2 / Copyright (c) 2003 Boost.Test contributors
4 / Distributed under the Boost Software License, Version 1.0. (See accompanying
5 / file LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt)
9 [section:testing_tool_ref Reference API for writing tests]
13 [/ ###############################################################################################]
14 [section:assertion_boost_test_universal_macro `BOOST_TEST`]
18 BOOST_TEST(statement);
19 BOOST_TEST_<level>(statement);
21 // replacement failure message, requires variadic macros
22 BOOST_TEST(statement, "failure message");
24 // Floating point comparison, requires variadic macros, auto and decltype
25 BOOST_TEST(statement, floating_point_comparison_manipulation);
27 // bitwise comparison, requires variadic macros, auto and decltype
28 BOOST_TEST(statement, boost::test_tools::bitwise() );
30 // element-wise comparison, for containers
31 BOOST_TEST(statement, boost::test_tools::per_element() );
33 // lexicographic comparison, for containers
34 BOOST_TEST(statement, boost::test_tools::lexicographic() );
37 The full documentation of this macro is located [link boost_test.testing_tools.boost_test_universal_macro here].
39 The macro is available in three variants, corresponding to different [link boost_test.testing_tools.tools_assertion_severity_level assertion severity levels]:
42 BOOST_TEST // or BOOST_TEST_CHECK
48 * `"failure message"` is a C-string printed in case of failure in place of the default message.
49 See [link boost_test.testing_tools.reports this section] for
51 * `floating_point_comparison_manipulation` is one of the floating point comparison manipulators.
52 See [link boost_test.testing_tools.boost_test_universal_macro this section]
54 * [classref boost::test_tools::bitwise] is a manipulator indicating that the comparison should be performed bitwise. See
55 [link boost_test.testing_tools.extended_comparison.bitwise this section] for more details
56 * [classref boost::test_tools::per_element] is a manipulator indicating that the comparison should be performed on each element, in sequence, rather
57 than on containers. See
58 [link boost_test_coll_perelement this section] for more details
59 * [classref boost::test_tools::lexicographic] is a manipulator indicating that the comparison should be performed with the lexicographic order. See
60 [link boost_test_coll_default_lex this section] for more details
62 [h3 Limitations and workaround]
63 There are some restrictions on the statements that are supported by this tool. Those are explained in details in
64 [link boost_test_statement_limitations this] section.
70 [/ DECORATORS ###############################################################################################]
71 [/-----------------------------------------------------------------]
73 [section:decorator_expected_failures expected_failures (decorator)]
76 expected_failures(counter_t number);
79 Indicates the expected failures for a test unit.
80 See [link boost_test.testing_tools.expected_failures here] for more details.
82 [endsect] [/ section expected_failures]
85 [/-----------------------------------------------------------------]
86 [section:decorator_timeout timeout (decorator)]
89 timeout(unsigned int seconds);
92 Specifies a time-out for a *test-case* or a *test-suite*, in wall-clock time.
94 If a test-case lasts longer than the timeout, the test is flagged as failed. On some systems (see below),
95 the test-case is forced to stop.
97 For test-suites, the mechanism is similar: every test-unit under the test-suite is allocated a maximum
98 duration time that is the remainder of the timeout after the previous tests have been executed. If a timeout occurs
99 during the execution of the suite, the suite is flagged as timed-out and the remaining test-units are skipped.
101 See [link boost_test.testing_tools.timeout here] for more details.
104 `BOOST_SIGACTION_BASED_SIGNAL_HANDLING` is defined
105 if Boost.Test is able to force the test-case to stop.]
107 [note The support for test suites has been added in [link ref_CHANGE_LOG_3_10 Boost 1.70 / __UTF__ v3.10]]
109 [endsect] [/ section timeout]
112 [/-----------------------------------------------------------------]
113 [section:decorator_tolerance tolerance (decorator)]
116 template <typename FPT>
119 template <typename FPT>
120 tolerance(test_tools::fpc::percent_tolerance_t<FPT> eps)
123 Decorator `tolerance` specifies the default comparison tolerance for floating point type `FTP` in the decorated test
124 unit. The default tolerance only applies to a particular type, so it makes sense to provide more than one `tolerance`
125 decorator if we are comparing different floating point types.
126 The variant with `percent_tolerance` uses value ``eps / 100`` as tolerance.
128 [note For more details see the
129 [link boost_test.testing_tools.extended_comparison.floating_point floating points comparison] section.
