newlib/libc/ctype/mkcaseconv

   1 #! /bin/sh -f
   2
   3 # generate a table for Unicode case conversion; entries:
   4 # struct caseconv_entry defined in towctrans_l.c
   5
   6 if [ -r UnicodeData.txt ]
   7 then    UnicodeData=UnicodeData.txt
   8 elif [ -r /usr/share/unicode/ucd/UnicodeData.txt ]
   9 then    UnicodeData=/usr/share/unicode/ucd/UnicodeData.txt
  10 else    echo UnicodeData.txt not found >&2
  11         exit 1
  12 fi
  13
  14 LC_ALL=C
  15 export LC_ALL
  16
  17 compact=true
  18
  19 #0041;LATIN CAPITAL LETTER A;Lu;0;L;;;;;N;;;;0061;
  20 #0061;LATIN SMALL LETTER A;Ll;0;L;;;;;N;;;0041;;0041
  21 #0130;LATIN CAPITAL LETTER I WITH DOT ABOVE;Lu;0;L;0049 0307;;;;N;LATIN CAPITAL LETTER I DOT;;;0069;
  22 #01C4;LATIN CAPITAL LETTER DZ WITH CARON;Lu;0;L;<compat> 0044 017D;;;;N;LATIN CAPITAL LETTER D Z HACEK;;;01C6;01C5
  23 #01C5;LATIN CAPITAL LETTER D WITH SMALL LETTER Z WITH CARON;Lt;0;L;<compat> 0044 017E;;;;N;LATIN LETTER CAPITAL D SMALL Z HACEK;;01C4;01C6;01C5
  24 #01C6;LATIN SMALL LETTER DZ WITH CARON;Ll;0;L;<compat> 0064 017E;;;;N;LATIN SMALL LETTER D Z HACEK;;01C4;;01C5
  25
  26 tr -d '\015' < $UnicodeData |
  27 sed \
  28 -e 's,^\([^;]*\);[^;]*;[^;]*;[^;]*;[^;]*;[^;]*;[^;]*;[^;]*;[^;]*;[^;]*;[^;]*;[^;]*;\([^;][^;]*\);\([^;]*\);\([^;]*\)$,src \1 upper "\2" lower "\3" title "\4",' \
  29 -e t \
  30 -e 's,^\([^;]*\);[^;]*;[^;]*;[^;]*;[^;]*;[^;]*;[^;]*;[^;]*;[^;]*;[^;]*;[^;]*;[^;]*;\([^;]*\);\([^;][^;]*\);\([^;]*\)$,src \1 upper "\2" lower "\3" title "\4",' \
  31 -e t \
  32 -e 's,^\([^;]*\);[^;]*;[^;]*;[^;]*;[^;]*;[^;]*;[^;]*;[^;]*;[^;]*;[^;]*;[^;]*;[^;]*;\([^;]*\);\([^;]*\);\([^;][^;]*\)$,src \1 upper "\2" lower "\3" title "\4",' \
  33 -e t \
  34 -e d |
  35 (#src 01C5 upper "01C4" lower "01C6" title "01C5"
  36 if $compact
  37 then
  38   (
  39   cat <<\/EOS
  40   src () {
  41     if [ -n "$3" ]
  42     then        tohi=$(( 0x0$3 - 0x0$1 ))
  43     else        tohi=0
  44     fi
  45     if [ -n "$5" ]
  46     then        tolo=$(( 0x0$5 - 0x0$1 ))
  47     else        tolo=0
  48     fi
  49     case "$tolo.$tohi" in
  50     0.0)        true;;
  51     0.*)
  52         case "$1.$tohi" in
  53         *[02468ACE].1)  echo "'#error' U+$1 ODDSML";;
  54         *[02468ACE].-1) echo "  0x$1 TO1 ODDCAP";;
  55         *[13579BDF].1)  echo "'#error' U+$1 EVENSML";;
  56         *[13579BDF].-1) echo "  0x$1 TO1 EVENCAP";;
  57         *)              echo "  0x$1 TOUP $tohi";;
  58         esac;;
  59     *.0)
  60         case "$1.$tolo" in
  61         *[02468ACE].1)  echo "  0x$1 TO1 EVENCAP";;
  62         *[02468ACE].-1) echo "'#error' U+$1 EVENSML";;
  63         *[13579BDF].1)  echo "  0x$1 TO1 ODDCAP";;
  64         *[13579BDF].-1) echo "'#error' U+$1 ODDSML";;
  65         *)              echo "  0x$1 TOLO $tolo";;
  66         esac;;
  67     *)  case "$tolo.$tohi" in
  68         1.-1)           echo "  0x$1 TOBOTH 0";;
  69         *)              echo "'#error' U+$1";;
  70         esac;;
  71     esac
  72   }
  73 /EOS
  74   cat
  75   ) | sh |
  76   uniq -f1 --group=append | sed -e "s,^$,range," -e t -e "s,^,item ," |
  77   (
  78   cat <<\/EOS
  79   first=
  80   diff=-1
  81   max=255
  82   range () {
  83         # $diff == $(($last - $first))
  84         if [ "$diff" -ge 0 ]
  85         then    # we have items at all
  86                 echo "  {$first, $diff, $v2, $v3},"
  87         fi
  88         first=
  89         diff=-1
  90   }
  91   item () {
  92         if [ "$1" = "#error" ]
  93         then    echo "$*"
  94                 return
  95         fi
  96
  97         if [ $diff -eq $max ]
  98         then    range
  99         elif [ -n "$first" ]
 100         then    if [ $(( $1 )) -ne $(( ${last-0} + 1 )) ]
 101                 then    range
 102                 fi
 103         fi
 104
 105         if [ -z "$first" ]
 106         then    first=$1
 107                 v2=$2
 108                 v3=$3
 109         fi
 110
 111         last=$1
 112         diff=$(( $diff + 1 ))
 113   }
 114 /EOS
 115   cat
 116   ) | sh
 117 elif false
 118 then
 119   sed -e 's/src \([^ ]*\) upper "\([^ ]*\)" lower "\([^ ]*\)" title "\([^ ]*\)"/  {0x\1, 0x\2 - 0x\1, 0x\3 - 0x\1},/' \
 120       -e 's/0x - 0x[^ ,}]*/0/g' -e 's/0x}/0}/' \
 121       -e 's/\(0x[0-9A-F][0-9A-F]*\) - \(0x[0-9A-F][0-9A-F]*\)/$((`printf %d \1` - `printf %d \2`))/g' \
 122       -e 's/^/echo "/' -e 's/$/"/' |
 123   sh
 124 else
 125   sed -e 's/src \([^ ]*\) upper "\([^ ]*\)" lower "\([^ ]*\)" title "\([^ ]*\)"/  {0x\1, 0x\2 - 0x\1, 0x\3 - 0x\1},/' \
 126       -e 's/0x - 0x[^ ,}]*/0/g' -e 's/0x}/0}/'
 127 fi
 128 ) > caseconv.t