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 "util/test.h"
6 #include "re2/prog.h"
7 #include "re2/regexp.h"
8 #include "re2/testing/tester.h"
9 #include "re2/testing/exhaustive_tester.h"
10 
11 // For target `log' in the Makefile.
12 #ifndef LOGGING
13 #define LOGGING 0
14 #endif
15 
16 namespace re2 {
17 
18 struct RegexpTest {
19   const char* regexp;
20   const char* text;
21 };
22 
23 RegexpTest simple_tests[] = {
24   { "a", "a" },
25   { "a", "zyzzyva" },
26   { "a+", "aa" },
27   { "(a+|b)+", "ab" },
28   { "ab|cd", "xabcdx" },
29   { "h.*od?", "hello\ngoodbye\n" },
30   { "h.*o", "hello\ngoodbye\n" },
31   { "h.*o", "goodbye\nhello\n" },
32   { "h.*o", "hello world" },
33   { "h.*o", "othello, world" },
34   { "[^\\s\\S]", "aaaaaaa" },
35   { "a", "aaaaaaa" },
36   { "a*", "aaaaaaa" },
37   { "a*", "" },
38   { "ab|cd", "xabcdx" },
39   { "a", "cab" },
40   { "a*b", "cab" },
41   { "((((((((((((((((((((x))))))))))))))))))))", "x" },
42   { "[abcd]", "xxxabcdxxx" },
43   { "[^x]", "xxxabcdxxx" },
44   { "[abcd]+", "xxxabcdxxx" },
45   { "[^x]+", "xxxabcdxxx" },
46   { "(fo|foo)", "fo" },
47   { "(foo|fo)", "foo" },
48 
49   { "aa", "aA" },
50   { "a", "Aa" },
51   { "a", "A" },
52   { "ABC", "abc" },
53   { "abc", "XABCY" },
54   { "ABC", "xabcy" },
55 
56   // Make sure ^ and $ work.
57   // The pathological cases didn't work
58   // in the original grep code.
59   { "foo|bar|[A-Z]", "foo" },
60   { "^(foo|bar|[A-Z])", "foo" },
61   { "(foo|bar|[A-Z])$", "foo\n" },
62   { "(foo|bar|[A-Z])$", "foo" },
63   { "^(foo|bar|[A-Z])$", "foo\n" },
64   { "^(foo|bar|[A-Z])$", "foo" },
65   { "^(foo|bar|[A-Z])$", "bar" },
66   { "^(foo|bar|[A-Z])$", "X" },
67   { "^(foo|bar|[A-Z])$", "XY" },
68   { "^(fo|foo)$", "fo" },
69   { "^(fo|foo)$", "foo" },
70   { "^^(fo|foo)$", "fo" },
71   { "^^(fo|foo)$", "foo" },
72   { "^$", "" },
73   { "^$", "x" },
74   { "^^$", "" },
75   { "^$$", "" },
76   { "^^$", "x" },
77   { "^$$", "x" },
78   { "^^$$", "" },
79   { "^^$$", "x" },
80   { "^^^^^^^^$$$$$$$$", "" },
81   { "^", "x" },
82   { "$", "x" },
83 
84   // Word boundaries.
85   { "\\bfoo\\b", "nofoo foo that" },
86   { "a\\b", "faoa x" },
87   { "\\bbar", "bar x" },
88   { "\\bbar", "foo\nbar x" },
89   { "bar\\b", "foobar" },
90   { "bar\\b", "foobar\nxxx" },
91   { "(foo|bar|[A-Z])\\b", "foo" },
92   { "(foo|bar|[A-Z])\\b", "foo\n" },
93   { "\\b", "" },
94   { "\\b", "x" },
95   { "\\b(foo|bar|[A-Z])", "foo" },
96   { "\\b(foo|bar|[A-Z])\\b", "X" },
97   { "\\b(foo|bar|[A-Z])\\b", "XY" },
98   { "\\b(foo|bar|[A-Z])\\b", "bar" },
99   { "\\b(foo|bar|[A-Z])\\b", "foo" },
100   { "\\b(foo|bar|[A-Z])\\b", "foo\n" },
101   { "\\b(foo|bar|[A-Z])\\b", "ffoo bbar N x" },
102   { "\\b(fo|foo)\\b", "fo" },
103   { "\\b(fo|foo)\\b", "foo" },
104   { "\\b\\b", "" },
105   { "\\b\\b", "x" },
106   { "\\b$", "" },
107   { "\\b$", "x" },
108   { "\\b$", "y x" },
109   { "\\b.$", "x" },
110   { "^\\b(fo|foo)\\b", "fo" },
111   { "^\\b(fo|foo)\\b", "foo" },
112   { "^\\b", "" },
113   { "^\\b", "x" },
114   { "^\\b\\b", "" },
115   { "^\\b\\b", "x" },
116   { "^\\b$", "" },
117   { "^\\b$", "x" },
118   { "^\\b.$", "x" },
119   { "^\\b.\\b$", "x" },
120   { "^^^^^^^^\\b$$$$$$$", "" },
121   { "^^^^^^^^\\b.$$$$$$", "x" },
122   { "^^^^^^^^\\b$$$$$$$", "x" },
123 
124   // Non-word boundaries.
