quick and dirty pdf ocr script

ocr

#!/bin/bash
set -e

# based on https://ryanfb.github.io/etc/2014/11/13/command_line_ocr_on_mac_os_x.html
# dependencies:
# brew install imagemagick tesseract pdftk-java
# make sure to chmod +x
# then ./ocr some/image.pdf output

source_file="$1"
dest_dir="$2"

[[ -z "$source_file" ]] && { echo "Source file required" ; exit 1; }
[[ -z "$dest_dir" ]] && { echo "Destination folder required" ; exit 1; }

source_name=$(basename $source_file)
processed_name=ocr-$source_name
tmp_dir=$dest_dir/$source_name.tmp

mkdir -p $dest_dir
mkdir -p $tmp_dir

echo Converting pdf to tiff...
convert -density 300 $source_file $tmp_dir/page_%03d.tif

pushd $tmp_dir
    echo Running OCR...
    for i in page_*.tif; do echo $i; tesseract $i $(basename $i .tif) pdf; done
    echo Merging pages...
    pdftk page_*.pdf cat output $processed_name
popd

cp $tmp_dir/$processed_name $dest_dir
rm -rf $tmp_dir

echo OCR complete: $dest_dir/$processed_name

open $dest_dir/$processed_name