Open Source Projects Europe

rclrtf    101 lines (87 with data), 2.4 kB

#!/bin/sh
# @(#$Id: rclrtf,v 1.5 2007-06-08 13:51:09 dockes Exp $  (C) 2004 J.F.Dockes
# Some inspiration from estraier
#================================================================
# convert rtf to html, by  executing the unrtf program:
#    http://www.gnu.org/software/unrtf/unrtf.html
#================================================================

# set variables
LANG=C ; export LANG
LC_ALL=C ; export LC_ALL
progname="rclrtl"
filetype=rtf

#RECFILTCOMMONCODE
##############################################################################
# !! Leave the previous line unmodified!! Code imported from the
# recfiltcommon file

# Utility code common to all shell filters. This could be sourced at run
# time, but it's slightly more efficient to include the code in the
# filters at build time (with a sed script).

# Describe error in a way that can be interpreted by our caller
senderror()
{
    echo RECFILTERROR $*
    # Also alert on stderr just in case
    echo ":2:$progname::: $*" 1>&2
    exit 1
}

iscmd()
{
    cmd=$1
    case $cmd in
    */*)
	if test -x $cmd -a ! -d $cmd ; then return 0; else return 1; fi ;;
    *)
      oldifs=$IFS; IFS=":"; set -- $PATH; IFS=$oldifs
      for d in $*;do test -x $d/$cmd -a ! -d $d/$cmd && return 0;done
      return 1 ;;
    esac
}

checkcmds()
{
    for cmd in $*;do
      if iscmd $cmd 
      then 
        a=1
      else 
        senderror HELPERNOTFOUND $cmd
      fi
    done
}

# show help message
if test $# -ne 1 -o "$1" = "--help" 
then
  echo "Convert a $filetype file to HTML text for Recoll indexing."
  echo "Usage: $progname [infile]"
  exit 1
fi

infile="$1"

# check the input file existence (may be '-' for stdin)
if test "X$infile" != X- -a ! -f "$infile"
then
  senderror INPUTNOSUCHFILE "$infile"
fi

# protect access to our temp files and directories
umask 77

##############################################################################
# !! Leave the following line unmodified !
#ENDRECFILTCOMMONCODE

checkcmds awk unrtf

# output the result
# The strange 'BEGIN' setup is to prevent 'file' from thinking this file
# is an awk program
unrtf --nopict --html "$infile" 2> /dev/null |
awk 'BEGIN'\
' {
  gothead = 0
}
/<\/head>/{
    if (gothead == 0) {
        printf("<meta http-equiv=\"Content-Type\" content=\"text/html; charset=iso-8859-1\">\n")
	gothead = 1
    }
}
{
  print
}
'