File: numbers-string-conversion-tests.scm

package info (click to toggle)
chicken 5.3.0-2
  • links: PTS
  • area: main
  • in suites: forky, sid, trixie
  • size: 32,892 kB
  • sloc: ansic: 580,083; lisp: 71,987; tcl: 1,445; sh: 588; makefile: 60
file content (519 lines) | stat: -rw-r--r-- 16,482 bytes parent folder | download | duplicates (3)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
;;;
;;; Numerical syntax "torture test"
;;;
;;; This tries to test a lot of edge cases in Scheme's numerical syntax.
;;;
;;; Output is written so that if you run it through "grep ERROR" it will
;;; output nothing (and exit status will be nonzero) if there are no errors.
;;; If you run it through "tail -n 1" you will just get the total error summary.
;;;
;;; This code assumes that string->number accepts numbers with embedded radix
;;; specifiers (R5RS mentions that it's allowed to return #f in those cases).
;;; It also doesn't try to support Schemes which support *only* integers or
;;; *only* flonums (which is also allowed by R5RS).
;;;

;;;
;; The prelude below is messy but required to get everything working
;; with some of the major Schemes.
;;
;; Also note that to test with Gambit, it appears you first need to type in
;; (load "~~/lib/syntax-case") and then load this file, or use gsi's -:s switch
;;;

(import (chicken format)) ; Chicken w/ numbers
;(use-syntax (ice-9 syncase)) ; Guile

;; Set this to #f if the Scheme has no compnums at all, 'inexact if it only
;; supports inexact compnums or 'exact if it supports exact compnums.
;; (Gauche, Guile, SCM: inexact, Chicken w/o numbers: #f)
(define compnum-type 'exact)

;; Set this to #f if the Scheme has no fractional number support,
;; 'exact if it supports rational numbers and 'inexact if it converts fractions
;; to floating-point, inexact numbers
(define fraction-type 'exact)

;; Fix these if your Scheme doesn't allow division by zero
(define the-nan (/ 0.0 0.0))
(define pos-inf (/ 1.0 0.0))
(define neg-inf (/ -1.0 0.0))

; Scheme48, Racket, Gambit, SCM
;(define (nan? x) (and (number? x) (not (= x x))))

(define total-errors 0)

(define (check-string-against-values! str . possible-values)
  (let ((res (string->number str)))
    (let lp ((values possible-values))
      (if (null? values)
          (begin (display       "PARSE ERROR         ")
                 (write (cons str possible-values))
                 (display " => ") (write res) (newline)
                 (set! total-errors (+ total-errors 1)))
          (let ((value (car values)))
            (if (not (or (and (not (string? value)) (equal? res value))
                         (and res (nan? res) (or (and value (nan? value))))))
                (lp (cdr values))
                (let ((re-str (and res (number->string res))))
                  (let lp2 ((values possible-values))
                    (if (null? values)
                        (begin (display "SERIALIZATION ERROR ")
                               (write (cons str possible-values))
                               (display " => ") (write re-str) (newline)
                               (set! total-errors (+ total-errors 1)))
                        (let ((value (car values)))
                          (if (not (or (and res (string=? re-str str))
                                       (and (not res) (not value))
                                       (and res (string? value) (string=? re-str value))))
                              (lp2 (cdr values))
                              (begin (display "OK                  ")
                                     (write (cons str possible-values))
                                     (newline)))))))))))))

(define-syntax test-numbers
  (syntax-rules (compnums fractions)
    ((_ (compnums (types e1 ...) ...) rest ...)
     (begin
       (case compnum-type (types (test-numbers e1 ... "no-totals")) ...)
       (test-numbers rest ...)))
    ((_ (fractions (types e1 ...) ...) rest ...)
     (begin
       (case fraction-type (types (test-numbers e1 ... "no-totals")) ...)
       (test-numbers rest ...)))
    ((_ (str value ...) rest ...)
     (begin
       (check-string-against-values! str value ...)
       (test-numbers rest ...)))
    ((_ "no-totals") #f)
    ((_ x rest ...)
     (begin (newline) (display "-> ") (display x) (newline)
            (display "-----------------------------------------------------")
            (newline)
            (test-numbers rest ...)))
    ((_)
     (if (= 0 total-errors)
         (begin (newline)
                (display "-----> Everything OK, no errors!")
                (newline))
         (begin (newline)
                (display "-----> TOTAL ERRORS: ")
                (display total-errors)
                (newline)
		(exit 1))))))

