scripts/convert_to_utf_8.sh

   1 #!/bin/sh
   2 #
   3 # convert_to_utf_8.sh
   4 #
   5 # Find man pages with encoding other than us-ascii, and convert them
   6 # to the utf-8 encoding.
   7 #
   8 # Example usage:
   9 #
  10 #    cd man-pages-x.yy
  11 #    sh convert_to_utf_8.sh <output_dir> man?/*
  12 #
  13 ######################################################################
  14 #
  15 # (C) Copyright 2013, Peter Schiffer <pschiffe@redhat.com>
  16 # This program is free software; you can redistribute it and/or
  17 # modify it under the terms of the GNU General Public License
  18 # as published by the Free Software Foundation; either version 2
  19 # of the License, or (at your option) any later version.
  20 #
  21 # This program is distributed in the hope that it will be useful,
  22 # but WITHOUT ANY WARRANTY; without even the implied warranty of
  23 # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  24 # GNU General Public License for more details
  25 # (http://www.gnu.org/licenses/gpl-2.0.html).
  26 #
  27
  28 if [[ $# -lt 2 ]]; then
  29     echo "Usage: ${0} <output_dir> man?/*" 1>&2
  30     exit 1
  31 fi
  32
  33 out_dir="$1"
  34 shift
  35
  36 enc_line=""
  37
  38 for f in "$@"; do
  39     enc=$(file -bi "$f" | cut -d = -f 2)
  40     if [[ $enc != "us-ascii" ]]; then
  41         dirn=$(dirname "$f")
  42         basen=$(basename "$f")
  43         new_dir="${out_dir}/${dirn}"
  44         if [[ ! -e "$new_dir" ]]; then
  45             mkdir -p "$new_dir"
  46         fi
  47         case "$basen" in
  48             armscii-8.7 | cp1251.7 | iso_8859-*.7 | koi8-?.7)
  49
  50                 # iconv does not understand some encoding names that
  51                 # start "iso_", but does understand the corresponding
  52                 # forms that start with "iso-"
  53
  54                 from_enc="$(echo $basen | sed 's/\.7$//;s/iso_/iso-/')"
  55                 ;;
  56             *)
  57                 echo "NULL TRANSFORM: $f"
  58                 from_enc=$enc
  59                 ;;
  60         esac
  61         printf "Converting %-23s from %s\n" "$f" "$from_enc"
  62         echo "$enc_line" > "${new_dir}/${basen}"
  63         iconv -f "$from_enc" -t utf-8 "$f" \
  64             | sed "/.*-\*- coding:.*/d;/.\\\" t$/d" >> "${new_dir}/${basen}"
  65     fi
  66 done
  67
  68 exit 0