#!/bin/bash
#<pre><b>
# Script     : titlesvsbkmarks
# Version    : 1.0
# Author     : Roland Rashleigh-Berry
# Date       : 12-May-2006
# Purpose    : Production script to display table/appendix reference numbers in
#              ALL.TITLES that can not be found in the PDF bookmarks.
# SubScripts : printpdfbookmarks
# Notes      : Must be run in the same directory where ALL.TITLES is held. If
#              -r option was used when creating the PDF bookmarks then the very
#              start of the bookmark will be a reference so the -r option must
#              also be used with this script to tell it that the reference is
#              the first word in the bookmark.
# Usage      : titlesvsbkmarks ONE.pdf TWO.pdf
#              titlesvsbkmarks BIG*.pdf
#===============================================================================
# OPTIONS:
#-opt- -------------------------------description-------------------------------
#  r   (switch) reference number at start of bookmarks
#===============================================================================
# PARAMETERS:
#-pos- -------------------------------description-------------------------------
#  1   List of final output PDFs or file pattern for these
#===============================================================================
# AMENDMENT HISTORY:
# init --date-- mod-id ----------------------description------------------------
# 
#===============================================================================
# This is public domain software. No guarantee as to suitability or accuracy is
# given or implied. User uses this code entirely at their own risk.
#===============================================================================

# Define give-usage-message-then-exit function. Edit to add any useful option
# information, to change valid values of options and to add parameter values
# ("xxxxxx" and onwards) if used. Split into multiple lines if required but
# be sure to echo every line to error output using 1>&2.
usage () {
  echo 'Usage: titlesvsbkmarks -r BIG*.pdf' 1>&2
  exit 1
}


# The following variable names and defaults have been generated for you
# to help you start. You can rename the variables to something more 
# meaningful but be sure to change all occurrences of that name in the code.
switchr=0                           # Default switch=off for reference number only comparison                  


# "case" statement for action to take on selected options
while getopts "r" param ; do
  case $param in
   "?") # bad option supplied
        usage
        ;;
   "r") # (switch) reference number only comparison
        switchr=1
        ;;
  esac
done


# shift to bring first parameter to position 1
shift $(($OPTIND - 1))

# Edit the test below for the number of expected parameters and call
# the give-usage-message-then-exit function if an incorrect number 
# supplied. Remove if parameters are optional.
if [ $# -lt 1 ] ; then
  echo "Error: no names of PDF file(s) supplied" 1>&2
  usage
fi

# position of the word in the pdf bookmarks containing the reference
pos=2
if [ $switchr -eq 1 ] ; then
  pos=1
fi

gawk '{
if (toupper(substr(keep,1,13))=="TITLES BELOW:")
  print $2
keep=$0
}' ALL.TITLES | sort > ~/trefs.tmp

# make sure file not there before we append to it
rm -f ~/prefs.tmp

# for each pdf, extract bookmarks, print the reference
# and append to a temporary file
for pdf in "$@" ; do
  printpdfbookmarks $pdf | gawk '{print $pos}' pos=$pos >> ~/prefs.tmp
done

# sort the pdf bookmark references in order
sort ~/prefs.tmp > ~/prefs2.tmp

# delete the unsorted temporary file
rm -f ~/prefs.tmp

# show title references not in PDF bookmarks
join -v 1 ~/trefs.tmp ~/prefs2.tmp

# delete the temporary files used in the join
rm -f ~/trefs.tmp ~/prefs2.tmp