newlib/libc/string/mkunidata

   1 #! /bin/sh
   2
   3 echo Generating Unicode width data for newlib/libc/string/wcwidth.c
   4
   5 cd `dirname $0`
   6 PATH="$PATH":.  # ensure access to uniset tool
   7
   8 #############################################################################
   9 # checks and (with option -u) downloads
  10
  11 case "$1" in
  12 -h)     echo "Usage: $0 [-h|-u|-i]"
  13         echo "Generate width data tables ambiguous.t, combining.t, wide.t"
  14         echo "from local Unicode files UnicodeData.txt, Blocks.txt, EastAsianWidth.txt."
  15         echo ""
  16         echo "Options:"
  17         echo "  -u    download files from unicode.org first, download uniset tool"
  18         echo "  -i    copy files from /usr/share/unicode/ucd first"
  19         echo "  -h    show this"
  20         exit
  21         ;;
  22 -u)
  23         wget () {
  24                 ref=`basename $1`
  25                 ref=`ls "$ref" 2> /dev/null || echo 01-Jan-1970`
  26                 curl -R -O --connect-timeout 55 -z "$ref" "$1"
  27         }
  28
  29         echo downloading uniset tool
  30         wget https://www.cl.cam.ac.uk/~mgk25/download/uniset.tar.gz
  31         gzip -dc uniset.tar.gz | tar xvf - uniset
  32
  33         echo downloading data from unicode.org
  34         for data in UnicodeData.txt Blocks.txt EastAsianWidth.txt
  35         do      wget http://unicode.org/Public/UNIDATA/$data
  36         done
  37         ;;
  38 -i)
  39         echo copying data from /usr/share/unicode/ucd
  40         for data in UnicodeData.txt Blocks.txt EastAsianWidth.txt
  41         do      cp /usr/share/unicode/ucd/$data .
  42         done
  43         ;;
  44 esac
  45
  46 echo checking uniset tool
  47 type uniset || exit 9
  48
  49 echo checking Unicode data files
  50 for data in UnicodeData.txt Blocks.txt EastAsianWidth.txt
  51 do      if [ -r $data ]
  52         then    true
  53         else    echo $data not available, skipping table generation
  54                 exit
  55         fi
  56 done
  57
  58 echo generating from Unicode version `sed -e 's,[^.0-9],,g' -e 1q Blocks.txt`
  59
  60 #############################################################################
  61 # table generation
  62
  63 echo generating combining characters table
  64 uniset +cat=Me +cat=Mn +cat=Cf -00AD +1160-11FF +200B +D7B0-D7C6 +D7CB-D7FB c > combining.t
  65
  66 echo generating ambiguous width characters table
  67 sh ./mkwidthA && uniset +WIDTH-A -cat=Me -cat=Mn -cat=Cf c > ambiguous.t
  68
  69 echo generating wide characters table
  70 sh ./mkwide
  71
  72 #############################################################################
  73 # end