iconvdata/tst-tables.sh

   1 #!/bin/sh
   2 # Copyright (C) 2000-2023 Free Software Foundation, Inc.
   3 # This file is part of the GNU C Library.
   4 #
   5
   6 # The GNU C Library is free software; you can redistribute it and/or
   7 # modify it under the terms of the GNU Lesser General Public
   8 # License as published by the Free Software Foundation; either
   9 # version 2.1 of the License, or (at your option) any later version.
  10
  11 # The GNU C Library is distributed in the hope that it will be useful,
  12 # but WITHOUT ANY WARRANTY; without even the implied warranty of
  13 # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  14 # Lesser General Public License for more details.
  15
  16 # You should have received a copy of the GNU Lesser General Public
  17 # License along with the GNU C Library; if not, see
  18 # <https://www.gnu.org/licenses/>.
  19
  20 # Checks that the iconv() implementation (in both directions) for the
  21 # stateless encodings agrees with the corresponding charmap table.
  22
  23 common_objpfx=$1
  24 objpfx=$2
  25 test_program_prefix=$3
  26
  27 status=0
  28
  29 cat <<EOF |
  30   # Single-byte and other "small" encodings come here.
  31   # Keep this list in the same order as gconv-modules.
  32   #
  33   # charset name    table name          comment
  34   ASCII             ANSI_X3.4-1968
  35   ISO646-GB         BS_4730
  36   ISO646-CA         CSA_Z243.4-1985-1
  37   ISO646-CA2        CSA_Z243.4-1985-2
  38   ISO646-DE         DIN_66003
  39   ISO646-DK         DS_2089
  40   ISO646-ES         ES
  41   ISO646-ES2        ES2
  42   ISO646-CN         GB_1988-80
  43   ISO646-IT         IT
  44   ISO646-JP         JIS_C6220-1969-RO
  45   ISO646-JP-OCR-B   JIS_C6229-1984-B
  46   ISO646-YU         JUS_I.B1.002
  47   ISO646-KR         KSC5636
  48   ISO646-HU         MSZ_7795.3
  49   ISO646-CU         NC_NC00-10
  50   ISO646-FR         NF_Z_62-010
  51   ISO646-FR1        NF_Z_62-010_1973
  52   ISO646-NO         NS_4551-1
  53   ISO646-NO2        NS_4551-2
  54   ISO646-PT         PT
  55   ISO646-PT2        PT2
  56   ISO646-SE         SEN_850200_B
  57   ISO646-SE2        SEN_850200_C
  58   ISO-8859-1
  59   ISO-8859-2
  60   ISO-8859-3
  61   ISO-8859-4
  62   ISO-8859-5
  63   ISO-8859-6
  64   ISO-8859-7
  65   ISO-8859-8
  66   ISO-8859-9
  67   ISO-8859-9E
  68   ISO-8859-10
  69   ISO-8859-11
  70   ISO-8859-13
  71   ISO-8859-14
  72   ISO-8859-15
  73   ISO-8859-16
  74   T.61-8BIT
  75   ISO_6937
  76   #ISO_6937-2        ISO-IR-90          Handling of combining marks is broken
  77   KOI-8
  78   KOI8-R
  79   LATIN-GREEK
  80   LATIN-GREEK-1
  81   HP-ROMAN8
  82   HP-ROMAN9
  83   HP-TURKISH8
  84   HP-THAI8
  85   HP-GREEK8
  86   EBCDIC-AT-DE
  87   EBCDIC-AT-DE-A
  88   EBCDIC-CA-FR
  89   EBCDIC-DK-NO
  90   EBCDIC-DK-NO-A
  91   EBCDIC-ES
  92   EBCDIC-ES-A
  93   EBCDIC-ES-S
  94   EBCDIC-FI-SE
  95   EBCDIC-FI-SE-A
  96   EBCDIC-FR
  97   EBCDIC-IS-FRISS
  98   EBCDIC-IT
  99   EBCDIC-PT
 100   EBCDIC-UK
 101   EBCDIC-US
 102   IBM037
 103   IBM038
 104   IBM256
 105   IBM273
 106   IBM274
 107   IBM275
 108   IBM277
 109   IBM278
 110   IBM280
 111   IBM281
 112   IBM284
 113   IBM285
 114   IBM290
 115   IBM297
 116   IBM420
 117   IBM423