132 [bt_example decorator_13..decorator tolerance..run-fail]
134 In the above example, in `test1`, checks on `double`s fail because they differ by more what tolerance for `double`s
135 specifies. In `test2` the tolerance for `double`s is greater and therefore the checks succeed. In `test3`, we specify
136 only tolerance for type `float`, and since the checks use type `double` the specified tolerance does not apply. Tolerance
137 in `test4` is equivalent to that in `test1`, therefore its checks also fail. Tolerance in `test5` is equivalent to
138 that in `test2`, therefore its checks also succeed.
139 [endsect] [/ section decorator_tolerance]
146 [/ ###############################################################################################]
147 [#ref_BOOST_level][section:assertion_boost_level `BOOST_<level>`]
151 BOOST_WARN(predicate);
152 BOOST_CHECK(predicate);
153 BOOST_REQUIRE(predicate);
156 These tools are used to validate the predicate value. The only parameter for these tools is a boolean predicate
157 value that gets validated. It could be any expression that could be evaluated and converted to boolean value. The
158 expression gets evaluated only once, so it's safe to pass complex expression for validation.
160 [bt_example example34..BOOST_<level> usage..run-fail]
164 * __BOOST_LEVEL_MESSAGE__
169 [/ ###############################################################################################]
170 [section:assertion_boost_level_bitwise_eq `BOOST_<level>_BITWISE_EQUAL`]
174 BOOST_WARN_BITWISE_EQUAL(left, right);
175 BOOST_CHECK_BITWISE_EQUAL(left, right);
176 BOOST_REQUIRE_BITWISE_EQUAL(left, right);
179 These tools are used to perform bitwise comparison of two values. The check shows all positions where left and
180 right value's bits mismatch.
182 The first parameter is the left compared value. The second parameter is the right compared value. Parameters are
183 not required to be of the same type, but warning is issued if their type's size does not coincide.
185 [bt_example example33..BOOST_<level>_BITWISE_EQUAL usage..run-fail]
189 * __BOOST_LEVEL_EQUAL__
193 [/ ###############################################################################################]
194 [section:assertion_boost_level_eq `BOOST_<level>_EQUAL`]
197 BOOST_WARN_EQUAL(left, right);
198 BOOST_CHECK_EQUAL(left, right);
199 BOOST_REQUIRE_EQUAL(left, right);
202 Check performed by these tools is the same as the one performed by `__BOOST_LEVEL__(left == right)`.
203 The difference is that the mismatched values are reported as well.
205 [note It is bad idea to use these tools to compare floating point values. Use __BOOST_LEVEL_CLOSE__ or
206 __BOOST_LEVEL_CLOSE_FRACTION__ tools instead.
209 [bt_example example35..BOOST_<level>_EQUAL usage..run-fail]
214 * __BOOST_LEVEL_CLOSE__
216 * __BOOST_LEVEL_EQUAL_COLLECTIONS__
220 [/ ###############################################################################################]
221 [section:assertion_boost_level_eq_collections `BOOST_<level>_EQUAL_COLLECTIONS`]
224 BOOST_WARN_EQUAL_COLLECTIONS(left_begin, left_end, right_begin, right_end);
225 BOOST_CHECK_EQUAL_COLLECTIONS(left_begin, left_end, right_begin, right_end);
226 BOOST_REQUIRE_EQUAL_COLLECTIONS(left_begin, left_end, right_begin, right_end);
229 These tools are used to perform an element by element comparison of two collections. They print all mismatched
230 positions, collection elements at these positions and check that the collections have the same size. The first two
231 parameters designate begin and end of the first collection. The two last parameters designate begin and end of the
234 [bt_example example36..BOOST_<level>_EQUAL_COLLECTIONS usage..run-fail]
238 * __BOOST_LEVEL_EQUAL__
242 [/ ###############################################################################################]
243 [section:assertion_boost_level_close `BOOST_<level>_CLOSE`]
246 BOOST_WARN_CLOSE(left, right, tolerance);
247 BOOST_CHECK_CLOSE(left, right, tolerance);
248 BOOST_REQUIRE_CLOSE(left, right, tolerance);
251 These tools are used to check on closeness using strong relationship defined by the predicate
252 ``check_is_close( left, right, tolerance )``
254 To check for the weak relationship use
255 __BOOST_LEVEL_PREDICATE__ family of tools with explicit `check_is_close` invocation.
258 The first parameter is the ['left] compared value. The second parameter is the
259 ['right] compared value. Last third parameter defines the tolerance for the comparison in
260 [link boost_test.testing_tools.extended_comparison.floating_point [*percentage units]].
