mailto:email@example.com. See instructions here to subscribe to the list. You can access previous messages via the archive of the mailing list.
This defines an API for writing test suites, to make it easy to portably test Scheme APIs, libraries, applications, and implementations. A test suite is a collection of test cases that execute in the context of a test-runner. This specifications also supports writing new test-runners, to allow customization of reporting and processing the result of running test suites.
There are other testing frameworks written in Scheme, including SchemeUnit. However SchemeUnit is not portable. It is also a bit on the verbose side. It would be useful to have a bridge between this framework and SchemeUnit so SchemeUnit tests could run under this framework and vice versa. However, that is not part of this specification.
There exists at least one Scheme wrapper providing a Scheme interface
standard JUnit API for Java.
It would be useful to have a bridge so that tests written using this
framework can run under a JUnit runner, and also that existing Scheme tests
run under the current framework.
However, that is not part of this specification.
We should have a testsuite for the testing framework. It should preferably be written using this specification, if that isn't too awkward. At the very least we need complete examples that exercise more of the API.
The implementation should be ported to other featureful Scheme implementations so they can make use of other than the lowest R5RS functionality.
Need to nail down definition of test specifier - specifically how a value gets coerced to a specifier procedure.
The implementation could be polished a bit more.
test-end could pop extra groups on the stack,
if the given group-name doesn't match the current name,
but matches on outer one. This is primarily for recovering from
test-suite errors, or exceptions that aren't caught.
be defined to test each specifier, or to short-circuit? This is
important for side-effecting specifiers, like
Similari;y, when there are multiple active
should we evaluate all or stop when we have a decision?
How should test-suite errors, such as incorrect counts or bad nesting, be handled? An exception? Or a call-back to the test-runner. Currently bad counts are handled using a call-back, but other errors may signal an exception.
Implemented a much more featureful
which necessitate more and better specifued callback functions,
as well as other changes.
Some support for other Scheme implementations by Alex Shinn.
Fixes some typos reported by Donovan Kolby.
Changes parameter order so expected operand comes before result.
Moved optional count from test-end to test-begin.
The PASS, FAIL, XPASS, XFAIL, SKIP symbols were lower-cased.
Added test-approximate, test-error, test-read-eval-string.
Renamed test-match-named to test-match-name.
The Scheme community needs a standard for writing test suites. Every SRFI or other library should come with a test suite. Such a test suite must be portable, without requiring any non-standard features, such as modules. The test suite implementation or "runner" need not be portable, but it is desirable that it be possible to write a portable basic implementation.
This API makes use of implicit dynamic state, including an
test runner. This makes the API convenient
and terse to use, but it may be a little less elegant and
than using explicit test objects, such as JUnit-style frameworks.
It is not claimed to follow either object-oriented or functional design
principles, but I hope it is useful and convenient to use and extend.
This proposal allows converting a Scheme source file to a test suite by just adding a few macros. You don't have to write the entire file in a new form, thus you don't have to re-indent it.
All names defined by the API start with the prefix
(Issue: Perhaps a colon prefix
testing: would be better.)
All function-like forms are defined as syntax. They may be implemented
as functions or macros or built-ins. The reason for specifying them as
syntax is to allow specific tests to be skipped without evaluating sub-expressions, or for implementations
to add features such as printing line numbers or catching exceptions.
Let's start with a simple example. This is a complete self-contained test-suite.
;; Initialize and give a name to a simple testsuite. (test-begin "vec-test") (define v (make-vector 5 99)) ;; Require that an expression evaluate to true. (test-assert (vector? v)) ;; Test that an expression is eqv? to some other expression. (test-eqv 99 (vector-ref v 2)) (vector-set! v 2 7) (test-eqv 7 (vector-ref v 2)) ;; Finish the testsuite, and report results. (test-end "vec-test")
This testsuite could be saved in its own source file. Nothing else is needed: We do not require any top-level forms, so it is easy to wrap an existing program or test to this form, without adding indentation. It is also easy to add new tests, without having to name individual tests (though that is optional).
Test cases are executed in the context of a test runner,
which is a object that accumulates and reports test results.
This specification defines how to create and use custom test runners,
but implementations should also provide a default test runner.
It is suggested (but not required) that loading the above
file in a top-level environment will cause the
tests to be executed using an implementation-specified default test runner,
test-end will cause a summary to be displayed
in an implementation-specified manner.
Primitive test cases test that a given condition is true.
They may have a name.
The core test case form is
(test-assert [test-name] expression)
This evaluates the expression.
The test passes if the result
is true; if the result is false, a test failure is reported.
The test also fails if an exception is raised, assuming the implementation
has a way to catch exceptions.
How the failure is reported depends on the test runner environment.
The test-name is a string that names the test case.
(Though the test-name is a string literal in the examples,
it is an expression. It is evaluated only once.)
