Tristan Matthews | 0461646 | 2013-11-14 16:09:34 -0500 | [diff] [blame] | 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) 1997-2009 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 | #ifdef HAVE_CONFIG_H |
| 46 | #include "config.h" |
| 47 | #endif |
| 48 | |
| 49 | #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 | check that a supposed UTF-8 string is actually valid. The early check means |
| 58 | that subsequent code can assume it is dealing with a valid string. The check |
| 59 | can be turned off for maximum performance, but the consequences of supplying an |
| 60 | invalid string are then undefined. |
| 61 | |
| 62 | 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 | obsoletes 2279), additional restrictions were applied. The values are now |
| 66 | limited to be between 0 and 0x0010ffff, no more than 4 bytes long, and the |
| 67 | subrange 0xd000 to 0xdfff is excluded. However, the format of 5-byte and 6-byte |
| 68 | characters is still checked. |
| 69 | |
| 70 | From release 8.13 more information about the details of the error are passed |
| 71 | back in the returned value: |
| 72 | |
| 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 | Arguments: |
| 97 | string points to the string |
| 98 | length length of string, or -1 if the string is zero-terminated |
| 99 | errp pointer to an error position offset variable |
| 100 | |
| 101 | Returns: = 0 if the string is a valid UTF-8 string |
| 102 | > 0 otherwise, setting the offset of the bad character |
| 103 | */ |
| 104 | |
| 105 | int |
| 106 | _pcre_valid_utf8(USPTR string, int length, int *erroroffset) |
| 107 | { |
| 108 | #ifdef SUPPORT_UTF8 |
| 109 | register USPTR p; |
| 110 | |
| 111 | if (length < 0) |
| 112 | { |
| 113 | for (p = string; *p != 0; p++); |
| 114 | length = (int)(p - string); |
| 115 | } |
| 116 | |
| 117 | for (p = string; length-- > 0; p++) |
| 118 | { |
| 119 | register int ab, c, d; |
| 120 | |
| 121 | c = *p; |
| 122 | if (c < 128) continue; /* ASCII character */ |
| 123 | |
| 124 | if (c < 0xc0) /* Isolated 10xx xxxx byte */ |
| 125 | { |
| 126 | *erroroffset = (int)(p - string); |
| 127 | return PCRE_UTF8_ERR20; |
| 128 | } |
| 129 | |
| 130 | if (c >= 0xfe) /* Invalid 0xfe or 0xff bytes */ |
| 131 | { |
| 132 | *erroroffset = (int)(p - string); |
| 133 | return PCRE_UTF8_ERR21; |
| 134 | } |
| 135 | |
| 136 | ab = _pcre_utf8_table4[c & 0x3f]; /* Number of additional bytes */ |
| 137 | if (length < ab) |
| 138 | { |
| 139 | *erroroffset = (int)(p - string); /* Missing bytes */ |
| 140 | return ab - length; /* Codes ERR1 to ERR5 */ |
| 141 | } |
| 142 | length -= ab; /* Length remaining */ |
| 143 | |
| 144 | /* Check top bits in the second byte */ |
| 145 | |
| 146 | if (((d = *(++p)) & 0xc0) != 0x80) |
| 147 | { |
| 148 | *erroroffset = (int)(p - string) - 1; |
| 149 | return PCRE_UTF8_ERR6; |
| 150 | } |
| 151 | |
| 152 | /* For each length, check that the remaining bytes start with the 0x80 bit |
| 153 | set and not the 0x40 bit. Then check for an overlong sequence, and for the |
| 154 | excluded range 0xd800 to 0xdfff. */ |
| 155 | |
| 156 | switch (ab) |
| 157 | { |
| 158 | /* 2-byte character. No further bytes to check for 0x80. Check first byte |
| 159 | for for xx00 000x (overlong sequence). */ |
| 160 | |
| 161 | case 1: if ((c & 0x3e) == 0) |
| 162 | { |
| 163 | *erroroffset = (int)(p - string) - 1; |
| 164 | return PCRE_UTF8_ERR15; |
| 165 | } |
| 166 | break; |
| 167 | |
| 168 | /* 3-byte character. Check third byte for 0x80. Then check first 2 bytes |
| 169 | for 1110 0000, xx0x xxxx (overlong sequence) or |
| 170 | 1110 1101, 1010 xxxx (0xd800 - 0xdfff) */ |
| 171 | |
| 172 | case 2: |
| 173 | if ((*(++p) & 0xc0) != 0x80) /* Third byte */ |
| 174 | { |
