(number->string): Added tests for non-radix-10 floating point
[bpt/guile.git] / test-suite / lib.scm
1 ;;;; test-suite/lib.scm --- generic support for testing
2 ;;;; Copyright (C) 1999, 2000, 2001 Free Software Foundation, Inc.
3 ;;;;
4 ;;;; This program is free software; you can redistribute it and/or modify
5 ;;;; it under the terms of the GNU General Public License as published by
6 ;;;; the Free Software Foundation; either version 2, or (at your option)
7 ;;;; any later version.
8 ;;;;
9 ;;;; This program is distributed in the hope that it will be useful,
10 ;;;; but WITHOUT ANY WARRANTY; without even the implied warranty of
11 ;;;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
12 ;;;; GNU General Public License for more details.
13 ;;;;
14 ;;;; You should have received a copy of the GNU General Public License
15 ;;;; along with this software; see the file COPYING. If not, write to
16 ;;;; the Free Software Foundation, Inc., 59 Temple Place, Suite 330,
17 ;;;; Boston, MA 02111-1307 USA
18
19 (define-module (test-suite lib)
20 :use-module (ice-9 stack-catch)
21 :use-module (ice-9 regex)
22 :export (
23
24 ;; Exceptions which are commonly being tested for.
25 exception:bad-variable
26 exception:missing-expression
27 exception:out-of-range exception:unbound-var
28 exception:wrong-num-args exception:wrong-type-arg
29
30 ;; Reporting passes and failures.
31 run-test
32 pass-if expect-fail
33 pass-if-exception expect-fail-exception
34
35 ;; Naming groups of tests in a regular fashion.
36 with-test-prefix with-test-prefix* current-test-prefix
37 format-test-name
38
39 ;; Reporting results in various ways.
40 register-reporter unregister-reporter reporter-registered?
41 make-count-reporter print-counts
42 make-log-reporter
43 full-reporter
44 user-reporter))
45
46
47 ;;;; If you're using Emacs's Scheme mode:
48 ;;;; (put 'with-test-prefix 'scheme-indent-function 1)
49
50 \f
51 ;;;; CORE FUNCTIONS
52 ;;;;
53 ;;;; The function (run-test name expected-result thunk) is the heart of the
54 ;;;; testing environment. The first parameter NAME is a unique name for the
55 ;;;; test to be executed (for an explanation of this parameter see below under
56 ;;;; TEST NAMES). The second parameter EXPECTED-RESULT is a boolean value
57 ;;;; that indicates whether the corresponding test is expected to pass. If
58 ;;;; EXPECTED-RESULT is #t the test is expected to pass, if EXPECTED-RESULT is
59 ;;;; #f the test is expected to fail. Finally, THUNK is the function that
60 ;;;; actually performs the test. For example:
61 ;;;;
62 ;;;; (run-test "integer addition" #t (lambda () (= 2 (+ 1 1))))
63 ;;;;
64 ;;;; To report success, THUNK should either return #t or throw 'pass. To
65 ;;;; report failure, THUNK should either return #f or throw 'fail. If THUNK
66 ;;;; returns a non boolean value or throws 'unresolved, this indicates that
67 ;;;; the test did not perform as expected. For example the property that was
68 ;;;; to be tested could not be tested because something else went wrong.
69 ;;;; THUNK may also throw 'untested to indicate that the test was deliberately
70 ;;;; not performed, for example because the test case is not complete yet.
71 ;;;; Finally, if THUNK throws 'unsupported, this indicates that this test
72 ;;;; requires some feature that is not available in the configured testing
73 ;;;; environment. All other exceptions thrown by THUNK are considered as
74 ;;;; errors.
75 ;;;;
76 ;;;;
77 ;;;; Convenience macros for tests expected to pass or fail
78 ;;;;
79 ;;;; * (pass-if name body) is a short form for
80 ;;;; (run-test name #t (lambda () body))
81 ;;;; * (expect-fail name body) is a short form for
82 ;;;; (run-test name #f (lambda () body))
83 ;;;;
84 ;;;; For example:
85 ;;;;
86 ;;;; (pass-if "integer addition" (= 2 (+ 1 1)))
87 ;;;;
88 ;;;;
89 ;;;; Convenience macros to test for exceptions
90 ;;;;
91 ;;;; The following macros take exception parameters which are pairs
92 ;;;; (type . message), where type is a symbol that denotes an exception type
93 ;;;; like 'wrong-type-arg or 'out-of-range, and message is a string holding a
94 ;;;; regular expression that describes the error message for the exception
95 ;;;; like "Argument .* out of range".
