Up

module OUnit

: sig

Unit test building blocks

Author Maas-Maarten Zeeman
Author Sylvain Le Gall

Assertions

Assertions are the basic building blocks of unittests.

#
val assert_failure : string -> 'a

Signals a failure. This will raise an exception with the specified string.

Raises Failure signal a failure
#
val assert_bool : string -> bool -> unit

Signals a failure when bool is false. The string identifies the failure.

Raises Failure signal a failure
#
val (@?) : string -> bool -> unit

Shorthand for assert_bool

Raises Failure to signal a failure
#
val assert_string : string -> unit

Signals a failure when the string is non-empty. The string identifies the failure.

Raises Failure signal a failure
#
val assert_command : ?exit_code:Unix.process_status -> ?sinput:char Stream.t -> ?foutput:(char Stream.t -> unit) -> ?use_stderr:bool -> ?env:string array -> ?verbose:bool -> string -> string list -> unit

assert_command prg args Run the command provided.

exit_code expected exit code
sinput provide this char Stream.t as input of the process
foutput run this function on output, it can contains an assert_equal to check it
use_stderr redirect stderr to stdout
env Unix environment
verbose if failed, dump stdout/stderr of the process to stderr
Since 1.1.0
#
val assert_equal : ?cmp:('a -> 'a -> bool) -> ?printer:('a -> string) -> ?pp_diff:(Format.formatter -> 'a * 'a -> unit) -> ?msg:string -> 'a -> 'a -> unit

assert_equal expected real Compares two values, when they are not equal a failure is signaled.

cmp customize function to compare, default is =
printer value printer, don't print value otherwise
pp_diff if not equal, ask a custom display of the difference using diff fmt exp real where fmt is the formatter to use
msg custom message to identify the failure
Raises Failure signal a failure
Version 1.1.0
#
val assert_raises : ?msg:string -> exn -> (unit -> 'a) -> unit

Asserts if the expected exception was raised.

msg identify the failure
Raises Failure description

Skipping tests

In certain condition test can be written but there is no point running it, because they are not significant (missing OS features for example). In this case this is not a failure nor a success. Following functions allow you to escape test, just as assertion but without the same error status.

A test skipped is counted as success. A test todo is counted as failure.

#
val skip_if : bool -> string -> unit

skip cond msg If cond is true, skip the test for the reason explain in msg. For example skip_if (Sys.os_type = "Win32") "Test a doesn't run on windows".

Since 1.0.3
#
val todo : string -> unit

The associated test is still to be done, for the reason given.

Since 1.0.3

Compare Functions

#
val cmp_float : ?epsilon:float -> float -> float -> bool

Compare floats up to a given relative error.

epsilon if the difference is smaller epsilon values are equal

Bracket

A bracket is a functional implementation of the commonly used setUp and tearDown feature in unittests. It can be used like this:

"MyTestCase" >:: (bracket test_set_up test_fun test_tear_down)

#
val bracket : (unit -> 'a) -> ('a -> unit) -> ('a -> unit) -> unit -> unit

bracket set_up test tear_down The set_up function runs first, then the test function runs and at the end tear_down runs. The tear_down function runs even if the test failed and help to clean the environment.

#
val bracket_tmpfile : ?prefix:string -> ?suffix:string -> ?mode:Pervasives.open_flag list -> (string * Pervasives.out_channel -> unit) -> unit -> unit

bracket_tmpfile test The test function takes a temporary filename and matching output channel as arguments. The temporary file is created before the test and removed after the test.

prefix see Filename.open_temp_file
suffix see Filename.open_temp_file
mode see Filename.open_temp_file
Since 1.1.0

Constructing Tests

#
type test_fun = unit -> unit

The type of test function

#
type test =
# | TestCase of test_fun
# | TestList of test list
# | TestLabel of string * test

The type of tests

#
val (>:) : string -> test -> test

Create a TestLabel for a test

#
val (>::) : string -> test_fun -> test

Create a TestLabel for a TestCase

#
val (>:::) : string -> test list -> test

Create a TestLabel for a TestList

Some shorthands which allows easy test construction.

Examples:

  • "test1" >: TestCase((fun _ -> ())) => TestLabel("test2", TestCase((fun _ -> ())))
  • "test2" >:: (fun _ -> ()) => TestLabel("test2", TestCase((fun _ -> ())))
  • "test-suite" >::: ["test2" >:: (fun _ -> ());] => TestLabel("test-suite", TestSuite([TestLabel("test2", TestCase((fun _ -> ())))]))
#
val test_decorate : (test_fun -> test_fun) -> test -> test

test_decorate g tst Apply g to test function contains in tst tree.

Since 1.0.3
#
val test_filter : ?skip:bool -> string list -> test -> test option

test_filter paths tst Filter test based on their path string representation.

skip if set, just use skip_if for the matching tests.
Since 1.0.3

Retrieve Information from Tests

#
val test_case_count : test -> int

Returns the number of available test cases

#
type node =
# | ListItem of int
# | Label of string

Types which represent the path of a test

#
type path = node list

The path to the test (in reverse order).

#
val string_of_node : node -> string

Make a string from a node

#
val string_of_path : path -> string

Make a string from a path. The path will be reversed before it is tranlated into a string

#
val test_case_paths : test -> path list

Returns a list with paths of the test

Performing Tests

#
type test_result =
# | RSuccess of path
# | RFailure of path * string
# | RError of path * string
# | RSkip of path * string
# | RTodo of path * string

The possible results of a test

#
type test_event =
# | EStart of path
(*A test start.*)
# | EEnd of path
(*A test end.*)
# | EResult of test_result
(*Result of a test.*)

Events which occur during a test run.

#
type test_results = test_result list

Results of a test run.

#
val perform_test : (test_event -> unit) -> test -> test_results

Perform the test, allows you to build your own test runner

#
val run_test_tt : ?verbose:bool -> test -> test_results

A simple text based test runner.

verbose print verbose message
#
val run_test_tt_main : ?arg_specs:(Arg.key * Arg.spec * Arg.doc) list -> ?set_verbose:(bool -> unit) -> test -> test_results

Main version of the text based test runner. It reads the supplied command line arguments to set the verbose level and limit the number of test to run.

arg_specs add extra command line arguments
set_verbose call a function to set verbosity
fexit call a final function after test, by default exit 1.
Version 1.1.0
#
val ounit2_of_ounit1 : test -> OUnit2.test
end