1/*************************************************
2* Perl-Compatible Regular Expressions *
3*************************************************/
4
5/* PCRE is a library of functions to support regular expressions whose syntax
6and semantics are as close as possible to those of the Perl 5 language.
7
8 Written by Philip Hazel
9 Original API code Copyright (c) 1997-2012 University of Cambridge
10 New API code Copyright (c) 2016-2020 University of Cambridge
11
12-----------------------------------------------------------------------------
13Redistribution and use in source and binary forms, with or without
14modification, are permitted provided that the following conditions are met:
15
16 * Redistributions of source code must retain the above copyright notice,
17 this list of conditions and the following disclaimer.
18
19 * Redistributions in binary form must reproduce the above copyright
20 notice, this list of conditions and the following disclaimer in the
21 documentation and/or other materials provided with the distribution.
22
23 * Neither the name of the University of Cambridge nor the names of its
24 contributors may be used to endorse or promote products derived from
25 this software without specific prior written permission.
26
27THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
28AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
29IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
30ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE
31LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
32CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
33SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
34INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
35CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
36ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
37POSSIBILITY OF SUCH DAMAGE.
38-----------------------------------------------------------------------------
39*/
40
41
42/* This module contains an internal function for validating UTF character
43strings. This file is also #included by the pcre2test program, which uses
44macros to change names from _pcre2_xxx to xxxx, thereby avoiding name clashes
45with the library. In this case, PCRE2_PCRE2TEST is defined. */
46
47#ifndef PCRE2_PCRE2TEST /* We're compiling the library */
48#ifdef HAVE_CONFIG_H
49#include "config.h"
50#endif
51#include "pcre2_internal.h"
52#endif /* PCRE2_PCRE2TEST */
53
54
55#ifndef SUPPORT_UNICODE
56/*************************************************
57* Dummy function when Unicode is not supported *
58*************************************************/
59
60/* This function should never be called when Unicode is not supported. */
61
62int
63PRIV(valid_utf)(PCRE2_SPTR string, PCRE2_SIZE length, PCRE2_SIZE *erroroffset)
64{
65(void)string;
66(void)length;
67(void)erroroffset;
68return 0;
69}
70#else /* UTF is supported */
71
72
73
74/*************************************************
75* Validate a UTF string *
76*************************************************/
77
78/* This function is called (optionally) at the start of compile or match, to
79check that a supposed UTF string is actually valid. The early check means
