Unit Testing

Base.runtests(tests=["all"]; ncores=ceil(Int, Sys.CPU_THREADS / 2),
              exit_on_error=false, revise=false, propagate_project=true, [seed])

Run the Julia unit tests listed in tests, which can be either a string or an array of strings, using ncores processors. If exit_on_error is false, when one test fails, all remaining tests in other files will still be run; they are otherwise discarded, when exit_on_error == true. If revise is true, the Revise package is used to load any modifications to Base or to the standard libraries before running the tests. If propagate_project is true the current project is propagated to the test environment. If a seed is provided via the keyword argument, it is used to seed the global RNG in the context where the tests are run; otherwise the seed is chosen randomly.

@test ex
@test f(args...) key=val ...
@test ex broken=true
@test ex skip=true

Test that the expression ex evaluates to true. If executed inside a @testset, return a Pass Result if it does, a Fail Result if it is false, and an Error Result if it could not be evaluated. If executed outside a @testset, throw an exception instead of returning Fail or Error.

Examples

julia> @test true
Test Passed

julia> @test [1, 2] + [2, 1] == [3, 3]
Test Passed

The @test f(args...) key=val... form is equivalent to writing @test f(args..., key=val...) which can be useful when the expression is a call using infix syntax such as approximate comparisons:

julia> @test π ≈ 3.14 atol=0.01
Test Passed

This is equivalent to the uglier test @test ≈(π, 3.14, atol=0.01). It is an error to supply more than one expression unless the first is a call expression and the rest are assignments (k=v).

You can use any key for the key=val arguments, except for broken and skip, which have special meanings in the context of @test:

• broken=cond indicates a test that should pass but currently consistently fails when cond==true. Tests that the expression ex evaluates to false or causes an exception. Returns a Broken Result if it does, or an Error Result if the expression evaluates to true. Regular @test ex is evaluated when cond==false.
• skip=cond marks a test that should not be executed but should be included in test summary reporting as Broken, when cond==true. This can be useful for tests that intermittently fail, or tests of not-yet-implemented functionality. Regular @test ex is evaluated when cond==false.

Examples

julia> @test 2 + 2 ≈ 6 atol=1 broken=true
Test Broken
  Expression: ≈(2 + 2, 6, atol = 1)

julia> @test 2 + 2 ≈ 5 atol=1 broken=false
Test Passed

julia> @test 2 + 2 == 5 skip=true
Test Broken
  Skipped: 2 + 2 == 5

julia> @test 2 + 2 == 4 skip=false
Test Passed

@test_throws exception expr

Tests that the expression expr throws exception. The exception may specify either a type, a string, regular expression, or list of strings occurring in the displayed error message, a matching function, or a value (which will be tested for equality by comparing fields). Note that @test_throws does not support a trailing keyword form.

Examples

julia> @test_throws BoundsError [1, 2, 3][4]
Test Passed
      Thrown: BoundsError

julia> @test_throws DimensionMismatch [1, 2, 3] + [1, 2]
Test Passed
      Thrown: DimensionMismatch

julia> @test_throws "Try sqrt(Complex" sqrt(-1)
Test Passed
     Message: "DomainError with -1.0:\nsqrt was called with a negative real argument but will only return a complex result if called with a complex argument. Try sqrt(Complex(x))."

In the final example, instead of matching a single string it could alternatively have been performed with:

• ["Try", "Complex"] (a list of strings)
• r"Try sqrt\([Cc]omplex" (a regular expression)
• str -> occursin("complex", str) (a matching function)

@testset [CustomTestSet] [options...] ["description"] begin test_ex end
@testset [CustomTestSet] [options...] ["description $v"] for v in itr test_ex end
@testset [CustomTestSet] [options...] ["description $v, $w"] for v in itrv, w in itrw test_ex end
@testset [CustomTestSet] [options...] ["description"] test_func()
@testset let v = v, w = w; test_ex; end

With begin/end or function call

When @testset is used, with begin/end or a single function call, the macro starts a new test set in which to evaluate the given expression.

If no custom testset type is given it defaults to creating a DefaultTestSet. DefaultTestSet records all the results and, if there are any Fails or Errors, throws an exception at the end of the top-level (non-nested) test set, along with a summary of the test results.

Any custom testset type (subtype of AbstractTestSet) can be given and it will also be used for any nested @testset invocations. The given options are only applied to the test set where they are given. The default test set type accepts three boolean options:

• verbose: if true, the result summary of the nested testsets is shown even when they all pass (the default is false).
• showtiming: if true, the duration of each displayed testset is shown (the default is true).
• failfast: if true, any test failure or error will cause the testset and any child testsets to return immediately (the default is false). This can also be set globally via the env var JULIA_TEST_FAILFAST.