262 [note It is required for left and right parameters to be of the same floating point type. You will need to explicitly
263 resolve any type mismatch to select which type to use for comparison.
266 [note The floating point comparison tools are automatically added if the __UTF__
267 is included as indicated in the previous sections. The tools are implemented is in the header
268 [headerref boost/test/tools/floating_point_comparison.hpp `boost/test/tools/floating_point_comparison.hpp`].]
270 [bt_example example42..BOOST_<level>_CLOSE usage with small values..run-fail]
271 [bt_example example43..BOOST_<level>_CLOSE usage with big values..run]
275 * __BOOST_LEVEL_CLOSE_FRACTION__
276 * __BOOST_LEVEL_SMALL__
277 * __BOOST_LEVEL_EQUAL__
278 * __floating_points_testing_tools__
282 [/ ###############################################################################################]
283 [section:assertion_boost_level_close_fraction `BOOST_<level>_CLOSE_FRACTION`]
286 BOOST_WARN_CLOSE_FRACTION(left, right, tolerance);
287 BOOST_CHECK_CLOSE_FRACTION(left, right, tolerance);
288 BOOST_REQUIRE_CLOSE_FRACTION(left, right, tolerance);
291 These tools are used to check on closeness using strong relationship defined by the predicate
292 ``check_is_close(left, right, tolerance)``
294 To check for the weak relationship use __BOOST_LEVEL_PREDICATE__ family of tools with explicit `check_is_close` invocation.
296 The first parameter is the ['left] compared value. The second parameter is the
297 ['right] compared value. Last third parameter defines the tolerance for the comparison as
298 [link boost_test.testing_tools.extended_comparison.floating_point [*fraction of absolute values being compared]].
300 [note It is required for left and right parameters to be of the same floating point type. You will need to explicitly
301 resolve any type mismatch to select which type to use for comparison.]
303 [note The floating point comparison tools are automatically added if the __UTF__
304 is included as indicated in the previous sections. The tools are implemented is in the header
305 [headerref boost/test/tools/floating_point_comparison.hpp `boost/test/tools/floating_point_comparison.hpp`].]
308 [bt_example example44..BOOST_<level>_CLOSE_FRACTION usage..run-fail]
312 * __BOOST_LEVEL_CLOSE__
313 * __BOOST_LEVEL_SMALL__
314 * __BOOST_LEVEL_EQUAL__
315 * __floating_points_testing_tools__
319 [/ ###############################################################################################]
320 [section:assertion_boost_level_ge `BOOST_<level>_GE`]
323 BOOST_WARN_GE(left, right);
324 BOOST_CHECK_GE(left, right);
325 BOOST_REQUIRE_GE(left, right);
328 Check performed by these tools is the same as the one performed by `__BOOST_LEVEL__( left >= right )`.
329 The difference is that the argument values are reported as well.
331 [bt_example example57..BOOST_<level>_GE usage..run-fail]
342 [/ ###############################################################################################]
343 [section:assertion_boost_level_gt `BOOST_<level>_GT`]
347 BOOST_WARN_GT(left, right);
348 BOOST_CHECK_GT(left, right);
349 BOOST_REQUIRE_GT(left, right);
352 Check performed by these tools is the same as the one performed by __BOOST_LEVEL__`( left > right )`.
353 The difference is that the argument values are reported as well.
355 [bt_example example58..BOOST_<level>_GT usage..run-fail]
365 [/ ###############################################################################################]
366 [section:assertion_boost_level_le `BOOST_<level>_LE`]
369 BOOST_WARN_LE(left, right);
370 BOOST_CHECK_LE(left, right);
371 BOOST_REQUIRE_LE(left, right);
374 Check performed by these tools is the same as the one performed by `__BOOST_LEVEL__( left <= right )`.
375 The difference is that the argument values are reported as well.
377 [bt_example example55..BOOST_<level>_LE usage..run-fail]
387 [/ ###############################################################################################]
388 [section:assertion_boost_level_lt `BOOST_<level>_LT`]
391 BOOST_WARN_LT(left, right);
392 BOOST_CHECK_LT(left, right);
393 BOOST_REQUIRE_LT(left, right);
396 Check performed by these tools is the same as the one performed by `__BOOST_LEVEL__( left < right )`.
397 The difference is that the argument values are reported as well.
399 [bt_example example56..BOOST_<level>_LT usage..run-fail]
409 [/ ###############################################################################################]
410 [section:assertion_boost_level_message `BOOST_<level>_MESSAGE`]
413 BOOST_WARN_MESSAGE(predicate, message);
414 BOOST_CHECK_MESSAGE(predicate, message);
415 BOOST_REQUIRE_MESSAGE(predicate, message);
418 These tools perform exactly the same check as __BOOST_LEVEL__ tools. The only difference is that
419 instead of generating an error/confirm message these use the supplied one.
