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