The description string accepts interpolation from the loop indices. If no description is provided, one is constructed based on the variables. If a function call is provided, its name will be used. Explicit description strings override this behavior.

By default the @testset macro will return the testset object itself, though this behavior can be customized in other testset types. If a for loop is used then the macro collects and returns a list of the return values of the finish method, which by default will return a list of the testset objects used in each iteration.

Before the execution of the body of a @testset, there is an implicit call to Random.seed!(seed) where seed is the current seed of the global RNG. Moreover, after the execution of the body, the state of the global RNG is restored to what it was before the @testset. This is meant to ease reproducibility in case of failure, and to allow seamless re-arrangements of @testsets regardless of their side-effect on the global RNG state.

Examples

julia> @testset "trigonometric identities" begin
           θ = 2/3*π
           @test sin(-θ) ≈ -sin(θ)
           @test cos(-θ) ≈ cos(θ)
           @test sin(2θ) ≈ 2*sin(θ)*cos(θ)
           @test cos(2θ) ≈ cos(θ)^2 - sin(θ)^2
       end;
Test Summary:            | Pass  Total  Time
trigonometric identities |    4      4  0.2s

@testset for

When @testset for is used, the macro starts a new test for each iteration of the provided loop. The semantics of each test set are otherwise identical to that of that begin/end case (as if used for each loop iteration).

@testset let

When @testset let is used, the macro starts a transparent test set with the given object added as a context object to any failing test contained therein. This is useful when performing a set of related tests on one larger object and it is desirable to print this larger object when any of the individual tests fail. Transparent test sets do not introduce additional levels of nesting in the test set hierarchy and are passed through directly to the parent test set (with the context object appended to any failing tests.)

Special implicit world age increment for @testset begin

World age inside @testset begin increments implicitly after every statement. This matches the behavior of ordinary toplevel code, but not that of ordinary begin/end blocks, i.e. with respect to world age, @testset begin behaves as if the body of the begin/end block was written at toplevel.

Examples

julia> @testset let logi = log(im)
           @test imag(logi) == π/2
           @test !iszero(real(logi))
       end
Test Failed at none:3
  Expression: !(iszero(real(logi)))
     Context: logi = 0.0 + 1.5707963267948966im

ERROR: There was an error during testing

julia> @testset let logi = log(im), op = !iszero
           @test imag(logi) == π/2
           @test op(real(logi))
       end
Test Failed at none:3
  Expression: op(real(logi))
     Context: logi = 0.0 + 1.5707963267948966im
              op = !iszero

ERROR: There was an error during testing

TestSetException

Thrown when a test set finishes and not all tests passed.

@test_logs [log_patterns...] [keywords] expression

Collect a list of log records generated by expression using collect_test_logs, check that they match the sequence log_patterns, and return the value of expression. The keywords provide some simple filtering of log records: the min_level keyword controls the minimum log level which will be collected for the test, the match_mode keyword defines how matching will be performed (the default :all checks that all logs and patterns match pairwise; use :any to check that the pattern matches at least once somewhere in the sequence.)

The most useful log pattern is a simple tuple of the form (level,message). A different number of tuple elements may be used to match other log metadata, corresponding to the arguments to passed to AbstractLogger via the handle_message function: (level,message,module,group,id,file,line). Elements which are present will be matched pairwise with the log record fields using == by default, with the special cases that Symbols may be used for the standard log levels, and Regexs in the pattern will match string or Symbol fields using occursin.

Examples

Consider a function which logs a warning, and several debug messages:

function foo(n)
    @info "Doing foo with n=$n"
    for i=1:n
        @debug "Iteration $i"
    end
    42
end

We can test the info message using

@test_logs (:info,"Doing foo with n=2") foo(2)

If we also wanted to test the debug messages, these need to be enabled with the min_level keyword:

using Logging
@test_logs (:info,"Doing foo with n=2") (:debug,"Iteration 1") (:debug,"Iteration 2") min_level=Logging.Debug foo(2)

If you want to test that some particular messages are generated while ignoring the rest, you can set the keyword match_mode=:any:

using Logging
@test_logs (:info,) (:debug,"Iteration 42") min_level=Logging.Debug match_mode=:any foo(100)

The macro may be chained with @test to also test the returned value:

@test (@test_logs (:info,"Doing foo with n=2") foo(2)) == 42

If you want to test for the absence of warnings, you can omit specifying log patterns and set the min_level accordingly:

# test that the expression logs no messages when the logger level is warn:
@test_logs min_level=Logging.Warn @info("Some information") # passes
@test_logs min_level=Logging.Warn @warn("Some information") # fails

If you want to test the absence of warnings (or error messages) in stderr which are not generated by @warn, see @test_nowarn.

TestLogger(; min_level=Info, catch_exceptions=false)

Create a TestLogger which captures logged messages in its logs::Vector{LogRecord} field.

