]>
Commit | Line | Data |
---|---|---|
1 | /////////////////////////////////////////////////////////////////////////////// | |
2 | // Name: tests/regex/regex.cpp | |
3 | // Purpose: Test the built-in regex lib and wxRegEx | |
4 | // Author: Mike Wetherell | |
5 | // RCS-ID: $Id$ | |
6 | // Copyright: (c) 2004 Mike Wetherell | |
7 | // Licence: wxWindows licence | |
8 | /////////////////////////////////////////////////////////////////////////////// | |
9 | ||
10 | // | |
11 | // Notes: | |
12 | // | |
13 | // To run just one section, say wx_1, do this: | |
14 | // test regex.wx_1 | |
15 | // | |
16 | // To run all the regex tests: | |
17 | // test regex | |
18 | // | |
19 | // Some tests must be skipped since they use features which we do not make | |
20 | // available through wxRegEx. To see the list of tests that have been skipped | |
21 | // turn on verbose logging, e.g.: | |
22 | // test --verbose regex | |
23 | // | |
24 | // The tests here are for the builtin library, tests for wxRegEx in general | |
25 | // should go in wxregex.cpp | |
26 | // | |
27 | // The tests are generated from Henry Spencer's reg.test, additional test | |
28 | // can be added in wxreg.test. These test files are then turned into a C++ | |
29 | // include file 'regex.inc' (included below) using a script 'regex.pl'. | |
30 | // | |
31 | ||
32 | // For compilers that support precompilation, includes "wx/wx.h". | |
33 | #include "testprec.h" | |
34 | ||
35 | #ifdef __BORLANDC__ | |
36 | #pragma hdrstop | |
37 | #endif | |
38 | ||
39 | #if wxUSE_REGEX | |
40 | ||
41 | // for all others, include the necessary headers | |
42 | #ifndef WX_PRECOMP | |
43 | #include "wx/wx.h" | |
44 | #endif | |
45 | ||
46 | ||
47 | // many of the tests are specific to the builtin regex lib, so only attempts | |
48 | // to do them when using the builtin regex lib. | |
49 | // | |
50 | #ifdef wxHAS_REGEX_ADVANCED | |
51 | ||
52 | #include "wx/regex.h" | |
53 | #include <string> | |
54 | #include <vector> | |
55 | ||
56 | using CppUnit::Test; | |
57 | using CppUnit::TestCase; | |
58 | using CppUnit::TestSuite; | |
59 | using CppUnit::Exception; | |
60 | ||
61 | using std::string; | |
62 | using std::vector; | |
63 | ||
64 | /////////////////////////////////////////////////////////////////////////////// | |
65 | // The test case - an instance represents a single test | |
66 | ||
67 | class RegExTestCase : public TestCase | |
68 | { | |
69 | public: | |
70 | // constructor - create a single testcase | |
71 | RegExTestCase( | |
72 | const string& name, | |
73 | const char *mode, | |
74 | const char *id, | |
75 | const char *flags, | |
76 | const char *pattern, | |
77 | const char *data, | |
78 | const vector<const char *>& expected); | |
79 | ||
80 | protected: | |
81 | // run this testcase | |
82 | void runTest(); | |
83 | ||
84 | private: | |
85 | // workers | |
86 | wxString Conv(const char *str); | |
87 | void parseFlags(const wxString& flags); | |
88 | void doTest(int flavor); | |
89 | static wxString quote(const wxString& arg); | |
90 | const wxChar *convError() const { return wxT("<cannot convert>"); } | |
91 | ||
92 | // assertions - adds some information about the test that failed | |
93 | void fail(const wxString& msg) const; | |
94 | void failIf(bool condition, const wxString& msg) const | |
95 | { if (condition) fail(msg); } | |