(test-numbers
 "Simple integers"
 ("1" 1)
 ("+1" 1 "1")
 ("-1" (- 1))
 ("#i1" 1.0 "1.0" "1.")
 ("#I1" 1.0 "1.0" "1.")
 ("#i-1" (- 1.0) "-1.0" "-1.")
 ("123\x00456" #f) 
 ("-#i1" #f)
 ("+-1" #f)
 ("" #f)
 ("-" #f)
 ("+" #f)
 ("+-" #f)

 "Basic decimal notation"
 ("1.0" (exact->inexact 1) "1.")
 ("1." 1.0 "1.0" "1.")
 ("1.#" 1.0 1.5 "1.0" "1." "1.5")
 (".1" 0.1 "0.1" "100.0e-3")
 ("-.1" (- 0.1) "-0.1" "-100.0e-3")
 ;; Some Schemes don't allow negative zero. This is okay with the standard
 ("-.0" -0.0 "-0." "-0.0" "0.0" "0." ".0")
 ("-0." -0.0 "-.0" "-0.0" "0.0" "0." ".0")
 ("." #f)
 (".1." #f)
 ("..1" #f)
 ("1.." #f)
 ("#i1.0" 1.0 "1.0" "1.")
 ("#e1.0" 1 "1")
 ("#e-.0" 0 "0")
 ("#e-0." 0 "0")
 ("-#e.0" #f)

 "Decimal notation with padding"
 ("1#" 10.0 15.0 "10.0" "15.0" "10." "15.")
 ("#e1#" 10 15 "10" "15")
 ("#E1#" 10 15 "10" "15")
 ("#1" #f)
 ("#" #f)
 ("1#2" #f)
 ("1.#2" #f)
 (".#" #f)
 ("#.#" #f)
 ("#.1" #f)
 ("1#.2" #f)
 ("1#." 10.0 15.0 "10.0" "15.0" "10." "15.")

 "Decimal notation with suffix"
 ("1e2" 100.0 "100.0" "100.")
 ("1E2" 100.0 "100.0" "100.")
 ("1s2" 100.0 "100.0" "100.")
 ("1S2" 100.0 "100.0" "100.")
 ("1f2" 100.0 "100.0" "100.")
 ("1F2" 100.0 "100.0" "100.")
 ("1d2" 100.0 "100.0" "100.")
 ("1D2" 100.0 "100.0" "100.")
 ("1l2" 100.0 "100.0" "100.")
 ("1L2" 100.0 "100.0" "100.")
 ("1e2e3" #f)
 ("1e2s3" #f)
 ("1e2.0" #f)

 "Decimal notation with suffix and padding"
 ("1#e2" 1000.0 1500.0 "1000.0" "1500.0" "1000." "1500." "1.0e3" "15.0e2")
 ("1e2#" #f)

 "NaN, Inf"
 ("+nan.0" the-nan "+NaN.0")
 ("+NAN.0" the-nan "+nan.0" "+NaN.0")
 ("+nan.1" #f)
 ("+nan.01" #f)
 ("+inf.0" pos-inf "+Inf.0")
 ("+InF.0" pos-inf "+inf.0" "+Inf.0")
 ("-inf.0" neg-inf "-Inf.0")
 ("-iNF.0" neg-inf "-inf.0" "-Inf.0")
 ("+inf.01" #f)
 ("+inf.1" #f)
 ("-inf.01" #f)
 ("-inf.1" #f)
 ("+inf.0/1" #f)
 ("1/+inf.0" #f)
 ("+nan" #f)
 ("+inf" #f)
 ("-inf" #f)
 ("nan.0" #f)
 ("inf.0" #f)
 ;; Thanks to John Cowan for these
 ("#e+nan.0" #f)
 ("#e+inf.0" #f)
 ("#e-inf.0" #f)
 ("#i+nan.0" the-nan "+nan.0" "+NaN.0")
 ("#i+inf.0" pos-inf "+inf.0" "+Inf.0")
 ("#i-inf.0" neg-inf "-inf.0" "-Inf.0")

