3 RackUnit API
(require rackunit) |
3.1 Overview of RackUnit
There are three basic concepts in RackUnit:
A check is the basic unit of a test. As the name suggests, it checks some condition is true.
A test case is a group of checks that form one conceptual unit. If any check within the case fails, the entire case fails.
A test suite is a group of test cases and test suites that has a name.
3.2 Checks
Checks are the basic building block of RackUnit. A check checks some condition. If the condition holds the check evaluates to (void). If the condition doesn’t hold the check raises an instance of exn:test:check with information detailing the failure.
Although checks are implemented as macros, which is necessary to grab source location, they are conceptually functions. This means, for instance, checks always evaluate their arguments. You can use check as first class functions, though you will lose precision in the reported source locations if you do so.
The following are the basic checks RackUnit provides. You can create your own checks using define-check.
(check-eq? v1 v2 [message]) → void? |
v1 : any |
v2 : any |
message : string? = "" |
(check-not-eq? v1 v2 [message]) → void? |
v1 : any |
v2 : any |
message : string? = "" |
(check-eqv? v1 v2 [message]) → void? |
v1 : any |
v2 : any |
message : string? = "" |
(check-not-eqv? v1 v2 [message]) → void? |
v1 : any |
v2 : any |
message : string? = "" |
(check-equal? v1 v2 [message]) → void? |
v1 : any |
v2 : any |
message : string? = "" |
(check-not-equal? v1 v2 [message]) → void? |
v1 : any |
v2 : any |
message : string? = "" |
For example, the following checks all fail:
(check-eq? (list 1) (list 1) "allocated data not eq?") |
(check-not-eq? 1 1 "fixnums are eq?") |
(check-eqv? 1 1.0 "not eqv?") |
(check-not-eqv? 1 1 "integers are eqv?") |
(check-equal? 1 1.0 "not equal?") |
(check-not-equal? (list 1) (list 1) "equal?") |
(check-= v1 v2 epsilon [message]) → void? |
v1 : any |
v2 : any |
epsilon : number? |
message : string? = "" |
For example, the following check passes:
(check-true v [message]) → void? |
v : any |
message : string? = "" |
(check-false v [message]) → void? |
v : any |
message : string? = "" |
(check-not-false v [message]) → void? |
v : any |
message : string? = "" |
For example, the following checks all fail:
(check-true 1) |
(check-false 1) |
(check-not-false #f) |
(check-exn exn-predicate thunk [message]) → void? |
exn-predicate : (or/c (-> any boolean?) regexp?) |
thunk : (-> any) |
message : string? = "" |
For example, the following check succeeds:
(check-exn exn:fail? |
(lambda () |
(raise (make-exn "Hi there" |
(current-continuation-marks))))) |
The following check fails:
(check-exn exn:fail? |
(lambda () |
(break-thread (current-thread)))) |
The following example is a common mistake. The call to error is not within a lambda, so it bypasses check-exn entirely.
; Forgot to wrap the expression in a thunk. Don't do this! |
(check-exn exn:fail? |
(error 'hi "there")) |
(check-not-exn thunk [message]) → void? |
thunk : (-> any) |
message : string? = "" |
(check-regexp-match regexp string) → void? |
regexp : regexp? |
string : string? |
For example, the following check succeeds:
(check-regexp-match "a+bba" "aaaaaabba")
The following check fails:
(check-regexp-match "a+bba" "aaaabbba")
For example, the following check succeeds:
The following check fails:
(check memq 'pine '(apple orange pear))
3.2.1 Augmenting Information on Check Failure
When a check fails it stores information including the name of the check, the location and message (if available), the expression the check is called with, and the parameters to the check. Additional information can be stored by using the with-check-info* function, and the with-check-info macro.
| ||||||
name : symbol? | ||||||
value : any |
The are several predefined functions that create check information structures with predefined names. This avoids misspelling errors:
(make-check-name name) → check-info? | ||||||||
name : string? | ||||||||
(make-check-params params) → check-info? | ||||||||
params : (listof any) | ||||||||
(make-check-location loc) → check-info? | ||||||||
| ||||||||
(make-check-expression msg) → check-info? | ||||||||
msg : any | ||||||||
(make-check-message msg) → check-info? | ||||||||
msg : string? | ||||||||
(make-check-actual param) → check-info? | ||||||||
param : any | ||||||||
(make-check-expected param) → check-info? | ||||||||
param : any |
(with-check-info* info thunk) → any |
info : (listof check-info?) |
thunk : (-> any) |
Example:
(with-check-info* |
(list (make-check-info 'time (current-seconds))) |
(lambda () (check = 1 2))) |
When this check fails the message
time: <current-seconds-at-time-of-running-check> |
is printed along with the usual information on an check failure.