80that subsequent code can assume it is dealing with a valid string. The check
81can be turned off for maximum performance, but the consequences of supplying an
82invalid string are then undefined.
83
84Arguments:
85 string points to the string
86 length length of string
87 errp pointer to an error position offset variable
88
89Returns: == 0 if the string is a valid UTF string
90 != 0 otherwise, setting the offset of the bad character
91*/
92
93int
94PRIV(valid_utf)(PCRE2_SPTR string, PCRE2_SIZE length, PCRE2_SIZE *erroroffset)
95{
96PCRE2_SPTR p;
97uint32_t c;
98
99/* ----------------- Check a UTF-8 string ----------------- */
100
101#if PCRE2_CODE_UNIT_WIDTH == 8
102
103/* Originally, this function checked according to RFC 2279, allowing for values
104in the range 0 to 0x7fffffff, up to 6 bytes long, but ensuring that they were
105in the canonical format. Once somebody had pointed out RFC 3629 to me (it
106obsoletes 2279), additional restrictions were applied. The values are now
107limited to be between 0 and 0x0010ffff, no more than 4 bytes long, and the
108subrange 0xd000 to 0xdfff is excluded. However, the format of 5-byte and 6-byte
109characters is still checked. Error returns are as follows:
110
111PCRE2_ERROR_UTF8_ERR1 Missing 1 byte at the end of the string
112PCRE2_ERROR_UTF8_ERR2 Missing 2 bytes at the end of the string
113PCRE2_ERROR_UTF8_ERR3 Missing 3 bytes at the end of the string
114PCRE2_ERROR_UTF8_ERR4 Missing 4 bytes at the end of the string
115PCRE2_ERROR_UTF8_ERR5 Missing 5 bytes at the end of the string
116PCRE2_ERROR_UTF8_ERR6 2nd-byte's two top bits are not 0x80
117PCRE2_ERROR_UTF8_ERR7 3rd-byte's two top bits are not 0x80
118PCRE2_ERROR_UTF8_ERR8 4th-byte's two top bits are not 0x80
119PCRE2_ERROR_UTF8_ERR9 5th-byte's two top bits are not 0x80
120PCRE2_ERROR_UTF8_ERR10 6th-byte's two top bits are not 0x80
121PCRE2_ERROR_UTF8_ERR11 5-byte character is not permitted by RFC 3629
122PCRE2_ERROR_UTF8_ERR12 6-byte character is not permitted by RFC 3629
123PCRE2_ERROR_UTF8_ERR13 4-byte character with value > 0x10ffff is not permitted
124PCRE2_ERROR_UTF8_ERR14 3-byte character with value 0xd800-0xdfff is not permitted
125PCRE2_ERROR_UTF8_ERR15 Overlong 2-byte sequence
126PCRE2_ERROR_UTF8_ERR16 Overlong 3-byte sequence
127PCRE2_ERROR_UTF8_ERR17 Overlong 4-byte sequence
128PCRE2_ERROR_UTF8_ERR18 Overlong 5-byte sequence (won't ever occur)
129PCRE2_ERROR_UTF8_ERR19 Overlong 6-byte sequence (won't ever occur)
130PCRE2_ERROR_UTF8_ERR20 Isolated 0x80 byte (not within UTF-8 character)
131PCRE2_ERROR_UTF8_ERR21 Byte with the illegal value 0xfe or 0xff
132*/
133
134for (p = string; length > 0; p++)
135 {
136 uint32_t ab, d;
137
138 c = *p;
139 length--;
140
141 if (c < 128) continue; /* ASCII character */
142
143 if (c < 0xc0) /* Isolated 10xx xxxx byte */
144 {
145 *erroroffset = (PCRE2_SIZE)(p - string);
146 return PCRE2_ERROR_UTF8_ERR20;
147 }
148
149 if (c >= 0xfe) /* Invalid 0xfe or 0xff bytes */
150 {
151 *erroroffset = (PCRE2_SIZE)(p - string);
152 return PCRE2_ERROR_UTF8_ERR21;
153 }
154
155 ab = PRIV(utf8_table4)[c & 0x3f]; /* Number of additional bytes (1-5) */
156 if (length < ab) /* Missing bytes */