 "Fractions"
 (fractions
  ((exact)
   ("1/2" (/ 1 2))
   ("#e1/2" (/ 1 2) "1/2")
   ("10/2" 5 "5")
   ("-1/2" (- (/ 1 2)))
   ("10/0" #f)
   ("0/10" 0 "0")
   ("#e0/10" 0 "0")
   ("#e1#/2" 5 (/ 15 2) "5" "15/2")
   ("#e1/2#" (/ 1 20) "1/20")
   ("#i3/2" (/ 3.0 2.0) "1.5"))
  ((inexact)
   ("1/2" (/ 1 2) "0.5" ".5" "500.0e-3")
   ("0/10" 0.0 "0.0")
   ("10/2" 5.0 "5.0" "5.")
   ;; Unsure what "#e1/2" is supposed to do in Scheme w/o exact fractions
   ("#i10/2" 5.0 "5.0" "5.")
   ("-1/2" (- (/ 1 2)) "-0.5" "-.5" "-500.0e-3")))
 (fractions
  ((inexact exact)
   ("#i1/0" pos-inf "+inf.0" "+Inf.0")
   ("#i-1/0" neg-inf "-inf.0" "-Inf.0")
   ("#i0/0" the-nan "+nan.0" "+NaN.0")
   ;; This _could_ be valid in some Schemes (but isn't as pretty)
   ;("#i1/0" #f)
   ;("#i-1/0" #f)
   ;("#i0/0" #f)
   
   ("1/-2" #f)
   ("1.0/2" #f)
   ("1/2.0" #f)
   ("1/2e2" #f)
   ("1/2e2" #f)
   ("1#/2" 5.0 7.5 "5.0" "5." "7.5")
   ("1/2#" 0.05 "0.05" ".05" "50.0e-3" "5.e-002")
   ("1#/#" #f)
   ("1/" #f)
   ("1/+" #f)
   ("+/1" #f)
   ("/1" #f)
   ("/" #f)))
 
 "Basic complex numbers (rectangular notation)"
 (compnums
  ((exact)
   ("1+2i" (make-rectangular 1 2))
   ("1+2I" (make-rectangular 1 2) "1+2i")
   ("1-2i" (make-rectangular 1 -2))
   ("-1+2i" (make-rectangular -1 2))
   ("-1-2i" (make-rectangular -1 -2))
   ("+i" (make-rectangular 0 1) "+1i" "0+i" "0+1i")
   ("0+i" (make-rectangular 0 1) "+i" "+1i" "0+1i")
   ("0+1i" (make-rectangular 0 1) "+i" "+1i" "0+i")
   ("-i" (make-rectangular 0 -1) "-1i" "0-i" "0-1i")
   ("0-i" (make-rectangular 0 -1) "-i" "-1i" "0-1i")
   ("0-1i" (make-rectangular 0 -1) "-i" "-1i" "0-i")
   ("+2i" (make-rectangular 0 2) "0+2i")
   ("-2i" (make-rectangular 0 -2) "-2i" "0-2i"))
  ((inexact)
   ("1+2i" (make-rectangular 1 2) "1.0+2.0i" "1.+2.i")
   ("1+2I" (make-rectangular 1 2) "1.0+2.0i" "1.+2.i")
   ("1-2i" (make-rectangular 1 -2) "1.0-2.0i" "1.-2.i")
   ("-1+2i" (make-rectangular -1 2) "-1.0+2.0i" "-1.+2.i")
   ("-1-2i" (make-rectangular -1 -2) "-1.0-2.0i" "-1.-2.i")
   ("+i" (make-rectangular 0 1) "+1.i" "+1.0i" "0.+1.i" "0.0+1.0i")
   ("0+i" (make-rectangular 0 1) "0+1i" "+1.i" "+1.0i" "0.+1.i" "0.0+1.0i")
   ("0+1i" (make-rectangular 0 1) "+1.i" "+1.0i" "0.+1.i" "0.0+1.0i")
   ("-i" (make-rectangular 0 -1) "-1.i" "-1.0i" "0.-1.i" "0.0-1.0i")
   ("0-i" (make-rectangular 0 -1) "-1.i" "-1.0i" "0.-1.i" "0.0-1.0i")
   ("0-1i" (make-rectangular 0 -1) "-1.i" "-1.0i" "0.-1.i" "0.0-1.0i")
   ("+2i" (make-rectangular 0 2) "+2.0i" "+2.i" "0.+2.i" "0.0+2.0i")
   ("-2i" (make-rectangular 0 -2) "-2.0i" "-2.i" "0.-2.i" "0.0-2.0i")))
 (compnums
  ((exact inexact)
   ("1#+1#i" (make-rectangular 10.0 10.0) (make-rectangular 15.0 15.0)
    "10.0+10.0i" "10.+10.i" "15.0+15.0i" "15.+15.i")))
 ("2i" #f)
 ("+-i" #f)
 ("i" #f)
 ("1+2i1" #f)
 ("1+2" #f)
 ("1#+#i" #f)