96 ;;;;
97 ;;;; * (pass-if-exception name exception body) will pass if the execution of
98 ;;;; body causes the given exception to be thrown. If no exception is
99 ;;;; thrown, the test fails. If some other exception is thrown, is is an
100 ;;;; error.
101 ;;;; * (expect-fail-exception name exception body) will pass unexpectedly if
102 ;;;; the execution of body causes the given exception to be thrown. If no
103 ;;;; exception is thrown, the test fails expectedly. If some other
104 ;;;; exception is thrown, it is an error.
105
106 \f
107 ;;;; TEST NAMES
108 ;;;;
109 ;;;; Every test in the test suite has a unique name, to help
110 ;;;; developers find tests that are failing (or unexpectedly passing),
111 ;;;; and to help gather statistics.
112 ;;;;
113 ;;;; A test name is a list of printable objects. For example:
114 ;;;; ("ports.scm" "file" "read and write back list of strings")
115 ;;;; ("ports.scm" "pipe" "read")
116 ;;;;
117 ;;;; Test names may contain arbitrary objects, but they always have
118 ;;;; the following properties:
119 ;;;; - Test names can be compared with EQUAL?.
120 ;;;; - Test names can be reliably stored and retrieved with the standard WRITE
121 ;;;; and READ procedures; doing so preserves their identity.
122 ;;;;
123 ;;;; For example:
124 ;;;;
125 ;;;; (pass-if "simple addition" (= 4 (+ 2 2)))
126 ;;;;
127 ;;;; In that case, the test name is the list ("simple addition").
128 ;;;;
129 ;;;; In the case of simple tests the expression that is tested would often
130 ;;;; suffice as a test name by itself. Therefore, the convenience macros
131 ;;;; pass-if and expect-fail provide a shorthand notation that allows to omit
132 ;;;; a test name in such cases.
133 ;;;;
134 ;;;; * (pass-if expression) is a short form for
135 ;;;; (run-test 'expression #t (lambda () expression))
136 ;;;; * (expect-fail expression) is a short form for
137 ;;;; (run-test 'expression #f (lambda () expression))
138 ;;;;
139 ;;;; For example:
140 ;;;;
141 ;;;; (pass-if (= 2 (+ 1 1)))
142 ;;;;
143 ;;;; The WITH-TEST-PREFIX syntax and WITH-TEST-PREFIX* procedure establish
144 ;;;; a prefix for the names of all tests whose results are reported
145 ;;;; within their dynamic scope. For example:
146 ;;;;
147 ;;;; (begin
148 ;;;; (with-test-prefix "basic arithmetic"
149 ;;;; (pass-if "addition" (= (+ 2 2) 4))
150 ;;;; (pass-if "subtraction" (= (- 4 2) 2)))
151 ;;;; (pass-if "multiplication" (= (* 2 2) 4)))
152 ;;;;
153 ;;;; In that example, the three test names are:
154 ;;;; ("basic arithmetic" "addition"),
155 ;;;; ("basic arithmetic" "subtraction"), and
156 ;;;; ("multiplication").
157 ;;;;
158 ;;;; WITH-TEST-PREFIX can be nested. Each WITH-TEST-PREFIX postpends
159 ;;;; a new element to the current prefix:
160 ;;;;
161 ;;;; (with-test-prefix "arithmetic"
162 ;;;; (with-test-prefix "addition"
163 ;;;; (pass-if "integer" (= (+ 2 2) 4))
164 ;;;; (pass-if "complex" (= (+ 2+3i 4+5i) 6+8i)))
165 ;;;; (with-test-prefix "subtraction"
166 ;;;; (pass-if "integer" (= (- 2 2) 0))
167 ;;;; (pass-if "complex" (= (- 2+3i 1+2i) 1+1i))))
168 ;;;;
169 ;;;; The four test names here are:
170 ;;;; ("arithmetic" "addition" "integer")
171 ;;;; ("arithmetic" "addition" "complex")
172 ;;;; ("arithmetic" "subtraction" "integer")
173 ;;;; ("arithmetic" "subtraction" "complex")
174 ;;;;
175 ;;;; To print a name for a human reader, we DISPLAY its elements,
176 ;;;; separated by ": ". So, the last set of test names would be
177 ;;;; reported as:
178 ;;;;
179 ;;;; arithmetic: addition: integer
180 ;;;; arithmetic: addition: complex
181 ;;;; arithmetic: subtraction: integer
182 ;;;; arithmetic: subtraction: complex
183 ;;;;
184 ;;;; The Guile benchmarks use with-test-prefix to include the name of
185 ;;;; the source file containing the test in the test name, to help
186 ;;;; developers to find failing tests, and to provide each file with its
187 ;;;; own namespace.