157 {
158 *erroroffset = (PCRE2_SIZE)(p - string);
159 switch(ab - length)
160 {
161 case 1: return PCRE2_ERROR_UTF8_ERR1;
162 case 2: return PCRE2_ERROR_UTF8_ERR2;
163 case 3: return PCRE2_ERROR_UTF8_ERR3;
164 case 4: return PCRE2_ERROR_UTF8_ERR4;
165 case 5: return PCRE2_ERROR_UTF8_ERR5;
166 }
167 }
168 length -= ab; /* Length remaining */
169
170 /* Check top bits in the second byte */
171
172 if (((d = *(++p)) & 0xc0) != 0x80)
173 {
174 *erroroffset = (int)(p - string) - 1;
175 return PCRE2_ERROR_UTF8_ERR6;
176 }
177
178 /* For each length, check that the remaining bytes start with the 0x80 bit
179 set and not the 0x40 bit. Then check for an overlong sequence, and for the
180 excluded range 0xd800 to 0xdfff. */
181
182 switch (ab)
183 {
184 /* 2-byte character. No further bytes to check for 0x80. Check first byte
185 for for xx00 000x (overlong sequence). */
186
187 case 1: if ((c & 0x3e) == 0)
188 {
189 *erroroffset = (int)(p - string) - 1;
190 return PCRE2_ERROR_UTF8_ERR15;
191 }
192 break;
193
194 /* 3-byte character. Check third byte for 0x80. Then check first 2 bytes
195 for 1110 0000, xx0x xxxx (overlong sequence) or
196 1110 1101, 1010 xxxx (0xd800 - 0xdfff) */
197
198 case 2:
199 if ((*(++p) & 0xc0) != 0x80) /* Third byte */
200 {
201 *erroroffset = (int)(p - string) - 2;
202 return PCRE2_ERROR_UTF8_ERR7;
203 }
204 if (c == 0xe0 && (d & 0x20) == 0)
205 {
206 *erroroffset = (int)(p - string) - 2;
207 return PCRE2_ERROR_UTF8_ERR16;
208 }
209 if (c == 0xed && d >= 0xa0)
210 {
211 *erroroffset = (int)(p - string) - 2;
212 return PCRE2_ERROR_UTF8_ERR14;
213 }
214 break;
215
216 /* 4-byte character. Check 3rd and 4th bytes for 0x80. Then check first 2
217 bytes for for 1111 0000, xx00 xxxx (overlong sequence), then check for a
218 character greater than 0x0010ffff (f4 8f bf bf) */
219
220 case 3:
221 if ((*(++p) & 0xc0) != 0x80) /* Third byte */
222 {
223 *erroroffset = (int)(p - string) - 2;
224 return PCRE2_ERROR_UTF8_ERR7;
225 }
226 if ((*(++p) & 0xc0) != 0x80) /* Fourth byte */
227 {
228 *erroroffset = (int)(p - string) - 3;
229 return PCRE2_ERROR_UTF8_ERR8;
230 }
231 if (c == 0xf0 && (d & 0x30) == 0)
232 {
233 *erroroffset = (int)(p - string) - 3;
234 return PCRE2_ERROR_UTF8_ERR17;
235 }
236 if (c > 0xf4 || (c == 0xf4 && d > 0x8f))
237 {
238 *erroroffset = (int)(p - string) - 3;
239 return PCRE2_ERROR_UTF8_ERR13;
240 }
241 break;
242
243 /* 5-byte and 6-byte characters are not allowed by RFC 3629, and will be
244 rejected by the length test below. However, we do the appropriate tests
245 here so that overlong sequences get diagnosed, and also in case there is
246 ever an option for handling these larger code points. */
247
248 /* 5-byte character. Check 3rd, 4th, and 5th bytes for 0x80. Then check for
249 1111 1000, xx00 0xxx */
250
251 case 4:
252 if ((*(++p) & 0xc0) != 0x80) /* Third byte */
253 {
254 *erroroffset = (int)(p - string) - 2;
255 return PCRE2_ERROR_UTF8_ERR7;
256 }
257 if ((*(++p) & 0xc0) != 0x80) /* Fourth byte */
258 {
259 *erroroffset = (int)(p - string) - 3;
260 return PCRE2_ERROR_UTF8_ERR8;
261 }
262 if ((*(++p) & 0xc0) != 0x80) /* Fifth byte */
263 {
264 *erroroffset = (int)(p - string) - 4;