125   { "\\Bfoo\\B", "n foo xfoox that" },
126   { "a\\B", "faoa x" },
127   { "\\Bbar", "bar x" },
128   { "\\Bbar", "foo\nbar x" },
129   { "bar\\B", "foobar" },
130   { "bar\\B", "foobar\nxxx" },
131   { "(foo|bar|[A-Z])\\B", "foox" },
132   { "(foo|bar|[A-Z])\\B", "foo\n" },
133   { "\\B", "" },
134   { "\\B", "x" },
135   { "\\B(foo|bar|[A-Z])", "foo" },
136   { "\\B(foo|bar|[A-Z])\\B", "xXy" },
137   { "\\B(foo|bar|[A-Z])\\B", "XY" },
138   { "\\B(foo|bar|[A-Z])\\B", "XYZ" },
139   { "\\B(foo|bar|[A-Z])\\B", "abara" },
140   { "\\B(foo|bar|[A-Z])\\B", "xfoo_" },
141   { "\\B(foo|bar|[A-Z])\\B", "xfoo\n" },
142   { "\\B(foo|bar|[A-Z])\\B", "foo bar vNx" },
143   { "\\B(fo|foo)\\B", "xfoo" },
144   { "\\B(foo|fo)\\B", "xfooo" },
145   { "\\B\\B", "" },
146   { "\\B\\B", "x" },
147   { "\\B$", "" },
148   { "\\B$", "x" },
149   { "\\B$", "y x" },
150   { "\\B.$", "x" },
151   { "^\\B(fo|foo)\\B", "fo" },
152   { "^\\B(fo|foo)\\B", "foo" },
153   { "^\\B", "" },
154   { "^\\B", "x" },
155   { "^\\B\\B", "" },
156   { "^\\B\\B", "x" },
157   { "^\\B$", "" },
158   { "^\\B$", "x" },
159   { "^\\B.$", "x" },
160   { "^\\B.\\B$", "x" },
161   { "^^^^^^^^\\B$$$$$$$", "" },
162   { "^^^^^^^^\\B.$$$$$$", "x" },
163   { "^^^^^^^^\\B$$$$$$$", "x" },
164 
165   // PCRE uses only ASCII for \b computation.
166   // All non-ASCII are *not* word characters.
167   { "\\bx\\b", "x" },
168   { "\\bx\\b", "x>" },
169   { "\\bx\\b", "<x" },
170   { "\\bx\\b", "<x>" },
171   { "\\bx\\b", "ax" },
172   { "\\bx\\b", "xb" },
173   { "\\bx\\b", "axb" },
174   { "\\bx\\b", "«x" },
175   { "\\bx\\b", "x»" },
176   { "\\bx\\b", "«x»" },
177   { "\\bx\\b", "axb" },
178   { "\\bx\\b", "áxβ" },
179   { "\\Bx\\B", "axb" },
180   { "\\Bx\\B", "áxβ" },
181 
182   // Weird boundary cases.
183   { "^$^$", "" },
184   { "^$^", "" },
185   { "$^$", "" },
186 
187   { "^$^$", "x" },
188   { "^$^", "x" },
189   { "$^$", "x" },
190 
191   { "^$^$", "x\ny" },
192   { "^$^", "x\ny" },
193   { "$^$", "x\ny" },
194 
195   { "^$^$", "x\n\ny" },
196   { "^$^", "x\n\ny" },
197   { "$^$", "x\n\ny" },
198 
199   { "^(foo\\$)$", "foo$bar" },
200   { "(foo\\$)", "foo$bar" },
201   { "^...$", "abc" },
202 
203   // UTF-8
204   { "^\xe6\x9c\xac$", "\xe6\x9c\xac" },
205   { "^...$", "\xe6\x97\xa5\xe6\x9c\xac\xe8\xaa\x9e" },
206   { "^...$", ".\xe6\x9c\xac." },
207 
208   { "^\\C\\C\\C$", "\xe6\x9c\xac" },
209   { "^\\C$", "\xe6\x9c\xac" },
210   { "^\\C\\C\\C$", "\xe6\x97\xa5\xe6\x9c\xac\xe8\xaa\x9e" },
211 
212   // Latin1
213   { "^...$", "\xe6\x97\xa5\xe6\x9c\xac\xe8\xaa\x9e" },
214   { "^.........$", "\xe6\x97\xa5\xe6\x9c\xac\xe8\xaa\x9e" },
215   { "^...$", ".\xe6\x9c\xac." },
216   { "^.....$", ".\xe6\x9c\xac." },
217 
218   // Perl v Posix
219   { "\\B(fo|foo)\\B", "xfooo" },
220   { "(fo|foo)", "foo" },
221 
222   // Octal escapes.
