iconvdata/tst-tables.sh

   1 #!/bin/sh
   2 # Copyright (C) 2000-2018 Free Software Foundation, Inc.
   3 # This file is part of the GNU C Library.
   4 # Contributed by Bruno Haible <haible@clisp.cons.org>, 2000.
   5 #
   6
   7 # The GNU C Library is free software; you can redistribute it and/or
   8 # modify it under the terms of the GNU Lesser General Public
   9 # License as published by the Free Software Foundation; either
  10 # version 2.1 of the License, or (at your option) any later version.
  11
  12 # The GNU C Library is distributed in the hope that it will be useful,
  13 # but WITHOUT ANY WARRANTY; without even the implied warranty of
  14 # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  15 # Lesser General Public License for more details.
  16
  17 # You should have received a copy of the GNU Lesser General Public
  18 # License along with the GNU C Library; if not, see
  19 # <http://www.gnu.org/licenses/>.
  20
  21 # Checks that the iconv() implementation (in both directions) for the
  22 # stateless encodings agrees with the corresponding charmap table.
  23
  24 common_objpfx=$1
  25 objpfx=$2
  26 test_program_prefix=$3
  27
  28 status=0
  29
  30 cat <<EOF |
  31   # Single-byte and other "small" encodings come here.
  32   # Keep this list in the same order as gconv-modules.
  33   #
  34   # charset name    table name          comment
  35   ASCII             ANSI_X3.4-1968
  36   ISO646-GB         BS_4730
  37   ISO646-CA         CSA_Z243.4-1985-1
  38   ISO646-CA2        CSA_Z243.4-1985-2
  39   ISO646-DE         DIN_66003
  40   ISO646-DK         DS_2089
  41   ISO646-ES         ES
  42   ISO646-ES2        ES2
  43   ISO646-CN         GB_1988-80
  44   ISO646-IT         IT
  45   ISO646-JP         JIS_C6220-1969-RO
  46   ISO646-JP-OCR-B   JIS_C6229-1984-B
  47   ISO646-YU         JUS_I.B1.002
  48   ISO646-KR         KSC5636
  49   ISO646-HU         MSZ_7795.3
  50   ISO646-CU         NC_NC00-10
  51   ISO646-FR         NF_Z_62-010
  52   ISO646-FR1        NF_Z_62-010_1973
  53   ISO646-NO         NS_4551-1
  54   ISO646-NO2        NS_4551-2
  55   ISO646-PT         PT
  56   ISO646-PT2        PT2
  57   ISO646-SE         SEN_850200_B
  58   ISO646-SE2        SEN_850200_C
  59   ISO-8859-1
  60   ISO-8859-2
  61   ISO-8859-3
  62   ISO-8859-4
  63   ISO-8859-5
  64   ISO-8859-6
  65   ISO-8859-7
  66   ISO-8859-8
  67   ISO-8859-9
  68   ISO-8859-9E
  69   ISO-8859-10
  70   ISO-8859-11
  71   ISO-8859-13
  72   ISO-8859-14
  73   ISO-8859-15
  74   ISO-8859-16
  75   T.61-8BIT
  76   ISO_6937
  77   #ISO_6937-2        ISO-IR-90          Handling of combining marks is broken
  78   KOI-8
  79   KOI8-R
  80   LATIN-GREEK
  81   LATIN-GREEK-1
  82   HP-ROMAN8
  83   HP-ROMAN9
  84   HP-TURKISH8
  85   HP-THAI8
  86   HP-GREEK8
  87   EBCDIC-AT-DE
  88   EBCDIC-AT-DE-A
  89   EBCDIC-CA-FR
  90   EBCDIC-DK-NO
  91   EBCDIC-DK-NO-A
  92   EBCDIC-ES
  93   EBCDIC-ES-A
  94   EBCDIC-ES-S
  95   EBCDIC-FI-SE
  96   EBCDIC-FI-SE-A
  97   EBCDIC-FR
  98   EBCDIC-IS-FRISS
  99   EBCDIC-IT
 100   EBCDIC-PT
 101   EBCDIC-UK
 102   EBCDIC-US
 103   IBM037
 104   IBM038
 105   IBM256
 106   IBM273
 107   IBM274
 108   IBM275
 109   IBM277
 110   IBM278
 111   IBM280
 112   IBM281
 113   IBM284
 114   IBM285
 115   IBM290
 116   IBM297