421 The first parameter is the boolean expression. The second parameter is the message reported in case of check
422 failure. The message argument can be constructed of components of any type supporting the
423 `std::ostream& operator<<(std::ostream&)`.
425 [bt_example example38..BOOST_<level>_MESSAGE usage..run]
434 [/ ###############################################################################################]
435 [section:assertion_boost_level_ne `BOOST_<level>_NE`]
439 BOOST_WARN_NE(left, right);
440 BOOST_CHECK_NE(left, right);
441 BOOST_REQUIRE_NE(left, right);
444 Check performed by these tools is the same as the one performed by `__BOOST_LEVEL__( left != right )`.
445 The difference is that the matched values are reported as well.
447 [bt_example example54..BOOST_<level>_NE usage..run-fail]
451 * __BOOST_LEVEL_EQUAL__
455 [/ ###############################################################################################]
456 [section:assertion_boost_level_predicate `BOOST_<level>_PREDICATE`]
460 BOOST_WARN_PREDICATE(predicate, arguments_list);
461 BOOST_CHECK_PREDICATE(predicate, arguments_list);
462 BOOST_REQUIRE_PREDICATE(predicate, arguments_list);
465 These are generic tools used to validate an arbitrary supplied predicate functor (there is a compile time limit on
466 predicate arity defined by the configurable macro `BOOST_TEST_MAX_PREDICATE_ARITY`). To
467 validate zero arity predicate use __BOOST_LEVEL__ tools. In other cases prefer theses tools. The
468 advantage of these tools is that they show arguments values in case of predicate failure.
470 The first parameter is the predicate itself. The second parameter is the list of predicate arguments each wrapped
471 in round brackets (`BOOST_PP` sequence format).
473 [bt_example example40..BOOST_<level>_PREDICATE usage..run]
475 [note Note difference in error log from __BOOST_LEVEL__]
483 [/ ###############################################################################################]
484 [section:assertion_boost_level_no_throw `BOOST_<level>_NO_THROW`]
487 BOOST_WARN_NO_THROW(expression);
488 BOOST_CHECK_NO_THROW(expression);
489 BOOST_REQUIRE_NO_THROW(expression);
492 These assertions validate that the execution of `expression` does not throw any exception.
493 To that extent, all possible exception are caught by assertion itself and no exception is propagated to
497 It is possible to test for complex expressions with the use of constructs such as `do { /* ... */} while(0)` block.
500 [bt_example exception_nothrow..BOOST_<level>_NO_THROW usage..run-fail]
504 * __BOOST_LEVEL_THROW__
505 * [link boost_test.testing_tools.exception_correctness Exception correctness] section
509 [/ ###############################################################################################]
510 [section:assertion_boost_level_throw `BOOST_<level>_THROW`]
513 BOOST_WARN_THROW(expression, exception_type);
514 BOOST_CHECK_THROW(expression, exception_type);
515 BOOST_REQUIRE_THROW(expression, exception_type);
518 These assertions validate that the execution of `expression` raises an /expected/ exception, which means an exception of
519 the supplied `exception_type` type or of any child type.
521 * If `expression` raises an unexpected exception, this exception is not caught by `BOOST_<level>_THROW` assertion and
522 might propagate to the test body. If not caught at all, the framework will catch it and terminate the test case
523 with the status /failed/.
524 * If `expression` does not raise any exception, the the assertion fails.
526 [warning the assertion catches only the expected exceptions.]
529 It is possible to test for complex expressions with the use of constructs such as `do { /* ... */} while(0)` block.
532 [bt_example exception_check..BOOST_<level>_THROW usage..run-fail]
536 * __BOOST_LEVEL_NO_THROW__
537 * [link boost_test.testing_tools.exception_correctness Exception correctness] section
542 [/ ###############################################################################################]
543 [section:assertion_boost_level_exception `BOOST_<level>_EXCEPTION`]
546 BOOST_WARN_EXCEPTION(expression, exception_type, predicate);
547 BOOST_CHECK_EXCEPTION(expression, exception_type, predicate);
548 BOOST_REQUIRE_EXCEPTION(expression, exception_type, predicate);
551 As for __BOOST_LEVEL_THROW__, these assertions validate that `expression` raises an exception of the
552 type specified by `exception_type` or any of its child type, with additional checks on the exception instance.
