Contents
Natural Expressions
Exceptions
Matcher expressions
Thread Safety
Expressions with commas
Most test frameworks have a large collection of assertion macros to capture all possible conditional forms (_EQUALS
, _NOTEQUALS
, _GREATER_THAN
etc).
Catch is different. Because it decomposes natural C-style conditional expressions most of these forms are reduced to one or two that you will use all the time. That said there is a rich set of auxiliary macros as well. We'll describe all of these here.
Most of these macros come in two forms:
The REQUIRE
family of macros tests an expression and aborts the test case if it fails.
The CHECK
family are equivalent but execution continues in the same test case even if the assertion fails. This is useful if you have a series of essentially orthogonal assertions and it is useful to see all the results rather than stopping at the first failure.
Evaluates the expression and records the result. If an exception is thrown, it is caught, reported, and counted as a failure. These are the macros you will use most of the time.
Examples:
CHECK( str == "string value" );
CHECK( thisReturnsTrue() );
REQUIRE( i == 42 );
Evaluates the expression and records the logical NOT of the result. If an exception is thrown it is caught, reported, and counted as a failure. (these forms exist as a workaround for the fact that ! prefixed expressions cannot be decomposed).
Example:
REQUIRE_FALSE( thisReturnsFalse() );
Do note that "overly complex" expressions cannot be decomposed and thus will not compile. This is done partly for practical reasons (to keep the underlying expression template machinery to minimum) and partly for philosophical reasons (assertions should be simple and deterministic).
Examples:
CHECK(a == 1 && b == 2);
This expression is too complex because of the &&
operator. If you want to check that 2 or more properties hold, you can either put the expression into parenthesis, which stops decomposition from working, or you need to decompose the expression into two assertions: CHECK( a == 1 ); CHECK( b == 2);
CHECK( a == 2 || b == 1 );
This expression is too complex because of the ||
operator. If you want to check that one of several properties hold, you can put the expression into parenthesis (unlike with &&
, expression decomposition into several CHECK
s is not possible).When comparing floating point numbers - especially if at least one of them has been computed - great care must be taken to allow for rounding errors and inexact representations.
Catch provides a way to perform tolerant comparisons of floating point values through use of a wrapper class called Approx
. Approx
can be used on either side of a comparison expression. It overloads the comparisons operators to take a tolerance into account. Here's a simple example:
REQUIRE( performComputation() == Approx( 2.1 ) );
Catch also provides a user-defined literal for Approx
; _a
. It resides in
the Catch::literals
namespace and can be used like so:
using namespace Catch::literals;
REQUIRE( performComputation() == 2.1_a );
Approx
is constructed with defaults that should cover most simple cases.
For the more complex cases, Approx
provides 3 customization points:
Approx
's value before it is rejected.
By default set to std::numeric_limits<float>::epsilon()*100
.Approx
's value before it is rejected.
By default set to 0.0
.Approx
for relative check.
By default set to 0.0
.Approx target = Approx(100).epsilon(0.01);
100.0 == target; // Obviously true
200.0 == target; // Obviously still false
100.5 == target; // True, because we set target to allow up to 1% difference
Approx target = Approx(100).margin(5);
100.0 == target; // Obviously true
200.0 == target; // Obviously still false
104.0 == target; // True, because we set target to allow absolute difference of at most 5
Scale can be useful if the computation leading to the result worked
on different scale than is used by the results. Since allowed difference
between Approx's value and compared value is based primarily on Approx's value
(the allowed difference is computed as
(Approx::scale + Approx::value) * epsilon
), the resulting comparison could
need rescaling to be correct.
Expects that no exception is thrown during evaluation of the expression.
Expects that an exception (of any type) is be thrown during evaluation of the expression.
Expects that an exception of the specified type is thrown during evaluation of the expression. Note that the exception type is extended with const&
and you should not include it yourself.
Expects that an exception is thrown that, when converted to a string, matches the string or string matcher provided (see next section for Matchers).
e.g.
REQUIRE_THROWS_WITH( openThePodBayDoors(), Contains( "afraid" ) && Contains( "can't do that" ) );
REQUIRE_THROWS_WITH( dismantleHal(), "My mind is going" );
Expects that exception of exception type is thrown and it matches provided matcher (see the documentation for Matchers).
Please note that the THROW
family of assertions expects to be passed a single expression, not a statement or series of statements. If you want to check a more complicated sequence of operations, you can use a C++11 lambda function.
REQUIRE_NOTHROW([&](){
int i = 1;
int j = 2;
auto k = i + j;
if (k == 3) {
throw 1;
}
}());
To support Matchers a slightly different form is used. Matchers have their own documentation.
Matchers can be composed using &&
, ||
and !
operators.
Currently assertions in Catch are not thread safe. For more details, along with workarounds, see the section on the limitations page.
Because the preprocessor parses code using different rules than the
compiler, multiple-argument assertions (e.g. REQUIRE_THROWS_AS
) have
problems with commas inside the provided expressions. As an example
REQUIRE_THROWS_AS(std::pair<int, int>(1, 2), std::invalid_argument);
will fail to compile, because the preprocessor sees 3 arguments provided,
but the macro accepts only 2. There are two possible workarounds.
1) Use typedef:
using int_pair = std::pair<int, int>;
REQUIRE_THROWS_AS(int_pair(1, 2), std::invalid_argument);
This solution is always applicable, but makes the meaning of the code less clear.
2) Parenthesize the expression:
TEST_CASE_METHOD((Fixture<int, int>), "foo", "[bar]") {
SUCCEED();
}
This solution is not always applicable, because it might require extra changes on the Catch's side to work.