 118   IBM424
 119   IBM437
 120   IBM500
 121   IBM850
 122   IBM851
 123   IBM852
 124   IBM855
 125   IBM856
 126   IBM857
 127   IBM858
 128   IBM860
 129   IBM861
 130   IBM862
 131   IBM863
 132   IBM864
 133   IBM865
 134   IBM866
 135   IBM866NAV
 136   IBM868
 137   IBM869
 138   IBM870
 139   IBM871
 140   IBM875
 141   IBM880
 142   IBM891
 143   IBM903
 144   IBM904
 145   IBM905
 146   IBM918
 147   IBM922
 148   IBM1004
 149   IBM1026
 150   #IBM1046                              Differs from the AIX and JDK converters
 151   IBM1047
 152   IBM1124
 153   IBM1129
 154   IBM1160
 155   IBM1161
 156   IBM1132
 157   IBM1133
 158   IBM1162
 159   IBM1163
 160   IBM1164
 161   CP1125
 162   CP1250
 163   CP1251
 164   CP1252
 165   CP1253
 166   CP1254
 167   CP1255
 168   CP1256
 169   CP1257
 170   CP1258
 171   IBM874
 172   CP737
 173   CP770
 174   CP771
 175   CP772
 176   CP773
 177   CP774
 178   CP775
 179   MACINTOSH
 180   IEC_P27-1
 181   ASMO_449
 182   ISO-IR-99         ANSI_X3.110-1983
 183   ISO-IR-139        CSN_369103
 184   CWI
 185   DEC-MCS
 186   ECMA-CYRILLIC
 187   ISO-IR-153        GOST_19768-74
 188   GREEK-CCITT
 189   GREEK7
 190   GREEK7-OLD
 191   INIS
 192   INIS-8
 193   INIS-CYRILLIC
 194   ISO_2033          ISO_2033-1983
 195   ISO_5427
 196   ISO_5427-EXT
 197   #ISO_5428                             Handling of combining marks is broken
 198   ISO_10367-BOX
 199   MAC-IS
 200   MAC-UK
 201   CP10007
 202   NATS-DANO
 203   NATS-SEFI
 204   WIN-SAMI-2        SAMI-WS2
 205   ISO-IR-197
 206   TIS-620
 207   KOI8-U
 208   #ISIRI-3342                         This charset concept is completely broken
 209   VISCII
 210   KOI8-T
 211   GEORGIAN-PS
 212   GEORGIAN-ACADEMY
 213   ISO-IR-209
 214   MAC-SAMI
 215   ARMSCII-8
 216   TCVN5712-1
 217   TSCII
 218   PT154
 219   RK1048
 220   MIK
 221   BRF
 222   MAC-CENTRALEUROPE
 223   KOI8-RU
 224   #
 225   # Multibyte encodings come here
 226   #
 227   SJIS              SHIFT_JIS
 228   CP932             WINDOWS-31J
 229   #IBM932                               This converter looks quite strange
 230   #IBM943                               This converter looks quite strange
 231   EUC-KR
 232   CP949
 233   JOHAB
 234   BIG5
 235   BIG5HKSCS         BIG5-HKSCS
 236   EUC-JP
 237   EUC-JP-MS
 238   EUC-CN            GB2312
 239   GBK
 240   EUC-TW
 241   GB18030
 242   EUC-JISX0213
 243   SHIFT_JISX0213
 244   #
 245   # Stateful encodings not testable this way
 246   #
 247   #IBM930
 248   #IBM933
 249   #IBM935
 250   #IBM937
 251   #IBM939
 252   #ISO-2022-JP
 253   #ISO-2022-JP-2
 254   #ISO-2022-JP-3
 255   #ISO-2022-KR
 256   #ISO-2022-CN
 257   #ISO-2022-CN-EXT
 258   #UTF-7
 259   #
 260 EOF
 261 while read charset charmap; do
 262   if test "$charset" = GB18030; then echo "This might take a while" 1>&2; fi
 263   case ${charset} in \#*) continue;; esac
 264   printf %s "Testing ${charset}" 1>&2
 265   if ./tst-table.sh ${common_objpfx} ${objpfx} "${test_program_prefix}" \
 266       ${charset} ${charmap} < /dev/null; then
 267     echo 1>&2
 268   else
 269     echo "failed: ./tst-table.sh ${common_objpfx} ${objpfx} ${charset} ${charmap}"
 270     echo " *** FAILED ***" 1>&2
 271     exit 1
 272   fi
 273 done
 274
 275 exit $?