554 * If an expected exception is raised by `expression`, the instance of the exception is passed to `predicate`
555 for further validation.
556 * It behaves like __BOOST_LEVEL_THROW__ if `expression` does not raise any exception, or an unrelated exception is raised.
558 `predicate` should be a unary function accepting an instance of `exception_type` or any of its child, and that should return
559 a boolean indicating a success (`true`) or a failure (`false`).
561 [warning the assertion catches only the expected exceptions.]
564 It is possible to test for complex expressions with the use of constructs such as `do { /* ... */} while(0)` block.
568 The example below checks that the exception carries the proper error code.
570 [bt_example exception_check_predicate..BOOST_<level>_EXCEPTION usage..run-fail]
574 * __BOOST_LEVEL_THROW__
575 * [link boost_test.testing_tools.exception_correctness Exception correctness] section
579 [/ ###############################################################################################]
580 [section:assertion_boost_level_small `BOOST_<level>_SMALL`]
583 BOOST_WARN_SMALL(value, tolerance);
584 BOOST_CHECK_SMALL(value, tolerance);
585 BOOST_REQUIRE_SMALL(value, tolerance);
588 These tools are used to check that supplied value is small enough. The "smallness" is defined by absolute value
589 of the tolerance supplied as a second argument. Use these tools with caution. To compare to values on closeness
590 it's preferable to use __BOOST_LEVEL_CLOSE__ tools instead.
592 The first parameter is the value to check. The second parameter is the tolerance.
594 [note The floating point comparison tools are automatically added if the __UTF__
595 is included as indicated in the previous sections. The tools are implemented is in the header
596 [headerref boost/test/tools/floating_point_comparison.hpp `boost/test/tools/floating_point_comparison.hpp`].]
598 [bt_example example41..BOOST_<level>_SMALL usage..run-fail]
602 * __BOOST_LEVEL_CLOSE__
603 * __BOOST_LEVEL_CLOSE_FRACTION__
604 * __floating_points_testing_tools__
610 [/ ###############################################################################################]
611 [section:test_org_boost_test_case_expected_failure `BOOST_AUTO_TEST_CASE_EXPECTED_FAILURES`]
612 Indicates the number of failures for a test case.
614 See [link boost_test.testing_tools.expected_failures here] for more details.
615 [endsect] [/ expected failures]
617 [/ ###############################################################################################]
618 [section:assertion_boost_error `BOOST_ERROR`]
621 BOOST_ERROR(message);
624 __BOOST_ERROR__ tool behave the same way as `__BOOST_TEST__(false, message)`. This tool is used for
625 an unconditional error counter increasing and message logging.
627 The tool's only parameter is an error message to log.
629 [bt_example example46..BOOST_ERROR usage..run-fail]
638 [/ ###############################################################################################]
639 [section:assertion_boost_fail `BOOST_FAIL`]
645 `__BOOST_FAIL__(message)` behave the same way as `__BOOST_TEST_REQUIRE__(false, message)`. This tool is used for an
646 unconditional error counter increasing, message logging and the current test case aborting.
648 The tool's only parameter is an error message to log.
650 [bt_example example47..BOOST_FAIL usage..run-fail]
660 [/ ###############################################################################################]
661 [section:assertion_boost_is_defined `BOOST_IS_DEFINED`]
664 BOOST_IS_DEFINED(symbol);
667 Unlike the rest of the tools in the toolbox this tool does not perform the logging itself. Its only purpose
668 is to check at runtime whether or not the supplied preprocessor symbol is defined. Use it in combination with
669 __BOOST_LEVEL__ to perform and log validation. Macros of any arity could be checked. To check the
670 macro definition with non-zero arity specify dummy arguments for it. See below for example.
672 The only tool's parameter is a preprocessor symbol that gets validated.
674 [bt_example example48..BOOST_IS_DEFINED usage..run-fail]
682 [/ ###############################################################################################]
683 [section:assertion_control_under_debugger `BOOST_TEST_TOOLS_UNDER_DEBUGGER`]
684 When defined, assertions evaluate their expression eagerly, as described [link boost_test.testing_tools.debugging here].
685 [endsect] [/ assertion_control_under_debugger]
687 [/ ###############################################################################################]
688 [section:assertion_control_under_debuggable `BOOST_TEST_TOOLS_DEBUGGABLE`]
689 When defined, test assertions are compiled in two modes (debugger-friendly and full-featured) and the version is selected at run-time, as described [link boost_test.testing_tools.debugging here].
690 [endsect] [/ assertion_control_under_debuggable]
692 [endsect] [/ testing_tool_ref]