| 175 | *erroroffset = (int)(p - string) - 2; |
| 176 | return PCRE_UTF8_ERR7; |
| 177 | } |
| 178 | if (c == 0xe0 && (d & 0x20) == 0) |
| 179 | { |
| 180 | *erroroffset = (int)(p - string) - 2; |
| 181 | return PCRE_UTF8_ERR16; |
| 182 | } |
| 183 | if (c == 0xed && d >= 0xa0) |
| 184 | { |
| 185 | *erroroffset = (int)(p - string) - 2; |
| 186 | return PCRE_UTF8_ERR14; |
| 187 | } |
| 188 | break; |
| 189 | |
| 190 | /* 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 | |
| 194 | case 3: |
| 195 | if ((*(++p) & 0xc0) != 0x80) /* Third byte */ |
| 196 | { |
| 197 | *erroroffset = (int)(p - string) - 2; |
| 198 | return PCRE_UTF8_ERR7; |
| 199 | } |
| 200 | if ((*(++p) & 0xc0) != 0x80) /* Fourth byte */ |
| 201 | { |
| 202 | *erroroffset = (int)(p - string) - 3; |
| 203 | return PCRE_UTF8_ERR8; |
| 204 | } |
| 205 | if (c == 0xf0 && (d & 0x30) == 0) |
| 206 | { |
| 207 | *erroroffset = (int)(p - string) - 3; |
| 208 | return PCRE_UTF8_ERR17; |
| 209 | } |
| 210 | if (c > 0xf4 || (c == 0xf4 && d > 0x8f)) |
| 211 | { |
| 212 | *erroroffset = (int)(p - string) - 3; |
| 213 | return PCRE_UTF8_ERR13; |
| 214 | } |
| 215 | break; |
| 216 | |
| 217 | /* 5-byte and 6-byte characters are not allowed by RFC 3629, and will be |
| 218 | rejected by the length test below. However, we do the appropriate tests |
| 219 | here so that overlong sequences get diagnosed, and also in case there is |
| 220 | ever an option for handling these larger code points. */ |
| 221 | |
| 222 | /* 5-byte character. Check 3rd, 4th, and 5th bytes for 0x80. Then check for |
| 223 | 1111 1000, xx00 0xxx */ |
| 224 | |
| 225 | case 4: |
| 226 | if ((*(++p) & 0xc0) != 0x80) /* Third byte */ |
| 227 | { |
| 228 | *erroroffset = (int)(p - string) - 2; |
| 229 | return PCRE_UTF8_ERR7; |
| 230 | } |
| 231 | if ((*(++p) & 0xc0) != 0x80) /* Fourth byte */ |
| 232 | { |
| 233 | *erroroffset = (int)(p - string) - 3; |
| 234 | return PCRE_UTF8_ERR8; |
| 235 | } |
| 236 | if ((*(++p) & 0xc0) != 0x80) /* Fifth byte */ |
| 237 | { |
| 238 | *erroroffset = (int)(p - string) - 4; |
| 239 | return PCRE_UTF8_ERR9; |
| 240 | } |
| 241 | if (c == 0xf8 && (d & 0x38) == 0) |
| 242 | { |
| 243 | *erroroffset = (int)(p - string) - 4; |
| 244 | return PCRE_UTF8_ERR18; |
| 245 | } |
| 246 | break; |
| 247 | |
| 248 | /* 6-byte character. Check 3rd-6th bytes for 0x80. Then check for |
| 249 | 1111 1100, xx00 00xx. */ |
| 250 | |
| 251 | case 5: |
| 252 | if ((*(++p) & 0xc0) != 0x80) /* Third byte */ |
| 253 | { |
| 254 | *erroroffset = (int)(p - string) - 2; |
| 255 | return PCRE_UTF8_ERR7; |
| 256 | } |
| 257 | if ((*(++p) & 0xc0) != 0x80) /* Fourth byte */ |
| 258 | { |
| 259 | *erroroffset = (int)(p - string) - 3; |
| 260 | return PCRE_UTF8_ERR8; |
| 261 | } |
| 262 | if ((*(++p) & 0xc0) != 0x80) /* Fifth byte */ |
| 263 | { |
| 264 | *erroroffset = (int)(p - string) - 4; |
| 265 | return PCRE_UTF8_ERR9; |
| 266 | } |
| 267 | if ((*(++p) & 0xc0) != 0x80) /* Sixth byte */ |
| 268 | { |
| 269 | *erroroffset = (int)(p - string) - 5; |
| 270 | return PCRE_UTF8_ERR10; |
| 271 | } |
| 272 | if (c == 0xfc && (d & 0x3c) == 0) |
| 273 | { |
| 274 | *erroroffset = (int)(p - string) - 5; |
| 275 | return PCRE_UTF8_ERR19; |
| 276 | } |
| 277 | break; |
| 278 | } |
| 279 | |
| 280 | /* 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 | |
| 284 | if (ab > 3) |
| 285 | { |
| 286 | *erroroffset = (int)(p - string) - ab; |
| 287 | return (ab == 4)? PCRE_UTF8_ERR11 : PCRE_UTF8_ERR12; |
| 288 | } |
| 289 | } |
| 290 | |
| 291 | #else /* SUPPORT_UTF8 */ |
| 292 | (void)(string); /* Keep picky compilers happy */ |
| 293 | (void)(length); |
| 294 | #endif |
| 295 | |
| 296 | return PCRE_UTF8_ERR0; /* This indicates success */ |
| 297 | } |
| 298 | |
| 299 | /* End of pcre_valid_utf8.c */ |