#!/bin/bash
#
# Syntax:  pdf_extract_pages_gs [-h] range pdf_file
#
# Purpose: Use gs to extract pages from a pdf file.
#          Page ranges can be specified as X-Y, X-end, or as begin-Y
#          Output is stored in the file pdf_file_pages.pdf
#          The -h flag shows the help info for this command.
#
# See also: pdf_extract_pages (based on pdftk)

# Created: Jan  1, 2011       by Bart De Schutter
# Last revised: Jan 11, 2011  by Bart De Schutter
#
# See http://www.deschutter.info/util/scripts.html for the latest version of
# this script.

# Status: public
# Category: pdf

# 
# Define default options
#
start=""
end=""
pdf_option=""

# 
# Process options
#
while getopts :h option
do
   case "$option" in
      h) sed '/^$/q;/^#!/d;s/^# //;s/^#//' $0 >&2
         exit 1;;
     \?) echo >&2
         echo >&2 "Unknown option: $OPTARG"
         echo >&2
         sed '/^$/q;/^#!/d;s/^# //;s/^#//' $0 >&2
         exit 1;;
   esac
done
shift $(($OPTIND-1))

if [ $# -ne 2 ];
then
   echo >&2
   echo >&2 "ERROR: ${0##*/} requires exactly 2 input arguments."
   echo >&2
   sed '/^$/q;/^#!/d;s/^# //;s/^#//' $0 >&2
   exit 1
fi

range="$1"
case "$range" in
   begin-[0-9]*)
      start=""
      end=${range#begin-};;
   [0-9]*-end)
      start=${range%-end}
      end="";;
   [0-9]*-[0-9]*)
      start=${range%-[0-9]*}
      end=${range#[0-9]*-};;
   *)
      echo >&2
      echo >&2 "ERROR: Invalid page range $range for ${0##*/}."
      echo >&2
      exit 1;;
esac

pdf_option=""
if [ -n "$start" ]
then
   pdf_option="$pdf_option -dFirstPage=$start"
fi
if [ -n "$end" ]
then
   pdf_option="$pdf_option -dLastPage=$end"
fi

pdf_file="$2"
if [ ! -e "$pdf_file" ];
then
   echo >&2
   echo >&2 "ERROR: File $pdf_file not found."
   echo >&2
   sed '/^$/q;/^#!/d;s/^# //;s/^#//' $0 >&2
   exit 1
fi

gs -q -sDEVICE=pdfwrite -dNOPAUSE -dBATCH -dSAFER \
   $pdf_option \
   -dCompatibilityLevel=1.4 \
   -dEmbedAllFonts=true -dSubsetFonts=true -dMaxSubsetPct=100 \
   -dAutoFilterColorImages=false -dAutoFilterGrayImages=false \
   -dColorImageFilter=/FlateEncode -dGrayImageFilter=/FlateEncode \
   -dAutoFilterMonoImages=false -dMonoImageFilter=/CCITTFaxEncode \
   -sOutputFile="${pdf_file%.pdf}_pages.pdf" "$pdf_file"

# Extra option: -sPAPERSIZE=a4
#               -dPDFSETTINGS=/prepress