It is used when reporting errors, and also when skipping tests,
as described below.
It is an error to invoke
if there is no current test runner.
The following forms may be more convenient than
(test-eqv [test-name] expected test-expr)
This is equivalent to:
(test-assert [test-name] (eqv? expected test-expr))
are shorthand for
test-assert combined with
(test-equal [test-name] expected test-expr) (test-eq [test-name] expected test-expr)
Here is a simple example:
(define (mean x y) (/ (+ x y) 2.0)) (test-eqv 4 (mean 3 5))
For testing approximate equality of inexact reals
we can use
(test-approximate [test-name] expected test-expr error)
This is equivalent to (except that each argument is only evaluated once):
(test-assert [test-name] (and (>= test-expr (- expected error)) (<= test-expr (+ expected error))))
We need a way to specify that evaluation should fail. This verifies that errors are detected when required.
(test-error [[test-name] error-type] test-expr)
Evaluating test-expr is expected to signal an error. The kind of error is indicated by error-type.
If the error-type is left out, or it is
#t, it means "some kind of unspecified error should be signaled".
(test-error #t (vector-ref '#(1 2) 9))
This specification leaves it implementation-defined (or for a future
specification) what form
test-error may take,
though all implementations must allow
Some implementations may support
but these are only standardized for
SRFI-36's I/O conditions, which are seldom useful in test suites.
An implementation may also allow implementation-specific
For example Java-based implementations may allow
the names of Java exception classes:
;; Kawa-specific example (test-error <java.lang.IndexOutOfBoundsException> (vector-ref '#(1 2) 9))
An implementation that cannot catch exceptions should skip
Testing syntax is tricky, especially if we want to check that invalid syntax is causes an error. The following utility function can help:
This function parses string (using
and evaluates the result.
The result of evaluation is returned from
An error is signalled if there are unread characters after the
read is done.
(test-read-eval-string "(+ 3 4)") evaluates to
(test-read-eval-string "(+ 3 4") signals an error.
(test-read-eval-string "(+ 3 4) ") signals an error,
because there is extra
junk (i.e. a space) after the
list is read.
test-read-eval-string used in tests:
(test-equal 7 (test-read-eval-string "(+ 3 4)")) (test-error (test-read-eval-string "(+ 3")) (test-equal #\newline (test-read-eval-string "#\\newline")) (test-error (test-read-eval-string "#\\newlin")) ;; Skip the next 2 tests unless srfi-62 is available. (test-skip (cond-expand (srfi-62 0) (else 2))) (test-equal 5 (test-read-eval-string "(+ 1 #;(* 2 3) 4)")) (test-equal '(x z) (test-read-string "(list 'x #;'y 'z)"))
A test group is a named sequence of forms containing testcases, expressions, and definitions. Entering a group sets the test group name; leaving a group restores the previous group name. These are dynamic (run-time) operations, and a group has no other effect or identity. Test groups are informal groupings: they are neither Scheme values, nor are they syntactic forms.
A test group may contain nested inner test groups. The test group path is a list of the currently-active (entered) test group names, oldest (outermost) first.
(test-begin suite-name [count])
test-begin enters a new test group.
The suite-name becomes the current test group name,
and is added to the end of the test group path.
Portable test suites should use a sting literal for suite-name;
the effect of expressions or other kinds of literals is unspecified.
Rationale: In some ways using symbols would be preferable. However, we want human-readable names, and standard Scheme does not provide a way to include spaces or mixed-case text in literal symbols.
The optional count must match the number of test-cases executed by this group. (Nested test groups count as a single test case for this count.) This extra test may be useful to catch cases where a test doesn't get executed because of some unexpected error.
Additionally, if there is no currently executing test runner, one is installed in an implementation-defined manner.
test-end leaves the current test group.
An error is reported if the suite-name does not
match the current test group name.
Additionally, if the matching
installed a new test-runner, then the
will de-install it, after reporting the accumulated test
results in an implementation-defined manner.
(test-group suite-name decl-or-expr ...)
(if (not (test-to-skip% suite-name)) (dynamic-wind (lambda () (test-begin suite-name)) (lambda () decl-or-expr ...) (lambda () (test-end suite-name))))
This is usually equivalent to executing the decl-or-exprs
within the named test group. However, the entire group is skipped
if it matched an active
test-skip (see later).
test-end is executed in case of an exception.
(test-group-with-cleanup suite-name decl-or-expr ... cleanup-form)
Execute each of the
decl-or-expr forms in order
(as in a
and then execute the
The latter should be executed even if
one of a
decl-or-expr forms raises an exception
(assuming the implementation has a way to catch exceptions).
(test-group-with-cleanup "test-file" (define f (open-output-file "log")) (do-a-bunch-of-tests f) (close-output-port f))
The following describes features for controlling which tests to execute, or specifing that some tests are expected to fail.
