pdf_rewrite.sh

#!/bin/bash

# name:    pdf_rewrite.sh
# author:  nbehrnd@yahoo.com
# license: 2019-2024 MIT
# date:    <2019-12-17 Tue>
# edit:    <2024-01-11 Thu>

# Aiming to reduce file size of a .pdf to be sent as an attachment,
# this bash script brings finds of different places together.  After
# provision of the executable bit by chmod, the call pattern is
#
# ./pdf_rewrite [-c | --colour | --color | --reprint | -g | --gray | --grey] input.pdf
#
# to yield the output .pdf file of same name as the original input, either
# + as a reprint retaining existing color scale (either -c, -r, --color,
# --colour, or --reprint)
# + as a reprint constrained to gray scale (either -g, or --gray, or --grey)
#
# Often, the later still is 'good enough'  especially with modern .pdf
# by journal publications.  It requires an installation of ghostscript
# as by typical Linuxes e.g., Xubuntu 18.04.3 LTS (which is gs 9.26).
#
# No warranties -- To be used on your own risk.

if [[ -z "$1" ]] ; then
    printf "Error. The minimal input is: pdf_rewrite.sh -h\n"

elif [[ "$1" == "-h" ]] ; then
    printf "
To reduce the file size of a .pdf, this script relays the file to
ghostscript.  Often, the reprint's file size is smaller than the of
the input.  Typically, a pdf's searchable text layer is retained.

The typical input pattern for this script is:

    ./pdf_rewrite --reprint input.pdf

Parameters available to this script are:

    -h                                         access this help menu
    -c, --colour, --color; -r, or --reprint    reprint in color
    -g, --gray, or --grey                      reprint in gray scale

On occasion, you may loose internal crosslinks (like the table of
contents), or hyperlinks to external references may be lost applying
this.  This depends on the input file -- see the test files of this
project, or an example like https://doi.org/10.1021/acs.jchemed.7b00361.
After all, ghostscript prepares a file eventually print.\n\n"


elif [[ "$1" == "-c" || "$1" == "-r" || "$1" == "--colour" || "$1" == "--color" || "$1" == "--reprint" ]] ; then

    # 'reprint as such'
    #
    # Sometimes, this removes cross- or / and hyperlinks.  Color and the
    # searchable text layer however typically remain.  It either states
    # the savings obtained, or that a further shrinkage isn't obtained.
    # The output file will have the same name as the inpupt file.
    #
    # Except of the loop to direct the work into this branch, and addition
    # of the -dPrinted=false instruction to reatin hyperlinks, this is the
    # answer by Evan Langlois to
    #
    # https://tex.stackexchange.com/questions/18987/how-to-make-the-pdfs-produced-by-pdflatex-smaller?rq=1
    #
    # so the credit to figuring out the following belongs to him:

    file="$2"
    filebase="$(basename "$file" .pdf)"
    optfile="/tmp/$$-${filebase}_opt.pdf"
    gs -sDEVICE=pdfwrite -dCompatibilityLevel=1.4 -dNOPAUSE -dQUIET -dBATCH \
            -dPrinted=false -sOutputFile="${optfile}" "${file}"

    if [ $? == '0' ]; then
        optsize=$(stat -c "%s" "${optfile}")
        orgsize=$(stat -c "%s" "${file}")
        if [ "${optsize}" -eq 0 ]; then
            echo "No output!  Keeping original"
            rm -f "${optfile}"
            exit;
        fi
        if [ ${optsize} -ge ${orgsize} ]; then
            echo "Didn't make it smaller! Keeping original"
            rm -f "${optfile}"
            exit;
        fi
        bytesSaved=$(expr $orgsize - $optsize)
        percent=$(expr $optsize '*' 100 / $orgsize)
        echo Saving $bytesSaved bytes \(now ${percent}% of old\)
        rm "${file}"
        mv "${optfile}" "${file}"
    fi

elif [[ "$1" == "-g" ]] || [[ "$1" == "--gray" ]] || [[ "$1" == "--grey" ]] ; then

    # A reprint in grayscale (often sufficent for modern journal .pdf),
    # while retaining the searchable text layer.  The script was found
    # on Unix Stackexchange under
    #
    # https://unix.stackexchange.com/questions/93959/how-to-convert-a-color-pdf-to-black-white
    #
    # as answer by slm (answered Oct 7 '13 at 18:18).  Contrasting to
    # the the question, result is grayscale, rather than the bitonal
    # black-white.  If using on journal publications, meta data such as
    # 'Subject', 'Keywords', 'Author', 'Pages' are retained from the
    # original.  'Producer' (e.g., GPL Ghostscript), 'CreationDate',
    # 'ModDate' however obviously are subject to change during this
    # modification.
    # Contrasting to the original code, the output is written on expense
    # of the original file.  This prevents overwriting 'grayed .pdf'
    # while processing multiple files in one session.  By dPrinted=false,
    # hyperlinks remain.

    gs \
     -sOutputFile=output.pdf \
     -sDEVICE=pdfwrite \
     -sColorConversionStrategy=Gray \
     -dProcessColorModel=/DeviceGray \
     -dCompatibilityLevel=1.4 \
     -dNOPAUSE \
     -dBATCH \
     -dPrinted=false \
     $2

    # addition, name the output like the original read:
    mv output.pdf $2

#  final coda:
fi