188
189 \f
190 ;;;; REPORTERS
191 ;;;;
192 ;;;; A reporter is a function which we apply to each test outcome.
193 ;;;; Reporters can log results, print interesting results to the
194 ;;;; standard output, collect statistics, etc.
195 ;;;;
196 ;;;; A reporter function takes two mandatory arguments, RESULT and TEST, and
197 ;;;; possibly additional arguments depending on RESULT; its return value
198 ;;;; is ignored. RESULT has one of the following forms:
199 ;;;;
200 ;;;; pass - The test named TEST passed.
201 ;;;; Additional arguments are ignored.
202 ;;;; upass - The test named TEST passed unexpectedly.
203 ;;;; Additional arguments are ignored.
204 ;;;; fail - The test named TEST failed.
205 ;;;; Additional arguments are ignored.
206 ;;;; xfail - The test named TEST failed, as expected.
207 ;;;; Additional arguments are ignored.
208 ;;;; unresolved - The test named TEST did not perform as expected, for
209 ;;;; example the property that was to be tested could not be
210 ;;;; tested because something else went wrong.
211 ;;;; Additional arguments are ignored.
212 ;;;; untested - The test named TEST was not actually performed, for
213 ;;;; example because the test case is not complete yet.
214 ;;;; Additional arguments are ignored.
215 ;;;; unsupported - The test named TEST requires some feature that is not
216 ;;;; available in the configured testing environment.
217 ;;;; Additional arguments are ignored.
218 ;;;; error - An error occurred while the test named TEST was
219 ;;;; performed. Since this result means that the system caught
220 ;;;; an exception it could not handle, the exception arguments
221 ;;;; are passed as additional arguments.
222 ;;;;
223 ;;;; This library provides some standard reporters for logging results
224 ;;;; to a file, reporting interesting results to the user, and
225 ;;;; collecting totals.
226 ;;;;
227 ;;;; You can use the REGISTER-REPORTER function and friends to add
228 ;;;; whatever reporting functions you like. If you don't register any
229 ;;;; reporters, the library uses FULL-REPORTER, which simply writes
230 ;;;; all results to the standard output.
231
232 \f
233 ;;;; MISCELLANEOUS
234 ;;;;
235
236 ;;; Define some exceptions which are commonly being tested for.
237 (define exception:bad-variable
238 (cons 'syntax-error "Bad variable"))
239 (define exception:missing-expression
240 (cons 'misc-error "^missing or extra expression"))
241 (define exception:out-of-range
242 (cons 'out-of-range "^Argument .*out of range"))
243 (define exception:unbound-var
244 (cons 'unbound-variable "^Unbound variable"))
245 (define exception:wrong-num-args
246 (cons 'wrong-number-of-args "^Wrong number of arguments"))
247 (define exception:wrong-type-arg
248 (cons 'wrong-type-arg "^Wrong type argument"))
249
250 ;;; Display all parameters to the default output port, followed by a newline.
251 (define (display-line . objs)
252 (for-each display objs)
253 (newline))
254
255 ;;; Display all parameters to the given output port, followed by a newline.
256 (define (display-line-port port . objs)
257 (for-each (lambda (obj) (display obj port)) objs)
258 (newline port))
259
260 \f
261 ;;;; CORE FUNCTIONS
262 ;;;;
263
264 ;;; The central testing routine.
265 ;;; The idea is taken from Greg, the GNUstep regression test environment.
266 (define run-test #f)
267 (let ((test-running #f))
268 (define (local-run-test name expect-pass thunk)
269 (if test-running
270 (error "Nested calls to run-test are not permitted.")
271 (let ((test-name (full-name name)))
272 (set! test-running #t)
273 (catch #t
274 (lambda ()
275 (let ((result (thunk)))
276 (if (eq? result #t) (throw 'pass))
277 (if (eq? result #f) (throw 'fail))
278 (throw 'unresolved)))
279 (lambda (key . args)
280 (case key
281 ((pass)
282 (report (if expect-pass 'pass 'upass) test-name))
283 ((fail)
284 (report (if expect-pass 'fail 'xfail) test-name))
285 ((unresolved untested unsupported)
286 (report key test-name))
287 ((quit)
288 (report 'unresolved test-name)
289 (quit))
290 (else
291 (report 'error test-name (cons key args))))))
292 (set! test-running #f))))
293 (set! run-test local-run-test))
294
295 ;;; A short form for tests that are expected to pass, taken from Greg.