Sometimes we want to only run certain tests, or we know that
certain tests are expected to fail.
A test specifier is one-argument function that takes a test-runner
and returns a boolean. The specifier may be run before a test is performed,
and the result may control whether the test is executed.
For convenience, a specifier may also be a non-procedure value,
which is coerced to a specifier procedure, as described below for
A simple example is:
(if some-condition (test-skip 2)) ;; skip next 2 tests
The resulting specifier matches if the current test name (as returned by
(test-match-nth n [count])
This evaluates to a stateful predicate: A counter keeps track of how many times it has been called. The predicate matches the n'th time it is called (where
1 is the first time), and
(- count 1) times,
where count defaults to
(test-match-any specifier ...)
The resulting specifier matches if any specifier matches.
(test-match-all specifier ...)
The resulting specifier matches if each specifier matches.
;; i.e. an integer
Convenience short-hand for:
(test-match-nth 1 count)
;; i.e. a string
Convenience short-hand for
In some cases you may want to skip a test.
test-skip adds the
to the set of currently active skip-specifiers.
Before each test (or
the set of active skip-specifiers are applied to the active test-runner.
If any specifier matches, then the test is skipped.
For convenience, if the specifier is a string that
is syntactic sugar for
(test-skip "test-b") (test-assert "test-a") ;; executed (test-assert "test-b") ;; skipped
Any skip specifiers introduced by a
are removed by a following non-nested
(test-begin "group1") (test-skip "test-a") (test-assert "test-a") ;; skipped (test-end "group1") ;; Undoes the prior test-skip (test-assert "test-a") ;; executed
Sometimes you know a test case will fail, but you don't have time to or can't fix it. Maybe a certain feature only works on certain platforms. However, you want the test-case to be there to remind you to fix it. You want to note that such tests are expected to fail.
Matching tests (where matching is defined as in
are expected to fail. This only affects test reporting,
not test execution. For example:
(test-expect-fail 2) (test-eqv ...) ;; expected to fail (test-eqv ...) ;; expected to fail (test-eqv ...) ;; expected to pass
A test-runner is an object that runs a test-suite, and manages the state. The test group path, and the sets skip and expected-fail specifiers are part of the test-runner. A test-runner will also typically accumulate statistics about executed tests,
Get or set the current test-runner. If an implementation supports parameter objects (as in SRFI-39), then
test-runner-current can be a parameter object.
test-runner-current may be implemented
as a macro or function
that uses a fluid or thread-local variable, or a plain global variable.
Creates a new simple test-runner, that prints errors and a summary on the standard output port.
Creates a new test-runner, that does nothing with the test results. This is mainly meant for extending when writing a custom runner.
Implementations may provide other test-runners, perhaps
Create a new test-runner. Equivalent to
Get or set the current test-runner factory. A factory is a zero-argument function that creates a new test-runner. The default value is
but implementations may provide a way to override the default.
test-runner-current, this may be a parameter object,
or use a per-thread, fluid, or global variable.
(test-apply [runner] specifier ... procedure)
Calls procedure with no arguments using the specified runner as the current test-runner. If runner is omitted, then
(test-runner-current) is used.
(If there is no current runner, one is created as in
If one or more specifiers are listed then only tests matching
the specifiers are executed. A specifier has the same form
as one used for
test-skip. A test is executed
if it matches any of the specifiers in the
test-apply and does not match any
(test-with-runner runner decl-or-expr ...)
Executes each decl-or-expr in order in a context where the current test-runner is runner.
Running a test sets various status properties in the current test-runner. This can be examined by a custom test-runner, or (more rarely) in a test-suite.
Return one of the above result codes from the most recent tests. Returns
#f if no tests have been run yet.
If we've started on a new test, but don't have a result yet,
then the result kind is
'xfail is the test is expected to fail,
'skip is the test is supposed to be skipped,
True if the value of
is one of
This is a convenient shorthand that might be useful
in a test suite to only run certain tests if the previous test passed.
A test runner also maintains a set of more detailed
associated with the current or most recent test. (I.e. the properties of the
most recent test are available as long as a new test hasn't started.)
Each property has a name (a symbol) and a value (any value).
Some properties are standard or set by the implementation;
implementations can add more.
(test-result-ref runner 'pname [default])
Returns the property value associated with the pname property name. If there is no value associated with
#f if default isn't specified.
(test-result-set! runner 'pname value)
Sets the property value associated with the pname property name to value. Usually implementation code should call this function, but it may be useful for a custom test-runner to add extra properties.
(test-result-remove runner 'pname)
Remove the property with the name
Remove all result properties. The implementation automatically calls
at the start of a
test-assert and similar procedures.
Returns an association list of the current result properties. It is unspecified if the result shares state with the test-runner. The result should not be modified, on the other hand the result may be implicitly modified by future
test-result-clear does not modify the returned
alist. Thus you can
archive result objects from previous runs.