96 | ||
97 | // mode, id, flags, pattern, test data, expected results... | |
98 | int m_mode; | |
99 | wxString m_id; | |
100 | wxString m_flags; | |
101 | wxString m_pattern; | |
102 | wxString m_data; | |
103 | wxArrayString m_expected; | |
104 | ||
105 | // the flag decoded | |
106 | int m_compileFlags; | |
107 | int m_matchFlags; | |
108 | bool m_basic; | |
109 | bool m_extended; | |
110 | bool m_advanced; | |
111 | }; | |
112 | ||
113 | // constructor - throws Exception on failure | |
114 | // | |
115 | RegExTestCase::RegExTestCase( | |
116 | const string& name, | |
117 | const char *mode, | |
118 | const char *id, | |
119 | const char *flags, | |
120 | const char *pattern, | |
121 | const char *data, | |
122 | const vector<const char *>& expected) | |
123 | : | |
124 | TestCase(name), | |
125 | m_mode(mode[0]), | |
126 | m_id(Conv(id)), | |
127 | m_flags(Conv(flags)), | |
128 | m_pattern(Conv(pattern)), | |
129 | m_data(Conv(data)), | |
130 | m_compileFlags(0), | |
131 | m_matchFlags(0), | |
132 | m_basic(false), | |
133 | m_extended(false), | |
134 | m_advanced(false) | |
135 | { | |
136 | bool badconv = m_pattern == convError() || m_data == convError(); | |
137 | //RN: Removing the std:: here will break MSVC6 compilation | |
138 | std::vector<const char *>::const_iterator it; | |
139 | ||
140 | for (it = expected.begin(); it != expected.end(); ++it) { | |
141 | m_expected.push_back(Conv(*it)); | |
142 | badconv = badconv || *m_expected.rbegin() == convError(); | |
143 | } | |
144 | ||
145 | failIf(badconv, wxT("cannot convert to default character encoding")); | |
146 | ||
147 | // the flags need further parsing... | |
148 | parseFlags(m_flags); | |
149 | ||
150 | #ifndef wxHAS_REGEX_ADVANCED | |
151 | failIf(!m_basic && !m_extended, wxT("advanced regexs not available")); | |
152 | #endif | |
153 | } | |
154 | ||
155 | int wxWcscmp(const wchar_t* s1, const wchar_t* s2) | |
156 | { | |
157 | size_t nLen1 = wxWcslen(s1); | |
158 | size_t nLen2 = wxWcslen(s2); | |
159 | ||
160 | if (nLen1 != nLen2) | |
161 | return nLen1 - nLen2; | |
162 | ||
163 | return memcmp(s1, s2, nLen1*sizeof(wchar_t)); | |
164 | } | |
165 | ||
166 | // convert a string from UTF8 to the internal encoding | |
167 | // | |
168 | wxString RegExTestCase::Conv(const char *str) | |
169 | { | |
170 | const wxWCharBuffer wstr = wxConvUTF8.cMB2WC(str); | |
171 | const wxWC2WXbuf buf = wxConvCurrent->cWC2WX(wstr); | |
172 | ||
173 | if (!buf || wxWcscmp(wxConvCurrent->cWX2WC(buf), wstr) != 0) | |
174 | return convError(); | |
175 | ||
176 | return buf; | |
177 | } | |
178 | ||
179 | // Parse flags | |
180 | // | |
181 | void RegExTestCase::parseFlags(const wxString& flags) | |
182 | { | |
183 | for ( wxString::const_iterator p = flags.begin(); p != flags.end(); ++p ) | |
184 | { | |
185 | switch ( (*p).GetValue() ) { | |
186 | // noop | |
187 | case '-': break; | |
188 | ||
189 | // we don't fully support these flags, but they don't stop us | |
190 | // checking for success of failure of the match, so treat as noop | |
191 | case 'A': case 'B': case 'E': case 'H': | |
192 | case 'I': case 'L': case 'M': case 'N': | |
193 | case 'P': case 'Q': case 'R': case 'S': | |
194 | case 'T': case 'U': case '%': | |
195 | break; | |
196 | ||
197 | // match options | |
198 | case '^': m_matchFlags |= wxRE_NOTBOL; break; | |
