77 case 5:
252 ph10 598 if ((*(++p) & 0xc0) != 0x80) /* Third byte */
253     {
254 ph10 654 *erroroffset = p - string - 2;
255 ph10 606 return PCRE_UTF8_ERR7;
256 ph10 654 }
257 ph10 598 if ((*(++p) & 0xc0) != 0x80) /* Fourth byte */
258     {
259 ph10 654 *erroroffset = p - string - 3;
260 ph10 606 return PCRE_UTF8_ERR8;
261 ph10 654 }
262 ph10 598 if ((*(++p) & 0xc0) != 0x80) /* Fifth byte */
263     {
264 ph10 654 *erroroffset = p - string - 4;
265 ph10 606 return PCRE_UTF8_ERR9;
266 ph10 654 }
267 ph10 598 if ((*(++p) & 0xc0) != 0x80) /* Sixth byte */
268     {
269 ph10 654 *erroroffset = p - string - 5;
270 ph10 606 return PCRE_UTF8_ERR10;
271 ph10 654 }
272     if (c == 0xfc && (d & 0x3c) == 0)
273 ph10 598 {
274 ph10 606 *erroroffset = p - string - 5;
275 ph10 654 return PCRE_UTF8_ERR19;
276     }
277 nigel 77 break;
278     }
279 ph10 654
280 ph10 598 /* Character is valid under RFC 2279, but 4-byte and 5-byte characters are
281     excluded by RFC 3629. The pointer p is currently at the last byte of the
282     character. */
283 nigel 77
284 ph10 654 if (ab > 3)
285 nigel 77 {
286 ph10 606 *erroroffset = p - string - ab;
287 ph10 654 return (ab == 4)? PCRE_UTF8_ERR11 : PCRE_UTF8_ERR12;
288     }
289 nigel 77 }
290 ph10 654
291 ph10 598 #else /* SUPPORT_UTF8 */
292 ph10 369 (void)(string); /* Keep picky compilers happy */
293 ph10 371 (void)(length);
294 ph10 110 #endif
295 nigel 77
296 ph10 606 return PCRE_UTF8_ERR0; /* This indicates success */
297 nigel 77 }
298    
299     /* End of pcre_valid_utf8.c */

Properties

Name Value
svn:eol-style native
svn:keywords "Author Date Id Revision Url"

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12