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 that is used to match an extended |
42 | class. It is used by both pcre_exec() and pcre_def_exec(). */ |
43 | |
44 | #include "pcre_config.h" |
45 | #include "pcre_internal.h" |
46 | |
47 | |
48 | /************************************************* |
49 | * Match character against an XCLASS * |
50 | *************************************************/ |
51 | |
52 | /* This function is called to match a character against an extended class that |
53 | might contain values > 255 and/or Unicode properties. |
54 | |
55 | Arguments: |
56 | c the character |
57 | data points to the flag byte of the XCLASS data |
58 | |
59 | Returns: TRUE if character matches, else FALSE |
60 | */ |
61 | |
62 | BOOL |
63 | PRIV(xclass)(pcre_uint32 c, const pcre_uchar *data, BOOL utf) |
64 | { |
65 | pcre_uchar t; |
66 | BOOL negated = (*data & XCL_NOT) != 0; |
67 | |
68 | (void)utf; |
69 | #ifdef COMPILE_PCRE8 |
70 | /* In 8 bit mode, this must always be TRUE. Help the compiler to know that. */ |
71 | utf = TRUE; |
72 | #endif |
73 | |
74 | /* Character values < 256 are matched against a bitmap, if one is present. If |
75 | not, we still carry on, because there may be ranges that start below 256 in the |
76 | additional data. */ |
77 | |
78 | if (c < 256) |
79 | { |
80 | if ((*data & XCL_HASPROP) == 0) |
81 | { |
82 | if ((*data & XCL_MAP) == 0) return negated; |
83 | return (((pcre_uint8 *)(data + 1))[c/8] & (1 << (c&7))) != 0; |
84 | } |
85 | if ((*data & XCL_MAP) != 0 && |
86 | (((pcre_uint8 *)(data + 1))[c/8] & (1 << (c&7))) != 0) |
87 | return !negated; /* char found */ |
88 | } |
89 | |
90 | /* First skip the bit map if present. Then match against the list of Unicode |
91 | properties or large chars or ranges that end with a large char. We won't ever |
92 | encounter XCL_PROP or XCL_NOTPROP when UCP support is not compiled. */ |
93 | |
94 | if ((*data++ & XCL_MAP) != 0) data += 32 / sizeof(pcre_uchar); |
95 | |
96 | while ((t = *data++) != XCL_END) |
97 | { |
98 | pcre_uint32 x, y; |
99 | if (t == XCL_SINGLE) |
100 | { |
101 | #ifdef SUPPORT_UTF |
102 | if (utf) |
103 | { |
104 | GETCHARINC(x, data); /* macro generates multiple statements */ |
105 | } |
106 | else |
107 | #endif |
108 | x = *data++; |
109 | if (c == x) return !negated; |
110 | } |
111 | else if (t == XCL_RANGE) |
112 | { |
113 | #ifdef SUPPORT_UTF |
114 | if (utf) |
115 | { |
116 | GETCHARINC(x, data); /* macro generates multiple statements */ |
117 | GETCHARINC(y, data); /* macro generates multiple statements */ |
118 | } |
119 | else |
120 | #endif |
121 | { |
122 | x = *data++; |
123 | y = *data++; |
124 | } |
125 | if (c >= x && c <= y) return !negated; |
126 | } |
127 | |
128 | #ifdef SUPPORT_UCP |
129 | else /* XCL_PROP & XCL_NOTPROP */ |
130 | { |
131 | const ucd_record *prop = GET_UCD(c); |
132 | BOOL isprop = t == XCL_PROP; |
133 | |
134 | switch(*data) |
135 | { |
136 | case PT_ANY: |
137 | if (isprop) return !negated; |
138 | break; |
139 | |
140 | case PT_LAMP: |
141 | if ((prop->chartype == ucp_Lu || prop->chartype == ucp_Ll || |
142 | prop->chartype == ucp_Lt) == isprop) return !negated; |
143 | break; |
144 | |
145 | case PT_GC: |
146 | if ((data[1] == PRIV(ucp_gentype)[prop->chartype]) == isprop) |
147 | return !negated; |
148 | break; |
149 | |
150 | case PT_PC: |
151 | if ((data[1] == prop->chartype) == isprop) return !negated; |
152 | break; |
153 | |