Set min_level to control the LogLevel, catch_exceptions for whether or not exceptions thrown as part of log event generation should be caught, and respect_maxlog for whether or not to follow the convention of logging messages with maxlog=n for some integer n at most n times.

See also: LogRecord.

Examples

julia> using Test, Logging

julia> f() = @info "Hi" number=5;

julia> test_logger = TestLogger();

julia> with_logger(test_logger) do
           f()
           @info "Bye!"
       end

julia> @test test_logger.logs[1].message == "Hi"
Test Passed

julia> @test test_logger.logs[1].kwargs[:number] == 5
Test Passed

julia> @test test_logger.logs[2].message == "Bye!"
Test Passed

LogRecord

Stores the results of a single log event. Fields:

• level: the LogLevel of the log message
• message: the textual content of the log message
• _module: the module of the log event
• group: the logging group (by default, the name of the file containing the log event)
• id: the ID of the log event
• file: the file containing the log event
• line: the line within the file of the log event
• kwargs: any keyword arguments passed to the log event

@inferred [AllowedType] f(x)

Tests that the call expression f(x) returns a value of the same type inferred by the compiler. It is useful to check for type stability.

f(x) can be any call expression. Returns the result of f(x) if the types match, and an Error Result if it finds different types.

Optionally, AllowedType relaxes the test, by making it pass when either the type of f(x) matches the inferred type modulo AllowedType, or when the return type is a subtype of AllowedType. This is useful when testing type stability of functions returning a small union such as Union{Nothing, T} or Union{Missing, T}.

julia> f(a) = a > 1 ? 1 : 1.0
f (generic function with 1 method)

julia> typeof(f(2))
Int64

julia> @code_warntype f(2)
MethodInstance for f(::Int64)
  from f(a) @ Main none:1
Arguments
  #self#::Core.Const(f)
  a::Int64
Body::UNION{FLOAT64, INT64}
1 ─ %1 = :>::Core.Const(>)
│   %2 = (%1)(a, 1)::Bool
└──      goto #3 if not %2
2 ─      return 1
3 ─      return 1.0

julia> @inferred f(2)
ERROR: return type Int64 does not match inferred return type Union{Float64, Int64}
[...]

julia> @inferred max(1, 2)
2

julia> g(a) = a < 10 ? missing : 1.0
g (generic function with 1 method)

julia> @inferred g(20)
ERROR: return type Float64 does not match inferred return type Union{Missing, Float64}
[...]

julia> @inferred Missing g(20)
1.0

julia> h(a) = a < 10 ? missing : f(a)
h (generic function with 1 method)

julia> @inferred Missing h(20)
ERROR: return type Int64 does not match inferred return type Union{Missing, Float64, Int64}
[...]

@test_deprecated [pattern] expression

When --depwarn=yes, test that expression emits a deprecation warning and return the value of expression. The log message string will be matched against pattern which defaults to r"deprecated"i.

When --depwarn=no, simply return the result of executing expression. When --depwarn=error, check that an ErrorException is thrown.

Examples

# Deprecated in julia 0.7
@test_deprecated num2hex(1)

# The returned value can be tested by chaining with @test:
@test (@test_deprecated num2hex(1)) == "0000000000000001"

@test_warn msg expr

Test whether evaluating expr results in stderr output that contains the msg string or matches the msg regular expression. If msg is a boolean function, tests whether msg(output) returns true. If msg is a tuple or array, checks that the error output contains/matches each item in msg. Returns the result of evaluating expr.

See also @test_nowarn to check for the absence of error output.

Note: Warnings generated by @warn cannot be tested with this macro. Use @test_logs instead.

@test_nowarn expr

Test whether evaluating expr results in empty stderr output (no warnings or other messages). Returns the result of evaluating expr.

Note: The absence of warnings generated by @warn cannot be tested with this macro. Use @test_logs instead.

@test_broken ex
@test_broken f(args...) key=val ...

Indicates a test that should pass but currently consistently fails. Tests that the expression ex evaluates to false or causes an exception. Returns a Broken Result if it does, or an Error Result if the expression evaluates to true. This is equivalent to @test ex broken=true.

The @test_broken f(args...) key=val... form works as for the @test macro.

Examples

julia> @test_broken 1 == 2
Test Broken
  Expression: 1 == 2

julia> @test_broken 1 == 2 atol=0.1
Test Broken
  Expression: ==(1, 2, atol = 0.1)

@test_skip ex
@test_skip f(args...) key=val ...

Marks a test that should not be executed but should be included in test summary reporting as Broken. This can be useful for tests that intermittently fail, or tests of not-yet-implemented functionality. This is equivalent to @test ex skip=true.

The @test_skip f(args...) key=val... form works as for the @test macro.