199 | case '$': m_matchFlags |= wxRE_NOTEOL; break; | |
200 | #if wxUSE_UNICODE | |
201 | case '*': break; | |
202 | #endif | |
203 | // compile options | |
204 | case '&': m_advanced = m_basic = true; break; | |
205 | case 'b': m_basic = true; break; | |
206 | case 'e': m_extended = true; break; | |
207 | case 'i': m_compileFlags |= wxRE_ICASE; break; | |
208 | case 'o': m_compileFlags |= wxRE_NOSUB; break; | |
209 | case 'n': m_compileFlags |= wxRE_NEWLINE; break; | |
210 | case 't': if (strchr("ep", m_mode)) break; // else fall through... | |
211 | ||
212 | // anything else we must skip the test | |
213 | default: | |
214 | fail(wxString::Format( | |
215 | wxT("requires unsupported flag '%c'"), *p)); | |
216 | } | |
217 | } | |
218 | } | |
219 | ||
220 | // Try test for all flavours of expression specified | |
221 | // | |
222 | void RegExTestCase::runTest() | |
223 | { | |
224 | if (m_basic) | |
225 | doTest(wxRE_BASIC); | |
226 | if (m_extended) | |
227 | doTest(wxRE_EXTENDED); | |
228 | #ifdef wxHAS_REGEX_ADVANCED | |
229 | if (m_advanced || (!m_basic && !m_extended)) | |
230 | doTest(wxRE_ADVANCED); | |
231 | #endif | |
232 | } | |
233 | ||
234 | // Try the test for a single flavour of expression | |
235 | // | |
236 | void RegExTestCase::doTest(int flavor) | |
237 | { | |
238 | wxRegEx re(m_pattern, m_compileFlags | flavor); | |
239 | ||
240 | // 'e' - test that the pattern fails to compile | |
241 | if (m_mode == 'e') { | |
242 | failIf(re.IsValid(), wxT("compile succeeded (should fail)")); | |
243 | return; | |
244 | } | |
245 | failIf(!re.IsValid(), wxT("compile failed")); | |
246 | ||
247 | bool matches = re.Matches(m_data, m_matchFlags); | |
248 | ||
249 | // 'f' or 'p' - test that the pattern does not match | |
250 | if (m_mode == 'f' || m_mode == 'p') { | |
251 | failIf(matches, wxT("match succeeded (should fail)")); | |
252 | return; | |
253 | } | |
254 | ||
255 | // otherwise 'm' or 'i' - test the pattern does match | |
256 | failIf(!matches, wxT("match failed")); | |
257 | ||
258 | if (m_compileFlags & wxRE_NOSUB) | |
259 | return; | |
260 | ||
261 | // check wxRegEx has correctly counted the number of subexpressions | |
262 | wxString msg; | |
263 | msg << wxT("GetMatchCount() == ") << re.GetMatchCount() | |
264 | << wxT(", expected ") << m_expected.size(); | |
265 | failIf(m_expected.size() != re.GetMatchCount(), msg); | |
266 | ||
267 | for (size_t i = 0; i < m_expected.size(); i++) { | |
268 | wxString result; | |
269 | size_t start, len; | |
270 | ||
271 | msg.clear(); | |
272 | msg << wxT("wxRegEx::GetMatch failed for match ") << i; | |
273 | failIf(!re.GetMatch(&start, &len, i), msg); | |
274 | ||
275 | // m - check the match returns the strings given | |
276 | if (m_mode == 'm') | |
277 | { | |
278 | if (start < INT_MAX) | |
279 | result = m_data.substr(start, len); | |
280 | else | |
281 | result = wxT(""); | |
282 | } | |
283 | ||
284 | // i - check the match returns the offsets given | |
285 | else if (m_mode == 'i') | |
286 | { | |
287 | if (start > INT_MAX) | |
288 | result = wxT("-1 -1"); | |
289 | else if (start + len > 0) | |
290 | result << start << wxT(" ") << start + len - 1; | |
291 | else | |
292 | result << start << wxT(" -1"); | |
293 | } | |
294 | ||
295 | msg.clear(); | |
296 | msg << wxT("match(") << i << wxT(") == ") << quote(result) | |
