pdf-ocr.sh

#!/bin/bash

set -x

mkdir -p output

mkdir -p merge split tiff


find ./originals -type f -name '*.pdf' -exec sh -c '
  for i do
    FILE_NAME=$(basename -- "$i")
    qpdf --split-pages=50 "${i}" "split/${FILE_NAME}"
  done
' sh {} +

find ./split -type f -name '*.pdf' -exec sh -c '
  for i do
    FILE_NAME=$(basename -- "$i")
    EXTENSION="${FILE_NAME##*.}"
    BASE_NAME="${FILE_NAME%.*}"

    gs -q -dNOPAUSE -r300x300 -sDEVICE=tiff24nc -sOutputFile="tiff/${BASE_NAME}.tiff" "split/${BASE_NAME}.pdf" -c quit
    rm -f "split/${BASE_NAME}.pdf"

    tesseract "tiff/${BASE_NAME}.tiff" "merge/${BASE_NAME}" pdf
    rm -f "tiff/${BASE_NAME}.tiff"

  done
' sh {} +

find ./originals -type f -name '*.pdf' -exec sh -c '
  for i do
    FILE_NAME=$(basename -- "$i")
    BASE_NAME="${FILE_NAME%.*}"

    MERGE_LIST=$(find ./merge -type f -name "*.pdf"  -printf "\"%p\" ")

    echo "Manual Step:"
    echo "qpdf --empty "output/${BASE_NAME}.pdf" --pages ${MERGE_LIST} --"
    # qpdf --empty "output/${BASE_NAME}.pdf" --pages ${MERGE_LIST} --

  done
' sh {} +

# rm -rf merge split tiff
rm -rf split tiff

# EOF