dotfiles/.local/bin/pdfctl

#!/bin/sh

usage () {
	cat <<EOF>&2
Usage: ${0##*/} COMMAND [-i] [-o OUTPUT] [OPTIONS] FILES...

Wrapper around various PDF operations.
OUTPUT is ignored when processing in-place (-i) or when more than one input
files is given.
Otherwise, or if OUTPUT is unspecified, a file with a random suffix is produced.

List of commands and their options:

  compress [OPTIONS]
    Compress PDFs.
    -c: Compress raster graphics at the expense of a quality loss.
    -n: Do not compress raster graphics at all. Output may be bigger.

  extract [-f FIRST-PAGE] [-l LAST-PAGE]
    Extract a range of pages to a single PDF.

  resize
    Resize to A4.

  size
    Same as pdfinfo, but append size in milimeters.

  xerox
    Convert a PDF to a blurred, black & white raster image.


For more details on compression:

  $ gs -sDEVICE=pdfwrite -o /dev/null \\
    -c "currentpagedevice { exch ==only ( ) print == } forall"

  http://partners.adobe.com/public/developer/en/acrobat/sdk/pdf/pdf_creation_apis_and_specs/DistillerParameters.pdf

EOF
}

[ $# -eq 0 ] && usage && exit

cmd="$1"
shift 1

NO_OUTPUT=false
INPLACE=false

COMPRESS_OPT="-dColorConversionStrategy=/LeaveColorUnchanged -dDownsampleMonoImages=false -dDownsampleGrayImages=false -dDownsampleColorImages=false -dAutoFilterColorImages=false -dAutoFilterGrayImages=false -dColorImageFilter=/FlateEncode -dGrayImageFilter=/FlateEncode"

EXTRACT_FIRST=1
EXTRACT_LAST=1

case $cmd in
compress)
	while getopts ":cino:" opt; do
		case $opt in
		c)
			COMPRESS_OPT="" ;;
		i)
			INPLACE=true ;;
		n)
			COMPRESS_OPT="-dColorConversionStrategy=/LeaveColorUnchanged -dEncodeColorImages=false -dEncodeGrayImages=false -dEncodeMonoImages=false" ;;
		o)
			output="$OPTARG" ;;
		\?)
			usage
			exit 1 ;;
		esac
	done
	;;

extract)
	while getopts ":f:il:o:" opt; do
		case $opt in
		f)
			EXTRACT_FIRST="$OPTARG" ;;
		i)
			INPLACE=true ;;
		l)
			EXTRACT_LAST="$OPTARG" ;;
		o)
			output="$OPTARG" ;;
		\?)
			usage
			exit 1 ;;
		esac
	done
	;;

resize|xerox)
	while getopts ":io:" opt; do
		case $opt in
		i)
			INPLACE=true ;;
		o)
			output="$OPTARG" ;;
		\?)
			usage
			exit 1 ;;
		esac
	done
	;;

size)
	NO_OUTPUT=true ;;

*)
	usage
	exit 1 ;;
esac

shift $((OPTIND - 1))
if [ $# -eq 0 ]; then
	usage
	exit 1
fi
[ $# -gt 1 ] &&	unset output

mkoutput () {
	if [ -z "$output" ] || [ -e "$output" ]; then
		case "$input" in
		*/*) ;;
		*)   input="./$input" ;;
		esac
		dirname="${input%/*}"
		basename="${input##*/}"
		output=$(mktemp "$dirname/${basename%.*}-$cmd"-XXXXXX.pdf)
	fi
	echo "$output"
}

## Some JPEG options
# -sDEVICE=jpeg
# -dUseCIEColor
# -dJPEGQ=100
# -r300
# -dTextAlphaBits=4
# -dGraphicsAlphaBits=4
# -dMaxStripSize=8192
compress () {
	gs -dBATCH -dNOPAUSE -q -sDEVICE=pdfwrite $COMPRESS_OPT -sOutputFile="$output" "$input"
}

extract () {
	gs -sDEVICE=pdfwrite -dNOPAUSE -dSAFER -dBATCH -dFirstPage="$EXTRACT_FIRST" -dLastPage="$EXTRACT_LAST" -sOutputFile="$output" "$input"
}

resize () {
	gs -q -o "$output" -sDEVICE=pdfwrite -sPAPERSIZE=a4 -dFIXEDMEDIA -dPDFFitPage -dCompatibilityLevel=1.4 "$input"
}

size () {
	pdfinfo "$input" | awk '/^Page size:/ {printf $0; print ", " $3*0.35278 " x " $5*0.35278 " mm"; next } 1'
}

xerox () {
	convert -threshold 90% -blur 1x1 -quality 100 -flatten -density 400 "$input" "$output"
}

for input; do
	echo "==> '$input'"
	$NO_OUTPUT ||	output=$(mkoutput)
	$cmd
	if $INPLACE; then
		rm -f "$input"
		mv "$output" "$input"
	fi
	echo
done