 (compnums
  ((exact inexact)
   "Decimal-notation complex numbers (rectangular notation)"
   ("1.0+2i" (make-rectangular 1.0 2) "1.0+2.0i" "1.0+2i" "1.+2i" "1.+2.i")
   ("1+2.0i" (make-rectangular 1 2.0) "1.0+2.0i" "1+2.0i" "1.+2.i" "1+2.i")
   ("1#.+1#.i" (make-rectangular 10.0 10.0) (make-rectangular 15.0 15.0)
    "10.0+10.0i" "10.+10.i" "15.0+15.0i" "15.+15.i")
   ("1e2+1.0i" (make-rectangular 100.0 1.0) "100.0+1.0i" "100.+1.i")
   ("1s2+1.0i" (make-rectangular 100.0 1.0) "100.0+1.0i" "100.+1.i")
   ("1.0+1e2i" (make-rectangular 1.0 100.0) "1.0+100.0i" "1.+100.i")
   ("1.0+1s2i" (make-rectangular 1.0 100.0) "1.0+100.0i" "1.+100.i")
   ("1#e2+1.0i" (make-rectangular 1000.0 1.0) (make-rectangular 1500.0 1.0)
    "1000.0+1.0i" "1000.+1.i" "1500.0+1.0i" "1500.+1.i" "1.0e3+1.0i" "15.0e2+1.0i")
   ("1.0+1#e2i" (make-rectangular 1.0 1000.0) (make-rectangular 1.0 1500.0)
    "1.0+1000.0i" "1.+1000.i" "1.0+1500.0i" "1.+1500.i" "1.0+1.0e3i" "1.0+15.0e2i")
   (".i" #f)
   ("+.i" #f)
   (".+i" #f)))

 (compnums
  ((exact)
   "Fractional complex numbers (rectangular notation)"
   ("1/2+3/4i" (make-rectangular (/ 1 2) (/ 3 4))))
  ((inexact)
   "Fractional complex numbers (rectangular notation)"
   ("1/2+3/4i" (make-rectangular (/ 1 2) (/ 3 4)) "0.5+0.75i" ".5+.75i" "500.0e-3+750.0e-3i")))

 (compnums
  ((exact inexact)
   "Mixed fractional/decimal notation complex numbers (rectangular notation)"
   ("1#/2+3/4i" (make-rectangular 5.0 (/ 3 4)) (make-rectangular 7.5 (/ 3 4))
    "5.0+0.75i" "5.+.75i" "7.5+0.75i" "5.0+3/4i" "5.+3/4i" "7.5+3/4i" "5.0+750.0e-3i")
   ("0.5+3/4i" (make-rectangular 0.5 (/ 3 4))
    "0.5+0.75i" ".5+.75i" "0.5+3/4i" ".5+3/4i" "500.0e-3+750.0e-3i")
   ("1.5+1#/4i" (make-rectangular 1.5 2.5) (make-rectangular 1.5 3.75)
    "1.5+2.5i" "1.5+3.75i")
   ("0.5+1/#i" #f)
   ("0.5+1/1#2i" #f)
   ("1/#+0.5i" #f)
   ("1/1#2+0.5i" #f)