296 (defmacro pass-if (name . rest)
297 (if (and (null? rest) (pair? name))
298 ;; presume this is a simple test, i.e. (pass-if (even? 2))
299 ;; where the body should also be the name.
300 `(run-test ',name #t (lambda () ,name))
301 `(run-test ,name #t (lambda () ,@rest))))
302
303 ;;; A short form for tests that are expected to fail, taken from Greg.
304 (defmacro expect-fail (name . rest)
305 (if (and (null? rest) (pair? name))
306 ;; presume this is a simple test, i.e. (expect-fail (even? 2))
307 ;; where the body should also be the name.
308 `(run-test ',name #f (lambda () ,name))
309 `(run-test ,name #f (lambda () ,@rest))))
310
311 ;;; A helper function to implement the macros that test for exceptions.
312 (define (run-test-exception name exception expect-pass thunk)
313 (run-test name expect-pass
314 (lambda ()
315 (stack-catch (car exception)
316 (lambda () (thunk) #f)
317 (lambda (key proc message . rest)
318 (cond
319 ;; handle explicit key
320 ((string-match (cdr exception) message)
321 #t)
322 ;; handle `(error ...)' which uses `misc-error' for key and doesn't
323 ;; yet format the message and args (we have to do it here).
324 ((and (eq? 'misc-error (car exception))
325 (list? rest)
326 (string-match (cdr exception)
327 (apply simple-format #f message (car rest))))
328 #t)
329 ;; handle syntax errors which use `syntax-error' for key and don't
330 ;; yet format the message and args (we have to do it here).
331 ((and (eq? 'syntax-error (car exception))
332 (list? rest)
333 (string-match (cdr exception)
334 (apply simple-format #f message (car rest))))
335 #t)
336 ;; unhandled; throw again
337 (else
338 (apply throw key proc message rest))))))))
339
340 ;;; A short form for tests that expect a certain exception to be thrown.
341 (defmacro pass-if-exception (name exception body . rest)
342 `(,run-test-exception ,name ,exception #t (lambda () ,body ,@rest)))
343
344 ;;; A short form for tests expected to fail to throw a certain exception.
345 (defmacro expect-fail-exception (name exception body . rest)
346 `(,run-test-exception ,name ,exception #f (lambda () ,body ,@rest)))
347
348 \f
349 ;;;; TEST NAMES
350 ;;;;
351
352 ;;;; Turn a test name into a nice human-readable string.
353 (define (format-test-name name)
354 (call-with-output-string
355 (lambda (port)
356 (let loop ((name name)
357 (separator ""))
358 (if (pair? name)
359 (begin
360 (display separator port)
361 (display (car name) port)
362 (loop (cdr name) ": ")))))))
363
364 ;;;; For a given test-name, deliver the full name including all prefixes.
365 (define (full-name name)
366 (append (current-test-prefix) (list name)))
367
368 ;;; A fluid containing the current test prefix, as a list.
369 (define prefix-fluid (make-fluid))
370 (fluid-set! prefix-fluid '())
371 (define (current-test-prefix)
372 (fluid-ref prefix-fluid))
373
374 ;;; Postpend PREFIX to the current name prefix while evaluting THUNK.
375 ;;; The name prefix is only changed within the dynamic scope of the
376 ;;; call to with-test-prefix*. Return the value returned by THUNK.
377 (define (with-test-prefix* prefix thunk)
378 (with-fluids ((prefix-fluid
379 (append (fluid-ref prefix-fluid) (list prefix))))
380 (thunk)))
381
382 ;;; (with-test-prefix PREFIX BODY ...)
383 ;;; Postpend PREFIX to the current name prefix while evaluating BODY ...
384 ;;; The name prefix is only changed within the dynamic scope of the
385 ;;; with-test-prefix expression. Return the value returned by the last
386 ;;; BODY expression.
387 (defmacro with-test-prefix (prefix . body)
388 `(with-test-prefix* ,prefix (lambda () ,@body)))
389
390 \f
391 ;;;; REPORTERS
392 ;;;;
393
394 ;;; The global list of reporters.
395 (define reporters '())
396
397 ;;; The default reporter, to be used only if no others exist.
398 (define default-reporter #f)
399
400 ;;; Add the procedure REPORTER to the current set of reporter functions.
401 ;;; Signal an error if that reporter procedure object is already registered.