(with-check-info ((name val) ...) body ...) |
Example:
(for-each |
(lambda (elt) |
(with-check-info |
(('current-element elt)) |
(check-pred odd? elt))) |
(list 1 3 5 7 8)) |
When this test fails the message
current-element: 8 |
is displayed along with the usual information on an check failure.
3.2.2 Custom Checks
Custom checks can be defined using define-check and its variants. To effectively use these macros it is useful to understand a few details about a check’s evaluation model.
Firstly, a check should be considered a function, even though most uses are actually macros. In particular, checks always evaluate their arguments exactly once before executing any expressions in the body of the checks. Hence if you wish to write checks that evaluate user defined code that code must be wrapped in a thunk (a function of no arguments) by the user. The predefined check-exn is an example of this type of check.
It is also useful to understand how the check information stack operates. The stack is stored in a parameter and the with-check-info forms evaluate to calls to parameterize. For this reason simple checks (see below) cannot usefully contain calls to with-check-info to report additional information. All checks created using define-simple-check or define-check grab some information by default: the name of the checks and the values of the parameters. Additionally the macro forms of checks grab location information and the expressions passed as parameters.
(define-simple-check (name param ...) expr ...) |
For example, the following code defines a check check-odd?
(define-simple-check (check-odd? number) |
(odd? number)) |
We can use these checks in the usual way:
(check-odd? 3) ; Success |
(check-odd? 2) ; Failure |
(define-binary-check (name pred actual expected)) |
(define-binary-check (name actual expected) expr ...) |
Examples:
Here’s the first form, where we use a predefined predicate to construct a binary check:
(define-binary-check (check-char=? char=? actual expected))
In use:
(check-char=? (read-char a-port) #\a)
If the expression is more complicated the second form should be used. For example, below we define a binary check that tests a number if within 0.01 of the expected value:
(define-binary-check (check-in-tolerance actual expected) |
(< (abs (- actual expected)) 0.01)) |
(define-check (name param ...) expr ...) |
3.3 Compound Testing Forms
3.3.1 Test Cases
As programs increase in complexity the unit of testing grows beyond a single check. For example, it may be the case that if one check fails it doesn’t make sense to run another. To solve this problem compound testing forms can be used to group expressions. If any expression in a group fails (by raising an exception) the remaining expressions will not be evaluated.
(test-begin expr ...) |
For example, in the following code the world is not destroyed as the preceding check fails:
(test-begin |
(check-eq? 'a 'b) |
; This line won't be run |
(destroy-the-world)) |
(test-case name expr ...) |
Here’s the above example rewritten to use test-case so the test can be named.
(test-case |
"Example test" |
(check-eq? 'a 'b) |
; This line won't be run |
(destroy-the-world)) |
(test-case? obj) → boolean? |
obj : any |
3.3.1.1 Shortcuts for Defining Test Cases
(test-check name operator v1 v2) → void? |
name : string? |
operator : (-> any/c any/c any/c) |
v1 : any/c |
v2 : any/c |
(test-pred name pred v) → void? |
name : string? |
pred : (-> any/c any/c) |
v : any/c |
(test-equal? name v1 v2) → (void?) |
name : string? |
v1 : any/c |
v2 : any/c |
(test-eq? name v1 v2) → void? |
name : string? |
v1 : any/c |
v2 : any/c |
(test-eqv? name v1 v2) → void? |
name : string? |
v1 : any/c |
v2 : any/c |
(test-= name v1 v2 epsilon) → void? |
name : string? |
v1 : real? |
v2 : real? |
epsilon : real? |
(test-true name v) → void? |
name : string? |
v : any/c |
(test-false name v) → void? |
name : string? |
v : any/c |
(test-not-false name v) → void? |
name : string? |
v : any/c |
(test-exn name pred thunk) → void? |
name : string? |
pred : (-> exn? any/c) |
thunk : (-> any) |
(test-not-exn name thunk) → void? |
name : string? |
thunk : (-> any) |
3.3.2 Test Suites
Test cases can themselves be grouped into test suites. A test suite can contain both test cases and test suites. Unlike a check or test case, a test suite is not immediately run. Instead use one of the functions described in User Interfaces or Programmatically Running Tests and Inspecting Results.
