1 // Copyright 2006-2007 The RE2 Authors. All Rights Reserved.
2 // Use of this source code is governed by a BSD-style
3 // license that can be found in the LICENSE file.
4
5 #include <stdlib.h>
6 #include <vector>
7 #include "util/test.h"
8 #include "re2/prog.h"
9 #include "re2/regexp.h"
10 #include "re2/testing/tester.h"
11 #include "re2/testing/exhaustive_tester.h"
12
13 namespace re2 {
14
15 struct RegexpTest {
16 const char* regexp;
17 const char* text;
18 };
19
20 RegexpTest simple_tests[] = {
21 { "a", "a" },
22 { "a", "zyzzyva" },
23 { "a+", "aa" },
24 { "(a+|b)+", "ab" },
25 { "ab|cd", "xabcdx" },
26 { "h.*od?", "hello\ngoodbye\n" },
27 { "h.*o", "hello\ngoodbye\n" },
28 { "h.*o", "goodbye\nhello\n" },
29 { "h.*o", "hello world" },
30 { "h.*o", "othello, world" },
31 { "[^\\s\\S]", "aaaaaaa" },
32 { "a", "aaaaaaa" },
33 { "a*", "aaaaaaa" },
34 { "a*", "" },
35 { "a*", NULL },
36 { "ab|cd", "xabcdx" },
37 { "a", "cab" },
38 { "a*b", "cab" },
39 { "((((((((((((((((((((x))))))))))))))))))))", "x" },
40 { "[abcd]", "xxxabcdxxx" },
41 { "[^x]", "xxxabcdxxx" },
42 { "[abcd]+", "xxxabcdxxx" },
43 { "[^x]+", "xxxabcdxxx" },
44 { "(fo|foo)", "fo" },
45 { "(foo|fo)", "foo" },
46
47 { "aa", "aA" },
48 { "a", "Aa" },
49 { "a", "A" },
50 { "ABC", "abc" },
51 { "abc", "XABCY" },
52 { "ABC", "xabcy" },
53
54 // Make sure ^ and $ work.
55 // The pathological cases didn't work
56 // in the original grep code.
57 { "foo|bar|[A-Z]", "foo" },
58 { "^(foo|bar|[A-Z])", "foo" },
59 { "(foo|bar|[A-Z])$", "foo\n" },
60 { "(foo|bar|[A-Z])$", "foo" },
61 { "^(foo|bar|[A-Z])$", "foo\n" },
62 { "^(foo|bar|[A-Z])$", "foo" },
63 { "^(foo|bar|[A-Z])$", "bar" },
64 { "^(foo|bar|[A-Z])$", "X" },
65 { "^(foo|bar|[A-Z])$", "XY" },
66 { "^(fo|foo)$", "fo" },
67 { "^(fo|foo)$", "foo" },
68 { "^^(fo|foo)$", "fo" },
69 { "^^(fo|foo)$", "foo" },
70 { "^$", "" },
71 { "^$", "x" },
72 { "^^$", "" },
73 { "^$$", "" },
74 { "^^$", "x" },
75 { "^$$", "x" },
76 { "^^$$", "" },
77 { "^^$$", "x" },
78 { "^^^^^^^^$$$$$$$$", "" },
79 { "^", "x" },
80 { "$", "x" },
81
82 // Word boundaries.
83 { "\\bfoo\\b", "nofoo foo that" },
84 { "a\\b", "faoa x" },
85 { "\\bbar", "bar x" },
86 { "\\bbar", "foo\nbar x" },
87 { "bar\\b", "foobar" },
88 { "bar\\b", "foobar\nxxx" },
89 { "(foo|bar|[A-Z])\\b", "foo" },
90 { "(foo|bar|[A-Z])\\b", "foo\n" },
91 { "\\b", "" },
92 { "\\b", "x" },
93 { "\\b(foo|bar|[A-Z])", "foo" },
94 { "\\b(foo|bar|[A-Z])\\b", "X" },
95 { "\\b(foo|bar|[A-Z])\\b", "XY" },
96 { "\\b(foo|bar|[A-Z])\\b", "bar" },
97 { "\\b(foo|bar|[A-Z])\\b", "foo" },
98 { "\\b(foo|bar|[A-Z])\\b", "foo\n" },
99 { "\\b(foo|bar|[A-Z])\\b", "ffoo bbar N x" },
100 { "\\b(fo|foo)\\b", "fo" },
101 { "\\b(fo|foo)\\b", "foo" },
102 { "\\b\\b", "" },
103 { "\\b\\b", "x" },
104 { "\\b$", "" },
105 { "\\b$", "x" },
106 { "\\b$", "y x" },
107 { "\\b.$", "x" },
108 { "^\\b(fo|foo)\\b", "fo" },
109 { "^\\b(fo|foo)\\b", "foo" },
110 { "^\\b", "" },
111 { "^\\b", "x" },
112 { "^\\b\\b", "" },
113 { "^\\b\\b", "x" },
114 { "^\\b$", "" },
115 { "^\\b$", "x" },
116 { "^\\b.$", "x" },
117 { "^\\b.\\b$", "x" },
118 { "^^^^^^^^\\b$$$$$$$", "" },
119 { "^^^^^^^^\\b.$$$$$$", "x" },
120 { "^^^^^^^^\\b$$$$$$$", "x" },
121
122 // Non-word boundaries.