Examples

julia> @test_skip 1 == 2
Test Broken
  Skipped: 1 == 2

julia> @test_skip 1 == 2 atol=0.1
Test Broken
  Skipped: ==(1, 2, atol = 0.1)

Test.Result

All tests produce a result object. This object may or may not be stored, depending on whether the test is part of a test set.

Test.Pass <: Test.Result

The test condition was true, i.e. the expression evaluated to true or the correct exception was thrown.

Test.Fail <: Test.Result

The test condition was false, i.e. the expression evaluated to false or the correct exception was not thrown.

Test.Error <: Test.Result

The test condition couldn't be evaluated due to an exception, or it evaluated to something other than a Bool. In the case of @test_broken it is used to indicate that an unexpected Pass Result occurred.

Test.Broken <: Test.Result

The test condition is the expected (failed) result of a broken test, or was explicitly skipped with @test_skip.

record(ts::AbstractTestSet, res::Result)

Record a result to a testset. This function is called by the @testset infrastructure each time a contained @test macro completes, and is given the test result (which could be an Error). This will also be called with an Error if an exception is thrown inside the test block but outside of a @test context.

finish(ts::AbstractTestSet)

Do any final processing necessary for the given testset. This is called by the @testset infrastructure after a test block executes.

Custom AbstractTestSet subtypes should call record on their parent (if there is one) to add themselves to the tree of test results. This might be implemented as:

if get_testset_depth() != 0
    # Attach this test set to the parent test set
    parent_ts = get_testset()
    record(parent_ts, self)
    return self
end

get_testset()

Retrieve the active test set from the task's local storage. If no test set is active, use the fallback default test set.

get_testset_depth()

Return the number of active test sets, not including the default test set

TestCounts

Holds the state for recursively gathering the results of a test set for display purposes.

Fields:

• customized: Whether the function get_test_counts was customized for the AbstractTestSet this counts object is for. If a custom method was defined, always pass true to the constructor.
• passes: The number of passing @test invocations.
• fails: The number of failing @test invocations.
• errors: The number of erroring @test invocations.
• broken: The number of broken @test invocations.
• passes: The cumulative number of passing @test invocations.
• fails: The cumulative number of failing @test invocations.
• errors: The cumulative number of erroring @test invocations.
• broken: The cumulative number of broken @test invocations.
• duration: The total duration the AbstractTestSet in question ran for, as a formatted String.

" gettestcounts(::AbstractTestSet) -> TestCounts

Recursive function that counts the number of test results of each type directly in the testset, and totals across the child testsets.

Custom AbstractTestSet should implement this function to get their totals counted & displayed with DefaultTestSet as well.

If this is not implemented for a custom TestSet, the printing falls back to reporting x for failures and ?s for the duration.

format_duration(::AbstractTestSet)

Return a formatted string for printing the duration the testset ran for.

If not defined, falls back to "?s".

print_test_results(ts::AbstractTestSet, depth_pad=0)

Print the results of an AbstractTestSet as a formatted table.

depth_pad refers to how much padding should be added in front of all output.

Called inside of Test.finish, if the finished testset is the topmost testset.

The GenericArray can be used to test generic array APIs that program to the AbstractArray interface, in order to ensure that functions can work with array types besides the standard Array type.

The GenericDict can be used to test generic dict APIs that program to the AbstractDict interface, in order to ensure that functions can work with associative types besides the standard Dict type.

The GenericOrder can be used to test APIs for their support of generic ordered types.

The GenericSet can be used to test generic set APIs that program to the AbstractSet interface, in order to ensure that functions can work with set types besides the standard Set and BitSet types.

The GenericString can be used to test generic string APIs that program to the AbstractString interface, in order to ensure that functions can work with string types besides the standard String type.

detect_ambiguities(mod1, mod2...; recursive=false,
                                  ambiguous_bottom=false,
                                  allowed_undefineds=nothing)

Return a vector of (Method,Method) pairs of ambiguous methods defined in the specified modules. Use recursive=true to test in all submodules.

ambiguous_bottom controls whether ambiguities triggered only by Union{} type parameters are included; in most cases you probably want to set this to false. See Base.isambiguous.

See Test.detect_unbound_args for an explanation of allowed_undefineds.

detect_unbound_args(mod1, mod2...; recursive=false, allowed_undefineds=nothing)

Return a vector of Methods which may have unbound type parameters. Use recursive=true to test in all submodules.

By default, any undefined symbols trigger a warning. This warning can be suppressed by supplying a collection of GlobalRefs for which the warning can be skipped. For example, setting

allowed_undefineds = Set([GlobalRef(Base, :active_repl),
                          GlobalRef(Base, :active_repl_backend)])

would suppress warnings about Base.active_repl and Base.active_repl_backend.