(test-suite name-expr maybe-before maybe-after test ...) | |||||||||||||||||||||||||||
| |||||||||||||||||||||||||||
|
The before-thunk and after-thunk are optional thunks (functions with no argument). They are run before and after the tests are run, respectively.
Unlike a check or test case, a test suite is not immediately run. Instead use one of the functions described in User Interfaces or Programmatically Running Tests and Inspecting Results.
For example, here is a test suite that displays Before before any tests are run, and After when the tests have finished.
(test-suite |
"An example suite" |
#:before (lambda () (display "Before")) |
#:after (lambda () (display "After")) |
(test-case |
"An example test" |
(check-eq? 1 1)) |
(test-suite "A nested test suite" |
(test-case "Another test" |
(check-< 1 2)))) |
| ||||||||||||||||||||||||||||
name : string? | ||||||||||||||||||||||||||||
tests : (listof (or/c test-case? test-suite?)) | ||||||||||||||||||||||||||||
before-thunk : (-> any) = void | ||||||||||||||||||||||||||||
after-thunk : (-> any) = void |
(test-suite? obj) → boolean? |
obj : any |
3.3.2.1 Utilities for Defining Test Suites
There are some macros that simplify the common cases of defining test suites:
(define-test-suite name test ...) |
For example, this code creates a binding for the name example-suite as well as creating a test suite with the name "example-suite":
(define-test-suite example-suite |
(check = 1 1)) |
(define/provide-test-suite name test ...) |
3.4 Test Control Flow
The before, after, and around macros allow you to specify code that is always run before, after, or around expressions in a test case.
(before before-expr expr1 expr2 ...) |
(after expr-1 expr-2 ... after-expr) |
(around before-expr expr-1 expr-2 ... after-expr) |
Example:
The test below checks that the file test.dat contains the string "foo". The before action writes to this file. The after action deletes it.
(around |
(with-output-to-file "test.dat" |
(lambda () |
(write "foo"))) |
(with-input-from-file "test.dat" |
(lambda () |
(check-equal? "foo" (read)))) |
(delete-file "test.dat")) |
(delay-test test1 test2 ...) |
3.5 Miscellaneous Utilities
The require/expose macro allows you to access bindings that a module does not provide. It is useful for testing the private functions of modules.
(require/expose module (id ...)) |
Note that require/expose can be a bit fragile, especially when mixed with compiled code. Use at your own risk!
This example gets make-failure-test, which is defined in a RackUnit test:
(require/expose rackunit/private/check-test (make-failure-test))
3.6 User Interfaces
RackUnit provides a textual and a graphical user interface
3.6.1 Textual User Interface
(require rackunit/text-ui) |
The textual UI is in the rackunit/text-ui module. It is run via the run-tests function.
(run-tests test [verbosity]) → natural-number/c |
test : (or/c test-case? test-suite?) |
verbosity : (symbols 'quiet 'normal 'verbose) = 'normal |
The optional verbosity is one of 'quiet, 'normal, or 'verbose. Quiet output displays only the number of successes, failures, and errors. Normal reporting suppresses some extraneous check information (such as the expression). Verbose reports all information.
run-tests returns the number of unsuccessful tests.
3.6.2 Graphical User Interface
(require rackunit/gui) |
RackUnit also provides a GUI test runner, available from the rackunit/gui module.
(test/gui test ...) → any |
test : (or/c test-case? test-suite?) |
(make-gui-runner) → (-> (or/c test-case? test-suite?) ... any) |