123 { "\\Bfoo\\B", "n foo xfoox that" },
124 { "a\\B", "faoa x" },
125 { "\\Bbar", "bar x" },
126 { "\\Bbar", "foo\nbar x" },
127 { "bar\\B", "foobar" },
128 { "bar\\B", "foobar\nxxx" },
129 { "(foo|bar|[A-Z])\\B", "foox" },
130 { "(foo|bar|[A-Z])\\B", "foo\n" },
131 { "\\B", "" },
132 { "\\B", "x" },
133 { "\\B(foo|bar|[A-Z])", "foo" },
134 { "\\B(foo|bar|[A-Z])\\B", "xXy" },
135 { "\\B(foo|bar|[A-Z])\\B", "XY" },
136 { "\\B(foo|bar|[A-Z])\\B", "XYZ" },
137 { "\\B(foo|bar|[A-Z])\\B", "abara" },
138 { "\\B(foo|bar|[A-Z])\\B", "xfoo_" },
139 { "\\B(foo|bar|[A-Z])\\B", "xfoo\n" },
140 { "\\B(foo|bar|[A-Z])\\B", "foo bar vNx" },
141 { "\\B(fo|foo)\\B", "xfoo" },
142 { "\\B(foo|fo)\\B", "xfooo" },
143 { "\\B\\B", "" },
144 { "\\B\\B", "x" },
145 { "\\B$", "" },
146 { "\\B$", "x" },
147 { "\\B$", "y x" },
148 { "\\B.$", "x" },
149 { "^\\B(fo|foo)\\B", "fo" },
150 { "^\\B(fo|foo)\\B", "foo" },
151 { "^\\B", "" },
152 { "^\\B", "x" },
153 { "^\\B\\B", "" },
154 { "^\\B\\B", "x" },
155 { "^\\B$", "" },
156 { "^\\B$", "x" },
157 { "^\\B.$", "x" },
158 { "^\\B.\\B$", "x" },
159 { "^^^^^^^^\\B$$$$$$$", "" },
160 { "^^^^^^^^\\B.$$$$$$", "x" },
161 { "^^^^^^^^\\B$$$$$$$", "x" },
162
163 // PCRE uses only ASCII for \b computation.
164 // All non-ASCII are *not* word characters.
165 { "\\bx\\b", "x" },
166 { "\\bx\\b", "x>" },
167 { "\\bx\\b", "<x" },
168 { "\\bx\\b", "<x>" },
169 { "\\bx\\b", "ax" },
170 { "\\bx\\b", "xb" },
171 { "\\bx\\b", "axb" },
172 { "\\bx\\b", "«x" },
173 { "\\bx\\b", "x»" },
174 { "\\bx\\b", "«x»" },
175 { "\\bx\\b", "axb" },
176 { "\\bx\\b", "áxβ" },
177 { "\\Bx\\B", "axb" },
178 { "\\Bx\\B", "áxβ" },
179
180 // Weird boundary cases.
181 { "^$^$", "" },
182 { "^$^", "" },
183 { "$^$", "" },
184
185 { "^$^$", "x" },
186 { "^$^", "x" },
187 { "$^$", "x" },
188
189 { "^$^$", "x\ny" },
190 { "^$^", "x\ny" },
191 { "$^$", "x\ny" },
192
193 { "^$^$", "x\n\ny" },
194 { "^$^", "x\n\ny" },
195 { "$^$", "x\n\ny" },
196
197 { "^(foo\\$)$", "foo$bar" },
198 { "(foo\\$)", "foo$bar" },
199 { "^...$", "abc" },
200
201 // UTF-8
202 { "^\xe6\x9c\xac$", "\xe6\x9c\xac" },
203 { "^...$", "\xe6\x97\xa5\xe6\x9c\xac\xe8\xaa\x9e" },
204 { "^...$", ".\xe6\x9c\xac." },
205
206 { "^\\C\\C\\C$", "\xe6\x9c\xac" },
207 { "^\\C$", "\xe6\x9c\xac" },
208 { "^\\C\\C\\C$", "\xe6\x97\xa5\xe6\x9c\xac\xe8\xaa\x9e" },
209
210 // Latin1
211 { "^...$", "\xe6\x97\xa5\xe6\x9c\xac\xe8\xaa\x9e" },
212 { "^.........$", "\xe6\x97\xa5\xe6\x9c\xac\xe8\xaa\x9e" },
213 { "^...$", ".\xe6\x9c\xac." },
214 { "^.....$", ".\xe6\x9c\xac." },
215
216 // Perl v Posix
217 { "\\B(fo|foo)\\B", "xfooo" },
218 { "(fo|foo)", "foo" },
219
220 // Octal escapes.