   "Mixed notation with infinity (might fail on mixed exactness compnums)"
   ;; This is a nasty one. Switch to inexact *after* reading the first number.
   ;; Note that it's perfectly acceptable for a scheme with *mixed* exactness
   ;; in complex values to return #f here.  TODO: How to parameterize this, we
   ;; *really* want to test that single-exactness compnums systems accept this.
   ("1/0+1.2i" (make-rectangular pos-inf 1.2) "+inf.0+1.2i" "+Inf.0+1.2i")
   ;; Less nasty, most get this right.  Same caveat as above re: mixed exactness
   ("1.2+1/0i" (make-rectangular 1.2 pos-inf) "1.2+inf.0i" "1.2+Inf.0")))

 (compnums
  ((exact inexact)
   "Complex NaN, Inf (rectangular notation)"
   ("+nan.0+nan.0i" (make-rectangular the-nan the-nan) "+NaN.0+NaN.0i")
   ("+inf.0+inf.0i" (make-rectangular pos-inf pos-inf) "+Inf.0+Inf.0i")
   ("-inf.0+inf.0i" (make-rectangular neg-inf pos-inf) "-Inf.0+Inf.0i")
   ("-inf.0-inf.0i" (make-rectangular neg-inf neg-inf) "-Inf.0-Inf.0i")
   ("+inf.0-inf.0i" (make-rectangular pos-inf neg-inf) "+Inf.0-Inf.0i")
 
   "Complex numbers (polar notation)"
   ;; TODO: Add some here. The problem is the representation
   ;;       is hard to nail down when echoed back as rectangular
   ;;       since they're floating point with many digits, and depend
   ;;       on the precision of PI used internally.
   ("1@2i" #f)
   ("0.5@1/#" #f)
   ("0.5@1/1#2" #f)
   ("1/#@0.5" #f)
   ("1/1#2@0.5" #f)
   ("1@" #f)
   ("1#@#" #f)
   ("1/@" #f)
   ("@/1" #f)
   ("@1" #f)
   ("1@+" #f)
   ("+@1" #f)
   ("@" #f)))

 "Base prefixes"
 ("#x11" 17 "17")
 ("#X11" 17 "17")
 ("#d11" 11 "11")
 ("#D11" 11 "11")
 ("#o11" 9 "9")
 ("#O11" 9 "9")
 ("#b11" 3 "3")
 ("#B11" 3 "3")
 ("#da1" #f)
 ("#o8" #f)
 ("#b2" #f)
 ("#o7" 7 "7")
 ("#xa" 10 "10")
 ("#xA" 10 "10")
 ("#xf" 15 "15")
 ("#xg" #f)
 ("#x-10" -16 "-16")
 ("#d-10" -10 "-10")
 ("#o-10" -8 "-8")
 ("#b-10" -2 "-2")
 ("-#x10" #f)
 ("-#d10" #f)
 ("-#o10" #f)
 ("-#b10" #f)
 ("#x-" #f)
 ("#x" #f)
 ("#d" #f)
 ("#d-" #f)
 ("#d+" #f)
 ("#o" #f)
 ("#o-" #f)
 ("#b" #f)
 ("#b-" #f)
 ("#e" #f)
 ("#e-" #f)
 ("#i" #f)
 ("#i-" #f)

 "Combination of prefixes"
 ("#x#x11" #f)
 ("#x#b11" #f)
 ("#b#o11" #f)
 ("#e#x10" 16 "16")
 ("#i#x10" 16.0 "16.0" "16.")
 ("#e#e10" #f)
 ("#e#e#x10" #f)
 ("#E#e#X10" #f)
 ("#i#e#x10" #f)
 ("#e#x#e10" #f)
 ("#x#x#e10" #f)
 ("#x#e#x10" #f)

 "Base prefixes with padding"
 ("#x1#0" #f)
 ("#d1#0" #f)
 ("#o1#0" #f)
 ("#b1#0" #f)
 ("#x1#" 16.0 24.0 "16.0" "24.0" "16." "24.")
 ("#d1#" 10.0 15.0 "10.0" "15.0" "10." "15.")
 ("#o1#" 8.0 12.0 "8.0" "12.0" "8." "12.")
 ("#b1#" 2.0 3.0 "2.0" "3.0" "2." "3.")