297 | << wxT(", expected == ") << quote(m_expected[i]); | |
298 | failIf(result != m_expected[i], msg); | |
299 | } | |
300 | } | |
301 | ||
302 | // assertion - adds some information about the test that failed | |
303 | // | |
304 | void RegExTestCase::fail(const wxString& msg) const | |
305 | { | |
306 | wxString str; | |
307 | wxArrayString::const_iterator it; | |
308 | ||
309 | str << (wxChar)m_mode << wxT(" ") << m_id << wxT(" ") << m_flags << wxT(" ") | |
310 | << quote(m_pattern) << wxT(" ") << quote(m_data); | |
311 | ||
312 | for (it = m_expected.begin(); it != m_expected.end(); ++it) | |
313 | str << wxT(" ") << quote(*it); | |
314 | ||
315 | if (str.length() > 77) | |
316 | str = str.substr(0, 74) + wxT("..."); | |
317 | ||
318 | str << wxT("\n ") << msg; | |
319 | ||
320 | // no lossy convs so using utf8 | |
321 | CPPUNIT_FAIL(string(str.mb_str(wxConvUTF8))); | |
322 | } | |
323 | ||
324 | // quote a string so that it can be displayed (static) | |
325 | // | |
326 | wxString RegExTestCase::quote(const wxString& arg) | |
327 | { | |
328 | const wxChar *needEscape = wxT("\a\b\t\n\v\f\r\"\\"); | |
329 | const wxChar *escapes = wxT("abtnvfr\"\\"); | |
330 | wxString str; | |
331 | ||
332 | for (size_t i = 0; i < arg.length(); i++) { | |
333 | wxChar ch = (wxChar)arg[i]; | |
334 | const wxChar *p = wxStrchr(needEscape, ch); | |
335 | ||
336 | if (p) | |
337 | str += wxString::Format(wxT("\\%c"), escapes[p - needEscape]); | |
338 | else if (wxIscntrl(ch)) | |
339 | str += wxString::Format(wxT("\\%03o"), ch); | |
340 | else | |
341 | str += (wxChar)ch; | |
342 | } | |
343 | ||
344 | return str.length() == arg.length() && str.find(' ') == wxString::npos ? | |
345 | str : wxT("\"") + str + wxT("\""); | |
346 | } | |
347 | ||
348 | ||
349 | /////////////////////////////////////////////////////////////////////////////// | |
350 | // Test suite | |
351 | ||
352 | class RegExTestSuite : public TestSuite | |
353 | { | |
354 | public: | |
355 | RegExTestSuite(string name) : TestSuite(name) { } | |
356 | void add(const char *mode, const char *id, const char *flags, | |
357 | const char *pattern, const char *data, const char *expected, ...); | |
358 | }; | |
359 | ||
360 | // Add a testcase to the suite | |
361 | // | |
362 | void RegExTestSuite::add( | |
363 | const char *mode, | |
364 | const char *id, | |
365 | const char *flags, | |
366 | const char *pattern, | |
367 | const char *data, | |
368 | const char *expected, ...) | |
369 | { | |
370 | string name = getName() + "." + id; | |
371 | ||
372 | vector<const char *> expected_results; | |
373 | va_list ap; | |
374 | ||
375 | for (va_start(ap, expected); expected; expected = va_arg(ap, const char *)) | |
376 | expected_results.push_back(expected); | |
377 | ||
378 | va_end(ap); | |
379 | ||
380 | try { | |
381 | addTest(new RegExTestCase( | |
382 | name, mode, id, flags, pattern, data, expected_results)); | |
383 | } | |
384 | catch (Exception& e) { | |
385 | wxLogInfo(wxString::Format(wxT("skipping: %s\n %s\n"), | |
386 | wxString(name.c_str(), wxConvUTF8).c_str(), | |
387 | wxString(e.what(), wxConvUTF8).c_str())); | |
388 | } | |
389 | } | |
390 | ||
391 | ||
392 | // Include the generated tests | |
393 | // | |
394 | #include "regex.inc" | |
395 | ||
396 | ||
397 | #endif // wxHAS_REGEX_ADVANCED | |
398 | ||
399 | #endif // wxUSE_REGEX |