223   { "\\141", "a" },
224   { "\\060", "0" },
225   { "\\0600", "00" },
226   { "\\608", "08" },
227   { "\\01", "\01" },
228   { "\\018", "\01" "8" },
229 
230   // Hexadecimal escapes
231   { "\\x{61}", "a" },
232   { "\\x61", "a" },
233   { "\\x{00000061}", "a" },
234 
235   // Unicode scripts.
236   { "\\p{Greek}+", "aαβb" },
237   { "\\P{Greek}+", "aαβb" },
238   { "\\p{^Greek}+", "aαβb" },
239   { "\\P{^Greek}+", "aαβb" },
240 
241   // Unicode properties.  Nd is decimal number.  N is any number.
242   { "[^0-9]+",  "abc123" },
243   { "\\p{Nd}+", "abc123²³¼½¾₀₉" },
244   { "\\p{^Nd}+", "abc123²³¼½¾₀₉" },
245   { "\\P{Nd}+", "abc123²³¼½¾₀₉" },
246   { "\\P{^Nd}+", "abc123²³¼½¾₀₉" },
247   { "\\pN+", "abc123²³¼½¾₀₉" },
248   { "\\p{N}+", "abc123²³¼½¾₀₉" },
249   { "\\p{^N}+", "abc123²³¼½¾₀₉" },
250 
251   { "\\p{Any}+", "abc123" },
252 
253   // Character classes & case folding.
254   { "(?i)[@-A]+", "@AaB" },  // matches @Aa but not B
255   { "(?i)[A-Z]+", "aAzZ" },
256   { "(?i)[^\\\\]+", "Aa\\" },  // \\ is between A-Z and a-z -
257                                // splits the ranges in an interesting way.
258 
259   // would like to use, but PCRE mishandles in full-match, non-greedy mode
260   // { "(?i)[\\\\]+", "Aa" },
261 
262   { "(?i)[acegikmoqsuwy]+", "acegikmoqsuwyACEGIKMOQSUWY" },
263 
264   // Character classes & case folding.
265   { "[@-A]+", "@AaB" },
266   { "[A-Z]+", "aAzZ" },
267   { "[^\\\\]+", "Aa\\" },
268   { "[acegikmoqsuwy]+", "acegikmoqsuwyACEGIKMOQSUWY" },
269 
270   // Anchoring.  (^abc in aabcdef was a former bug)
271   // The tester checks for a match in the text and
272   // subpieces of the text with a byte removed on either side.
273   { "^abc", "abcdef" },
274   { "^abc", "aabcdef" },
275   { "^[ay]*[bx]+c", "abcdef" },
276   { "^[ay]*[bx]+c", "aabcdef" },
277   { "def$", "abcdef" },
278   { "def$", "abcdeff" },
279   { "d[ex][fy]$", "abcdef" },
280   { "d[ex][fy]$", "abcdeff" },
281   { "[dz][ex][fy]$", "abcdef" },
282   { "[dz][ex][fy]$", "abcdeff" },
283   { "(?m)^abc", "abcdef" },
284   { "(?m)^abc", "aabcdef" },
285   { "(?m)^[ay]*[bx]+c", "abcdef" },
286   { "(?m)^[ay]*[bx]+c", "aabcdef" },
287   { "(?m)def$", "abcdef" },
288   { "(?m)def$", "abcdeff" },
289   { "(?m)d[ex][fy]$", "abcdef" },
290   { "(?m)d[ex][fy]$", "abcdeff" },
291   { "(?m)[dz][ex][fy]$", "abcdef" },
292   { "(?m)[dz][ex][fy]$", "abcdeff" },
293   { "^", "a" },
294   { "^^", "a" },
295 
296   // Context.
297   // The tester checks for a match in the text and
298   // subpieces of the text with a byte removed on either side.
299   { "a", "a" },
300   { "ab*", "a" },
301   { "a\\C*", "a" },
302   { "a\\C+", "a" },
303   { "a\\C?", "a" },
304   { "a\\C*?", "a" },
305   { "a\\C+?", "a" },
306   { "a\\C??", "a" },
307 
308   // Former bugs.
309   { "a\\C*|ba\\C", "baba" },
310   { "\\w*I\\w*", "Inc." },
311 };
312 
TEST(Regexp,SearchTests)313 TEST(Regexp, SearchTests) {
314   int failures = 0;
315   for (size_t i = 0; i < arraysize(simple_tests); i++) {
316     const RegexpTest& t = simple_tests[i];
317     if (!TestRegexpOnText(t.regexp, t.text))
318       failures++;
319 
320     if (LOGGING) {
321       // Build a dummy ExhaustiveTest call that will trigger just
322       // this one test, so that we log the test case.
323       std::vector<std::string> atom, alpha, ops;
324       atom.push_back(t.regexp);
325       alpha.push_back(t.text);
326       ExhaustiveTest(1, 0, atom, ops, 1, alpha, "", "");
327     }
328   }
329   EXPECT_EQ(failures, 0);
330 }
331 
332 }  // namespace re2
333