src/ute-afmtodit/make-afmtodit-tables

   1 #!/bin/sh
   2 #@ make-afmtodit-tables -- script for creating the `unicode_decomposed'
   3 #@                         and `AGL_to_unicode' tables
   4 #@ Synopsis:
   5 #@   make-afmtodit-tables \
   6 #@     UnicodeData.txt version-string glyphlist.txt > afmtodit.in
   7 #@
   8 #@ `UnicodeData.txt' is the central database file from the Unicode standard.
   9 #@ Unfortunately, it doesn't contain a version number which must be thus
  10 #@ provided manually as an additional parameter.
  11 #@ `glyphlist.txt' holds the Adobe Glyph List (AGL).
  12 #@ This program needs a C preprocessor.
  13 #@ TODO remove in favour of S-CText
  14 #
  15 # Copyright (c) 2014 - 2015 Steffen (Daode) Nurpmeso <sdaoden@users.sf.net>.
  16 #
  17 # Copyright (C) 2005
  18 # Free Software Foundation, Inc.
  19 #      Written by Werner Lemberg <wl@gnu.org>
  20 #
  21 # This is free software; you can redistribute it and/or modify it under
  22 # the terms of the GNU General Public License as published by the Free
  23 # Software Foundation; either version 2, or (at your option) any later
  24 # version.
  25 #
  26 # This is distributed in the hope that it will be useful, but WITHOUT ANY
  27 # WARRANTY; without even the implied warranty of MERCHANTABILITY or
  28 # FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License
  29 # for more details.
  30 #
  31 # You should have received a copy of the GNU General Public License along
  32 # with groff; see the file COPYING.  If not, write to the Free Software
  33 # Foundation, 51 Franklin St - Fifth Floor, Boston, MA 02110-1301, USA.
  34
  35 CPP=cpp
  36
  37 prog="$0"
  38
  39 if test $# -ne 3; then
  40   echo "Synopsis: $0 UnicodeData.txt <version-string> glyphlist.txt > afmtodit.in"
  41   exit 1
  42 fi
  43
  44 unicode_data="$1"
  45 version_string="$2"
  46 glyph_list="$3"
  47
  48 if test ! -f "$1"; then
  49   echo "File \`$1' doesn't exist" >&2
  50   exit 2
  51 fi
  52 if test ! -f "$3"; then
  53   echo "File \`$3' doesn't exist" >&2
  54   exit 2
  55 fi
  56
  57 # Handle UnicodeData.txt.
  58 #
  59 # Remove ranges and control characters,
  60 # then extract the decomposition field,
  61 # then remove lines without decomposition,
  62 # then remove all compatibility decompositions.
  63 cat "$1" \
  64 | sed -e '/^[^;]*;</d' \
  65 | sed -e 's/;[^;]*;[^;]*;[^;]*;[^;]*;\([^;]*\);.*$/;\1/' \
  66 | sed -e '/^[^;]*;$/d' \
  67 | sed -e '/^[^;]*;</d' > $$1
  68
  69 # Prepare input for running cpp.
  70 cat $$1 \
  71 | sed -e 's/^\([^;]*\);/#define \1 /' \
  72       -e 's/ / u/g' > $$2
  73 cat $$1 \
  74 | sed -e 's/^\([^;]*\);.*$/\1 u\1/' >> $$2
  75
  76 # Run C preprocessor to recursively decompose.
  77 $CPP $$2 $$3
  78
  79 # Convert it back to original format.
  80 cat $$3 \
  81 | sed -e '/#/d' \
  82       -e '/^$/d' \
  83       -e 's/ \+/ /g' \
  84       -e 's/ *$//' \
  85       -e 's/u//g' \
  86       -e 's/^\([^ ]*\) /\1;/' > $$4
  87
  88 # Write comment.
  89 cat <<END
  90 # This table has been algorithmically derived from the file
  91 # UnicodeData.txt, version $version_string, available from unicode.org,
  92 # on `date '+%Y-%m-%d'`.
  93 END
  94
  95 # Emit first table.
  96 echo 'my %unicode_decomposed = ('
  97 cat $$4 \
  98 | sed -e 's/ /_/g' \
  99       -e 's/\(.*\);\(.*\)/  "\1", "\2",/'
 100 echo ');'
 101 echo ''
 102
 103 # Write comment.
 104 cat <<END
 105 # This table has been algorithmically derived from the file
 106 # glyphlist.txt, version 2.0, available from partners.adobe.com,
 107 # on `date '+%Y-%m-%d'`.
 108 END
 109
 110 # Emit second table.
 111 echo 'my %AGL_to_unicode = ('
 112 cat "$3" \
 113 | sed -e '/#/d' \
 114       -e 's/ /_/g' \
 115       -e '/;\(E\|F[0-8]\)/d' \
 116       -e 's/\(.*\);\(.*\)/  "\1", "\2",/'
 117 echo ');'
 118
 119 # Remove temporary files.
 120 rm $$1 $$2 $$3 $$4
 121
 122 # s-sh-mode