221 { "\\141", "a" },
222 { "\\060", "0" },
223 { "\\0600", "00" },
224 { "\\608", "08" },
225 { "\\01", "\01" },
226 { "\\018", "\01" "8" },
227
228 // Hexadecimal escapes
229 { "\\x{61}", "a" },
230 { "\\x61", "a" },
231 { "\\x{00000061}", "a" },
232
233 // Unicode scripts.
234 { "\\p{Greek}+", "aαβb" },
235 { "\\P{Greek}+", "aαβb" },
236 { "\\p{^Greek}+", "aαβb" },
237 { "\\P{^Greek}+", "aαβb" },
238
239 // Unicode properties. Nd is decimal number. N is any number.
240 { "[^0-9]+", "abc123" },
241 { "\\p{Nd}+", "abc123²³¼½¾₀₉" },
242 { "\\p{^Nd}+", "abc123²³¼½¾₀₉" },
243 { "\\P{Nd}+", "abc123²³¼½¾₀₉" },
244 { "\\P{^Nd}+", "abc123²³¼½¾₀₉" },
245 { "\\pN+", "abc123²³¼½¾₀₉" },
246 { "\\p{N}+", "abc123²³¼½¾₀₉" },
247 { "\\p{^N}+", "abc123²³¼½¾₀₉" },
248
249 { "\\p{Any}+", "abc123" },
250
251 // Character classes & case folding.
252 { "(?i)[@-A]+", "@AaB" }, // matches @Aa but not B
253 { "(?i)[A-Z]+", "aAzZ" },
254 { "(?i)[^\\\\]+", "Aa\\" }, // \\ is between A-Z and a-z -
255 // splits the ranges in an interesting way.
256
257 // would like to use, but PCRE mishandles in full-match, non-greedy mode
258 // { "(?i)[\\\\]+", "Aa" },
259
260 { "(?i)[acegikmoqsuwy]+", "acegikmoqsuwyACEGIKMOQSUWY" },
261
262 // Character classes & case folding.
263 { "[@-A]+", "@AaB" },
264 { "[A-Z]+", "aAzZ" },
265 { "[^\\\\]+", "Aa\\" },
266 { "[acegikmoqsuwy]+", "acegikmoqsuwyACEGIKMOQSUWY" },
267
268 // Anchoring. (^abc in aabcdef was a former bug)
269 // The tester checks for a match in the text and
270 // subpieces of the text with a byte removed on either side.
271 { "^abc", "abcdef" },
272 { "^abc", "aabcdef" },
273 { "^[ay]*[bx]+c", "abcdef" },
274 { "^[ay]*[bx]+c", "aabcdef" },
275 { "def$", "abcdef" },
276 { "def$", "abcdeff" },
277 { "d[ex][fy]$", "abcdef" },
278 { "d[ex][fy]$", "abcdeff" },
279 { "[dz][ex][fy]$", "abcdef" },
280 { "[dz][ex][fy]$", "abcdeff" },
281 { "(?m)^abc", "abcdef" },
282 { "(?m)^abc", "aabcdef" },
283 { "(?m)^[ay]*[bx]+c", "abcdef" },
284 { "(?m)^[ay]*[bx]+c", "aabcdef" },
285 { "(?m)def$", "abcdef" },
286 { "(?m)def$", "abcdeff" },
287 { "(?m)d[ex][fy]$", "abcdef" },
288 { "(?m)d[ex][fy]$", "abcdeff" },
289 { "(?m)[dz][ex][fy]$", "abcdef" },
290 { "(?m)[dz][ex][fy]$", "abcdeff" },
291 { "^", "a" },
292 { "^^", "a" },
293
294 // Context.
295 // The tester checks for a match in the text and
296 // subpieces of the text with a byte removed on either side.
297 { "a", "a" },
298 { "ab*", "a" },
299 { "a\\C*", "a" },
300
301 // Former bugs.
302 { "a\\C*|ba\\C", "baba" },
303 };
304
TEST(Regexp,SearchTests)305 TEST(Regexp, SearchTests) {
306 int failures = 0;
307 for (int i = 0; i < arraysize(simple_tests); i++) {
308 const RegexpTest& t = simple_tests[i];
309 if (!TestRegexpOnText(t.regexp, t.text))
310 failures++;
311
312 #ifdef LOGGING
313 // Build a dummy ExhaustiveTest call that will trigger just
314 // this one test, so that we log the test case.
315 vector<string> atom, alpha, ops;
316 atom.push_back(StringPiece(t.regexp).as_string());
317 alpha.push_back(StringPiece(t.text).as_string());
318 ExhaustiveTest(1, 0, atom, ops, 1, alpha, "", "");
319 #endif
320
321 }
322 EXPECT_EQ(failures, 0);
323 }
324
325 } // namespace re2
326