154 | case PT_SC: |
155 | if ((data[1] == prop->script) == isprop) return !negated; |
156 | break; |
157 | |
158 | case PT_ALNUM: |
159 | if ((PRIV(ucp_gentype)[prop->chartype] == ucp_L || |
160 | PRIV(ucp_gentype)[prop->chartype] == ucp_N) == isprop) |
161 | return !negated; |
162 | break; |
163 | |
164 | /* Perl space used to exclude VT, but from Perl 5.18 it is included, |
165 | which means that Perl space and POSIX space are now identical. PCRE |
166 | was changed at release 8.34. */ |
167 | |
168 | case PT_SPACE: /* Perl space */ |
169 | case PT_PXSPACE: /* POSIX space */ |
170 | switch(c) |
171 | { |
172 | HSPACE_CASES: |
173 | VSPACE_CASES: |
174 | if (isprop) return !negated; |
175 | break; |
176 | |
177 | default: |
178 | if ((PRIV(ucp_gentype)[prop->chartype] == ucp_Z) == isprop) |
179 | return !negated; |
180 | break; |
181 | } |
182 | break; |
183 | |
184 | case PT_WORD: |
185 | if ((PRIV(ucp_gentype)[prop->chartype] == ucp_L || |
186 | PRIV(ucp_gentype)[prop->chartype] == ucp_N || c == CHAR_UNDERSCORE) |
187 | == isprop) |
188 | return !negated; |
189 | break; |
190 | |
191 | case PT_UCNC: |
192 | if (c < 0xa0) |
193 | { |
194 | if ((c == CHAR_DOLLAR_SIGN || c == CHAR_COMMERCIAL_AT || |
195 | c == CHAR_GRAVE_ACCENT) == isprop) |
196 | return !negated; |
197 | } |
198 | else |
199 | { |
200 | if ((c < 0xd800 || c > 0xdfff) == isprop) |
201 | return !negated; |
202 | } |
203 | break; |
204 | |
205 | /* The following three properties can occur only in an XCLASS, as there |
206 | is no \p or \P coding for them. */ |
207 | |
208 | /* Graphic character. Implement this as not Z (space or separator) and |
209 | not C (other), except for Cf (format) with a few exceptions. This seems |
210 | to be what Perl does. The exceptional characters are: |
211 | |
212 | U+061C Arabic Letter Mark |
213 | U+180E Mongolian Vowel Separator |
214 | U+2066 - U+2069 Various "isolate"s |
215 | */ |
216 | |
217 | case PT_PXGRAPH: |
218 | if ((PRIV(ucp_gentype)[prop->chartype] != ucp_Z && |
219 | (PRIV(ucp_gentype)[prop->chartype] != ucp_C || |
220 | (prop->chartype == ucp_Cf && |
221 | c != 0x061c && c != 0x180e && (c < 0x2066 || c > 0x2069)) |
222 | )) == isprop) |
223 | return !negated; |
224 | break; |
225 | |
226 | /* Printable character: same as graphic, with the addition of Zs, i.e. |
227 | not Zl and not Zp, and U+180E. */ |
228 | |
229 | case PT_PXPRINT: |
230 | if ((prop->chartype != ucp_Zl && |
231 | prop->chartype != ucp_Zp && |
232 | (PRIV(ucp_gentype)[prop->chartype] != ucp_C || |
233 | (prop->chartype == ucp_Cf && |
234 | c != 0x061c && (c < 0x2066 || c > 0x2069)) |
235 | )) == isprop) |
236 | return !negated; |
237 | break; |
238 | |
239 | /* Punctuation: all Unicode punctuation, plus ASCII characters that |
240 | Unicode treats as symbols rather than punctuation, for Perl |
241 | compatibility (these are $+<=>^`|~). */ |
242 | |
243 | case PT_PXPUNCT: |
244 | if ((PRIV(ucp_gentype)[prop->chartype] == ucp_P || |
245 | (c < 128 && PRIV(ucp_gentype)[prop->chartype] == ucp_S)) == isprop) |
246 | return !negated; |
247 | break; |
248 | |
249 | /* This should never occur, but compilers may mutter if there is no |
250 | default. */ |
251 | |
252 | default: |
253 | return FALSE; |
254 | } |
255 | |
256 | data += 2; |
257 | } |
258 | #endif /* SUPPORT_UCP */ |
259 | } |
260 | |
261 | return negated; /* char did not match */ |
262 | } |
263 | |
264 | /* End of pcre_xclass.c */ |
265 | |