 "(Attempted) decimal notation with base prefixes"
 ("#x1.0" #f)
 ("#d1.0" 1.0 "1.0" "1.")
 ("#o1.0" #f)
 ("#b1.0" #f)
 ("#x1.#" #f)
 ("#d1.#" 1.0 1.5 "1.0" "1.5" "1.")
 ("#o1.#" #f)
 ("#b1.#" #f)
 ("#x1." #f)
 ("#d1." 1.0 "1.0" "1.")
 ("#o1." #f)
 ("#b1." #f)
 ("#x.1" #f)
 ("#d.1" 0.1 "0.1" ".1" "100.0e-3")
 ("#o.1" #f)
 ("#b.1" #f)
 ("#x1e2" 482 "482")
 ("#d1e2" 100.0 "100.0" "100.")
 ("#o1e2" #f)
 ("#b1e2" #f)

 "Fractions with prefixes"
 (fractions
  ((inexact)
   ("#x10/2" 8.0 "8.0" "8.")
   ("#x11/2" 8.5 "8.5")
   ("#d11/2" 5.5 "5.5")
   ("#o11/2" 4.5 "4.5")
   ("#b11/10" 1.5 "1.5"))
  ((exact)
   ("#x10/2" 8 "8")
   ("#x11/2" (/ 17 2) "17/2")
   ("#d11/2" (/ 11 2) "11/2")
   ("#o11/2" (/ 9 2) "9/2")
   ("#b11/10" (/ 3 2) "3/2")))
 (fractions
  ((inexact exact)
   ("#b11/2" #f)
   ("#x10/#o10" #f)
   ("10/#o10" #f)
   ("#x1#/2" 8.0 12.0 "8.0" "8." "12.0" "12.")
   ("#d1#/2" 5.0 7.5 "5.0" "5." "7.5")
   ("#o1#/2" 4.0 6.0 "4.0" "4." "6.0" "6.")
   ("#b1#/2" #f)
   ("#b1#/10" 1.0 1.5 "1.0" "1." "1.5")))

 (compnums
  ((exact inexact)
   "Complex numbers with prefixes"
   ("#x1#+1#i" (make-rectangular 16.0 16.0) (make-rectangular 24.0 24.0)
    "16.0+16.0i" "16.+16.i" "24.0+24.0i" "24.+24.i")
   ("#x1.0+1.0i" #f)
   ("#d1.0+1.0i" (make-rectangular 1.0 1.0) "1.0+1.0i" "1.+1.i")
   ("#o1.0+1.0i" #f)
   ("#b1.0+1.0i" #f)
   ("#x10+#o10i" #f)
   ("10+#o10i" #f)
   ("#x1@#x1" #f)
   ("1@#x1" #f)))
 (compnums
  ((exact)
   ("#x10+11i" (make-rectangular 16 17) "16+17i")
   ("#d10+11i" (make-rectangular 10 11) "10+11i")
   ("#o10+11i" (make-rectangular 8 9) "8+9i")
   ("#b10+11i" (make-rectangular 2 3) "2+3i")
   ("#e1.0+1.0i" (make-rectangular 1 1) "1+1i" "1+i")
   ("#i1.0+1.0i" (make-rectangular 1.0 1.0) "1.0+1.0i" "1.+1.i"))
  ((inexact)
   ("#x10+11i" (make-rectangular 16 17) "16.0+17.0i" "16.+17.i")
   ("#d10+11i" (make-rectangular 10 11) "10.0+11.0i" "10.+11.i")
   ("#o10+11i" (make-rectangular 8 9) "8.0+9.0i" "8.+9.i")
   ("#b10+11i" (make-rectangular 2 3) "2.0+3.0i" "2.+3.i")))

 )

;; #1272 - Bases not in [2,36] throw errors.
(let ((check-base (lambda (b)
                    (string->number "123" b)
                    (error "No error on invalid base" b))))
  (condition-case (check-base 1)  ((exn type) 'ok))
  (condition-case (check-base 37) ((exn type) 'ok)))

;; #1627 - Even though R7RS Scheme allows not distinguishing negative
;; zero (as in the test above), we do.
(assert (string=? "-0.0" (number->string -0.0)))
(assert (string=? "0.0" (number->string +0.0)))
(assert (eqv? -0.0 (string->number "-0.0")))
(assert (eqv? 0.0 (string->number "+0.0")))
(assert (eqv? 0.0 (string->number "0.0")))
(assert (eqv? -0.0 (string->number "-0e1")))
(assert (eqv? 0.0 (string->number "0e-1")))