 117   IBM420
 118   IBM423
 119   IBM424
 120   IBM437
 121   IBM500
 122   IBM850
 123   IBM851
 124   IBM852
 125   IBM855
 126   IBM856
 127   IBM857
 128   IBM858
 129   IBM860
 130   IBM861
 131   IBM862
 132   IBM863
 133   IBM864
 134   IBM865
 135   IBM866
 136   IBM866NAV
 137   IBM868
 138   IBM869
 139   IBM870
 140   IBM871
 141   IBM875
 142   IBM880
 143   IBM891
 144   IBM903
 145   IBM904
 146   IBM905
 147   IBM918
 148   IBM922
 149   IBM1004
 150   IBM1026
 151   #IBM1046                              Differs from the AIX and JDK converters
 152   IBM1047
 153   IBM1124
 154   IBM1129
 155   IBM1160
 156   IBM1161
 157   IBM1132
 158   IBM1133
 159   IBM1162
 160   IBM1163
 161   IBM1164
 162   CP1125
 163   CP1250
 164   CP1251
 165   CP1252
 166   CP1253
 167   CP1254
 168   CP1255
 169   CP1256
 170   CP1257
 171   CP1258
 172   IBM874
 173   CP737
 174   CP770
 175   CP771
 176   CP772
 177   CP773
 178   CP774
 179   CP775
 180   MACINTOSH
 181   IEC_P27-1
 182   ASMO_449
 183   ISO-IR-99         ANSI_X3.110-1983
 184   ISO-IR-139        CSN_369103
 185   CWI
 186   DEC-MCS
 187   ECMA-CYRILLIC
 188   ISO-IR-153        GOST_19768-74
 189   GREEK-CCITT
 190   GREEK7
 191   GREEK7-OLD
 192   INIS
 193   INIS-8
 194   INIS-CYRILLIC
 195   ISO_2033          ISO_2033-1983
 196   ISO_5427
 197   ISO_5427-EXT
 198   #ISO_5428                             Handling of combining marks is broken
 199   ISO_10367-BOX
 200   MAC-IS
 201   MAC-UK
 202   CP10007
 203   NATS-DANO
 204   NATS-SEFI
 205   WIN-SAMI-2        SAMI-WS2
 206   ISO-IR-197
 207   TIS-620
 208   KOI8-U
 209   #ISIRI-3342                         This charset concept is completely broken
 210   VISCII
 211   KOI8-T
 212   GEORGIAN-PS
 213   GEORGIAN-ACADEMY
 214   ISO-IR-209
 215   MAC-SAMI
 216   ARMSCII-8
 217   TCVN5712-1
 218   TSCII
 219   PT154
 220   RK1048
 221   MIK
 222   BRF
 223   MAC-CENTRALEUROPE
 224   KOI8-RU
 225   #
 226   # Multibyte encodings come here
 227   #
 228   SJIS              SHIFT_JIS
 229   CP932             WINDOWS-31J
 230   #IBM932                               This converter looks quite strange
 231   #IBM943                               This converter looks quite strange
 232   EUC-KR
 233   CP949
 234   JOHAB
 235   BIG5
 236   BIG5HKSCS         BIG5-HKSCS
 237   EUC-JP
 238   EUC-JP-MS
 239   EUC-CN            GB2312
 240   GBK
 241   EUC-TW
 242   GB18030
 243   EUC-JISX0213
 244   SHIFT_JISX0213
 245   #
 246   # Stateful encodings not testable this way
 247   #
 248   #IBM930
 249   #IBM933
 250   #IBM935
 251   #IBM937
 252   #IBM939
 253   #ISO-2022-JP
 254   #ISO-2022-JP-2
 255   #ISO-2022-JP-3
 256   #ISO-2022-KR
 257   #ISO-2022-CN
 258   #ISO-2022-CN-EXT
 259   #UTF-7
 260   #
 261 EOF
 262 while read charset charmap; do
 263   if test "$charset" = GB18030; then echo "This might take a while" 1>&2; fi
 264   case ${charset} in \#*) continue;; esac
 265   printf %s "Testing ${charset}" 1>&2
 266   if ./tst-table.sh ${common_objpfx} ${objpfx} "${test_program_prefix}" \
 267       ${charset} ${charmap} < /dev/null; then
 268     echo 1>&2
 269   else
 270     echo "failed: ./tst-table.sh ${common_objpfx} ${objpfx} ${charset} ${charmap}"
 271     echo " *** FAILED ***" 1>&2
 272     exit 1
 273   fi
 274 done
 275
 276 exit $?