The set of available result properties is implementation-specific. However, it is suggested that the following might be provided:
(test-result-kind runner)is equivalent to:
(test-result-ref runner 'result-kind)
test-assert) in test suite source code..
error-typespecified in a
test-error, if it meaningful and known.
This section specifies how to write a test-runner. It can be ignored if you just want to write test-cases.
These call-back functions are
methods (in the object-oriented sense)
of a test-runner. A method
is called by the implementation when event happens.
To define (set) the callback function for event use the following expression.
(This is normally done when initializing a test-runner.)
(test-runner-on-event! runner event-function)
An event-function takes a test-runner argument, and possibly other arguments, depending on the event.
To extract (get) the callback function for event do this:
To extract call the callback function for event use the following expression.
(This is normally done by the implementation core.)
((test-runner-on-event runner) runner other-args ...)
The following call-back hooks are available.
(test-runner-on-test-begin! runner on-test-begin-function)
The on-test-begin-function is called at the start of an individual testcase, before the test expression (and expected value) are evaluated.
(test-runner-on-test-end! runner on-test-end-function)
The on-test-end-function is called at the end of an individual testcase, when the result of the test is available.
(test-runner-on-group-begin! runner on-group-begin-function)
(on-group-begin-function runner suite-name count)
The on-group-begin-function is called by a
including at the start of a
The suite-name is a Scheme string,
and count is an integer or
(test-runner-on-group-end! runner on-group-end-function)
The on-group-end-function is called by a
including at the end of a
(test-runner-on-bad-count! runner on-bad-count-function)
(on-bad-count-function runner actual-count expected-count)
test-end (before the on-group-end-function
is called) if an expected-count was specified by the matching
test-begin and the expected-count does not match
the actual-count of tests actually executed or skipped.
(test-runner-on-final! runner on-final-function)
The on-final-function takes one parameter (a test-runner) and typically displays a summary (count) of the tests. The on-final-function is called after called the on-group-end-function correspondiong to the outermost
The default value is
test-on-final-simple which writes
to the standard output port the number of tests of the various kinds.
The following functions are for accessing the other components of a test-runner. They would normally only be used to write a new test-runner or a match-predicate.
Returns the number of tests that passed, and were expected to pass.
Returns the number of tests that failed, but were expected to pass.
Returns the number of tests that passed, but were expected to fail.
Returns the number of tests that failed, and were expected to pass.
Returns the number of tests or test groups that were skipped.
Returns the name of the current test or test group, as a string. During execution of
test-begin this is the name of the
test group; during the execution of an actual test, this is the name
of the test-case.
If no name was specified, the name is the empty string.
A list of names of groups we're nested in, with the outermost group first.
A list of names of groups we're nested in, with the outermost group last. (This is more efficient than
since it doesn't require any copying.)
(test-runner-aux-value! runner on-test)
Get or set the
aux-value field of a test-runner.
This field is not used by this API or the
test-runner, but may be used by custom test-runners to store extra state.
Resets the state of the runner to its initial state.
This is an example of a simple custom test-runner. Loading this program before running a test-suite will install it as the default test runner.
(define (my-simple-runner filename) (let ((runner (test-runner-null)) (port (open-output-file filename)) (num-passed 0) (num-failed 0)) (test-runner-on-test! runner (lambda (runner result) (case (cdr (assq 'result-kind result)) ((pass xpass) (set! num-passed (+ num-passed 1))) ((fail xfail) (set! num-failed (+ num-failed 1))) (else #t)))) (test-runner-on-final! runner (lambda (runner) (format port "Passing tests: ~d.~%Failing tests: ~d.~%" num-passed num-failed) (close-output-port port))) runner)) (test-runner-factory (lambda () (my-simple-runner "/tmp/my-test.log")))
The test implementation uses
to select different code depending on certain SRFI names (
or implementations (
It should otherwise be portable to any R5RS implementation.
srfi-25-test.scm, based converted from Jussi Piitulainen's
Of course we need a test suite for the testing framework itself.
Copyright (C) Per Bothner (2005, 2006)
Permission is hereby granted, free of charge, to any person obtaining
a copy of this software and associated documentation files (the
"Software"), to deal in the Software without restriction, including
without limitation the rights to use, copy, modify, merge, publish,
distribute, sublicense, and/or sell copies of the Software, and to
permit persons to whom the Software is furnished to do so, subject to
the following conditions:
The above copyright notice and this permission notice shall be
included in all copies or substantial portions of the Software.
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE
LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION
OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION
WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
was contributed by Donovan Kolbly
Per Bothner Last modified: Tue Feb 21 16:20:11 PST 2006
Copyright (C) Per Bothner (2005, 2006)
Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:
The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.