test-suite/tests/encoding-utf8.test

   1 ;;;; strings.test --- test suite for Guile's string functions    -*- mode: scheme; coding: utf-8 -*-
   2 ;;;;
   3 ;;;; Copyright (C) 2009 Free Software Foundation, Inc.
   4 ;;;;
   5 ;;;; This program is free software; you can redistribute it and/or modify
   6 ;;;; it under the terms of the GNU General Public License as published by
   7 ;;;; the Free Software Foundation; either version 2, or (at your option)
   8 ;;;; any later version.
   9 ;;;;
  10 ;;;; This program is distributed in the hope that it will be useful,
  11 ;;;; but WITHOUT ANY WARRANTY; without even the implied warranty of
  12 ;;;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  13 ;;;; GNU General Public License for more details.
  14 ;;;;
  15 ;;;; You should have received a copy of the GNU General Public License
  16 ;;;; along with this software; see the file COPYING.  If not, write to
  17 ;;;; the Free Software Foundation, Inc., 51 Franklin Street, Fifth Floor,
  18 ;;;; Boston, MA 02110-1301 USA
  19
  20 (define-module (test-strings)
  21   #:use-module (test-suite lib)
  22   #:use-module (srfi srfi-1))
  23
  24 (define exception:conversion
  25   (cons 'misc-error "^cannot convert to output locale"))
  26
  27 ;; Create a string from integer char values, eg. (string-ints 65) => "A"
  28 (define (string-ints . args)
  29   (apply string (map integer->char args)))
  30
  31 (define oldlocale #f)
  32 (if (defined? 'setlocale)
  33     (set! oldlocale (setlocale LC_ALL "")))
  34
  35 (define ascii-a (integer->char 65))     ; LATIN CAPITAL LETTER A
  36 (define a-acute (integer->char #x00c1)) ; LATIN CAPITAL LETTER A WITH ACUTE
  37 (define alpha (integer->char #x03b1))   ; GREEK SMALL LETTER ALPHA
  38 (define cherokee-a (integer->char #x13a0)) ; CHEROKEE LETTER A
  39
  40 (with-test-prefix "characters"
  41   (pass-if "input A"
  42     (char=? ascii-a #\A))
  43
  44   (pass-if "input A acute"
  45     (char=? a-acute #\Á))
  46
  47   (pass-if "input alpha"
  48     (char=? alpha #\α))
  49
  50   (pass-if "input Cherokee A"
  51     (char=? cherokee-a #\Ꭰ))
  52
  53  (pass-if "display A"
  54            (let ((pt (open-output-string)))
  55              (set-port-encoding! pt "UTF-8")
  56              (set-port-conversion-strategy! pt 'substitute)
  57              (display ascii-a pt)
  58              (string=? "A"
  59                        (get-output-string pt))))
  60
  61   (pass-if "display A acute"
  62            (let ((pt (open-output-string)))
  63              (set-port-encoding! pt "UTF-8")
  64              (set-port-conversion-strategy! pt 'substitute)
  65              (display a-acute pt)
  66              (string=? "Á"
  67                        (get-output-string pt))))
  68
  69   (pass-if "display alpha"
  70            (let ((pt (open-output-string)))
  71              (set-port-encoding! pt "UTF-8")
  72              (set-port-conversion-strategy! pt 'substitute)
  73              (display alpha pt)
  74              (string-ci=? "α"
  75                        (get-output-string pt))))
  76
  77   (pass-if "display Cherokee A"
  78            (let ((pt (open-output-string)))
  79              (set-port-encoding! pt "UTF-8")
  80              (set-port-conversion-strategy! pt 'substitute)
  81              (display cherokee-a pt)
  82              (string-ci=? "Ꭰ"
  83                        (get-output-string pt))))
  84
  85   (pass-if "write A"
  86            (let ((pt (open-output-string)))
  87              (set-port-encoding! pt "UTF-8")
  88              (set-port-conversion-strategy! pt 'escape)
  89              (write ascii-a pt)
  90              (string=? "#\\A"
  91                        (get-output-string pt))))
  92
  93   (pass-if "write A acute"
  94            (let ((pt (open-output-string)))
  95              (set-port-encoding! pt "UTF-8")
  96              (set-port-conversion-strategy! pt 'escape)
  97              (write a-acute pt)
  98              (string=? "#\\Á"
  99                        (get-output-string pt))))
 100
 101   (pass-if "write alpha"
 102            (let ((pt (open-output-string)))
 103              (set-port-encoding! pt "UTF-8")
 104              (set-port-conversion-strategy! pt 'escape)
 105              (write alpha pt)
 106              (string=? "#\\α"
 107                        (get-output-string pt))))
 108
 109   (pass-if "write Cherokee A"
 110            (let ((pt (open-output-string)))
 111              (set-port-encoding! pt "UTF-8")
 112              (set-port-conversion-strategy! pt 'escape)
 113              (write cherokee-a pt)
 114              (string=? "#\\Ꭰ"
 115                        (get-output-string pt)))))
 116
 117 (define s1 "última")
 118 (define s2 "cédula")
 119 (define s3 "años")
 120 (define s4 "羅生門")
 121
 122 (with-test-prefix "string length"
 123
 124   (pass-if "última"
 125            (eq? (string-length s1) 6))
 126
 127   (pass-if "cédula"
 128            (eq? (string-length s2) 6))
 129
 130   (pass-if "años"
 131            (eq? (string-length s3) 4))
 132
 133   (pass-if "羅生門"
 134            (eq? (string-length s4) 3)))
 135
 136 (with-test-prefix "internal encoding"
 137
 138   (pass-if "última"
 139            (string=? s1 (string-ints #xfa #x6c #x74 #x69 #x6d #x61)))
 140
 141   (pass-if "cédula"
 142            (string=? s2 (string-ints #x63 #xe9 #x64 #x75 #x6c #x61)))
 143
 144   (pass-if "años"
 145            (string=? s3 (string-ints #x61 #xf1 #x6f #x73)))
 146
 147   (pass-if "羅生門"
 148            (string=? s4 (string-ints #x7f85 #x751f #x9580))))
 149
 150 (with-test-prefix "chars"
 151
 152   (pass-if "última"
 153            (list= eqv? (string->list s1)
 154                   (list #\ú #\l #\t #\i #\m #\a)))
 155
 156   (pass-if "cédula"
 157            (list= eqv? (string->list s2)
 158                   (list #\c #\é #\d #\u #\l #\a)))
 159
 160   (pass-if "años"
 161            (list= eqv? (string->list s3)
 162                   (list #\a #\ñ #\o #\s)))
 163
 164   (pass-if "羅生門"
 165            (list= eqv? (string->list s4)
 166                   (list #\羅 #\生 #\門))))
 167
 168 (with-test-prefix "symbols == strings"
 169
 170   (pass-if "última"
 171            (eq? (string->symbol s1) 'última))
 172
 173   (pass-if "cédula"
 174            (eq? (string->symbol s2) 'cédula))
 175
 176   (pass-if "años"
 177            (eq? (string->symbol s3) 'años))
 178
 179   (pass-if "羅生門"
 180            (eq? (string->symbol s4) '羅生門)))
 181
 182 (with-test-prefix "non-ascii variable names"
 183
 184   (pass-if "1"
 185            (let ((芥川龍之介  1)
 186                  (ñ 2))
 187              (eq? (+  芥川龍之介 ñ) 3))))
 188
 189 (if (defined? 'setlocale)
 190     (setlocale LC_ALL oldlocale))