402 (define (register-reporter reporter)
403 (if (memq reporter reporters)
404 (error "register-reporter: reporter already registered: " reporter))
405 (set! reporters (cons reporter reporters)))
406
407 ;;; Remove the procedure REPORTER from the current set of reporter
408 ;;; functions. Signal an error if REPORTER is not currently registered.
409 (define (unregister-reporter reporter)
410 (if (memq reporter reporters)
411 (set! reporters (delq! reporter reporters))
412 (error "unregister-reporter: reporter not registered: " reporter)))
413
414 ;;; Return true iff REPORTER is in the current set of reporter functions.
415 (define (reporter-registered? reporter)
416 (if (memq reporter reporters) #t #f))
417
418 ;;; Send RESULT to all currently registered reporter functions.
419 (define (report . args)
420 (if (pair? reporters)
421 (for-each (lambda (reporter) (apply reporter args))
422 reporters)
423 (apply default-reporter args)))
424
425 \f
426 ;;;; Some useful standard reporters:
427 ;;;; Count reporters count the occurrence of each test result type.
428 ;;;; Log reporters write all test results to a given log file.
429 ;;;; Full reporters write all test results to the standard output.
430 ;;;; User reporters write interesting test results to the standard output.
431
432 ;;; The complete list of possible test results.
433 (define result-tags
434 '((pass "PASS" "passes: ")
435 (fail "FAIL" "failures: ")
436 (upass "UPASS" "unexpected passes: ")
437 (xfail "XFAIL" "expected failures: ")
438 (unresolved "UNRESOLVED" "unresolved test cases: ")
439 (untested "UNTESTED" "untested test cases: ")
440 (unsupported "UNSUPPORTED" "unsupported test cases: ")
441 (error "ERROR" "errors: ")))
442
443 ;;; The list of important test results.
444 (define important-result-tags
445 '(fail upass unresolved error))
446
447 ;;; Display a single test result in formatted form to the given port
448 (define (print-result port result name . args)
449 (let* ((tag (assq result result-tags))
450 (label (if tag (cadr tag) #f)))
451 (if label
452 (begin
453 (display label port)
454 (display ": " port)
455 (display (format-test-name name) port)
456 (if (pair? args)
457 (begin
458 (display " - arguments: " port)
459 (write args port)))
460 (newline port))
461 (error "(test-suite lib) FULL-REPORTER: unrecognized result: "
462 result))))
463
464 ;;; Return a list of the form (COUNTER RESULTS), where:
465 ;;; - COUNTER is a reporter procedure, and
466 ;;; - RESULTS is a procedure taking no arguments which returns the
467 ;;; results seen so far by COUNTER. The return value is an alist
468 ;;; mapping outcome symbols (`pass', `fail', etc.) onto counts.
469 (define (make-count-reporter)
470 (let ((counts (map (lambda (tag) (cons (car tag) 0)) result-tags)))
471 (list
472 (lambda (result name . args)
473 (let ((pair (assq result counts)))
474 (if pair
475 (set-cdr! pair (+ 1 (cdr pair)))
476 (error "count-reporter: unexpected test result: "
477 (cons result (cons name args))))))
478 (lambda ()
479 (append counts '())))))
480
481 ;;; Print a count reporter's results nicely. Pass this function the value
482 ;;; returned by a count reporter's RESULTS procedure.
483 (define (print-counts results . port?)
484 (let ((port (if (pair? port?)
485 (car port?)
486 (current-output-port))))
487 (newline port)
488 (display-line-port port "Totals for this test run:")
489 (for-each
490 (lambda (tag)
491 (let ((result (assq (car tag) results)))
492 (if result
493 (display-line-port port (caddr tag) (cdr result))
494 (display-line-port port
495 "Test suite bug: "
496 "no total available for `" (car tag) "'"))))
497 result-tags)
498 (newline port)))
499
500 ;;; Return a reporter procedure which prints all results to the file
501 ;;; FILE, in human-readable form. FILE may be a filename, or a port.
502 (define (make-log-reporter file)
503 (let ((port (if (output-port? file) file
504 (open-output-file file))))
505 (lambda args
506 (apply print-result port args)
507 (force-output port))))
508
509 ;;; A reporter that reports all results to the user.
510 (define (full-reporter . args)
511 (apply print-result (current-output-port) args))
512
513 ;;; A reporter procedure which shows interesting results (failures,
514 ;;; unexpected passes etc.) to the user.
515 (define (user-reporter result name . args)
516 (if (memq result important-result-tags)
517 (apply full-reporter result name args)))
518
519 (set! default-reporter full-reporter)