265 return PCRE2_ERROR_UTF8_ERR9;
266 }
267 if (c == 0xf8 && (d & 0x38) == 0)
268 {
269 *erroroffset = (int)(p - string) - 4;
270 return PCRE2_ERROR_UTF8_ERR18;
271 }
272 break;
273
274 /* 6-byte character. Check 3rd-6th bytes for 0x80. Then check for
275 1111 1100, xx00 00xx. */
276
277 case 5:
278 if ((*(++p) & 0xc0) != 0x80) /* Third byte */
279 {
280 *erroroffset = (int)(p - string) - 2;
281 return PCRE2_ERROR_UTF8_ERR7;
282 }
283 if ((*(++p) & 0xc0) != 0x80) /* Fourth byte */
284 {
285 *erroroffset = (int)(p - string) - 3;
286 return PCRE2_ERROR_UTF8_ERR8;
287 }
288 if ((*(++p) & 0xc0) != 0x80) /* Fifth byte */
289 {
290 *erroroffset = (int)(p - string) - 4;
291 return PCRE2_ERROR_UTF8_ERR9;
292 }
293 if ((*(++p) & 0xc0) != 0x80) /* Sixth byte */
294 {
295 *erroroffset = (int)(p - string) - 5;
296 return PCRE2_ERROR_UTF8_ERR10;
297 }
298 if (c == 0xfc && (d & 0x3c) == 0)
299 {
300 *erroroffset = (int)(p - string) - 5;
301 return PCRE2_ERROR_UTF8_ERR19;
302 }
303 break;
304 }
305
306 /* Character is valid under RFC 2279, but 4-byte and 5-byte characters are
307 excluded by RFC 3629. The pointer p is currently at the last byte of the
308 character. */
309
310 if (ab > 3)
311 {
312 *erroroffset = (int)(p - string) - ab;
313 return (ab == 4)? PCRE2_ERROR_UTF8_ERR11 : PCRE2_ERROR_UTF8_ERR12;
314 }
315 }
316return 0;
317
318
319/* ----------------- Check a UTF-16 string ----------------- */
320
321#elif PCRE2_CODE_UNIT_WIDTH == 16
322
323/* There's not so much work, nor so many errors, for UTF-16.
324PCRE2_ERROR_UTF16_ERR1 Missing low surrogate at the end of the string
325PCRE2_ERROR_UTF16_ERR2 Invalid low surrogate
326PCRE2_ERROR_UTF16_ERR3 Isolated low surrogate
327*/
328
329for (p = string; length > 0; p++)
330 {
331 c = *p;
332 length--;
333
334 if ((c & 0xf800) != 0xd800)
335 {
336 /* Normal UTF-16 code point. Neither high nor low surrogate. */
337 }
338 else if ((c & 0x0400) == 0)
339 {
340 /* High surrogate. Must be a followed by a low surrogate. */
341 if (length == 0)
342 {
343 *erroroffset = p - string;
344 return PCRE2_ERROR_UTF16_ERR1;
345 }
346 p++;
347 length--;
348 if ((*p & 0xfc00) != 0xdc00)
349 {
350 *erroroffset = p - string - 1;
351 return PCRE2_ERROR_UTF16_ERR2;
352 }
353 }
354 else
355 {
356 /* Isolated low surrogate. Always an error. */
357 *erroroffset = p - string;
358 return PCRE2_ERROR_UTF16_ERR3;
359 }
360 }
361return 0;
362
363
364
365/* ----------------- Check a UTF-32 string ----------------- */
366
367#else
368
369/* There is very little to do for a UTF-32 string.
370PCRE2_ERROR_UTF32_ERR1 Surrogate character
371PCRE2_ERROR_UTF32_ERR2 Character > 0x10ffff
372*/
373
374for (p = string; length > 0; length--, p++)
375 {
376 c = *p;
377 if ((c & 0xfffff800u) != 0xd800u)
378 {
379 /* Normal UTF-32 code point. Neither high nor low surrogate. */
380 if (c > 0x10ffffu)
381 {
382 *erroroffset = p - string;
383 return PCRE2_ERROR_UTF32_ERR2;
384 }
385 }
386 else
387 {
388 /* A surrogate */
389 *erroroffset = p - string;
390 return PCRE2_ERROR_UTF32_ERR1;
391 }
392 }
393return 0;
394#endif /* CODE_UNIT_WIDTH */
395}
396#endif /* SUPPORT_UNICODE */
397
398/* End of pcre2_valid_utf.c */
399