unscramble pdf on gcloud/wsl

pdf_unscramble.sh

#!/bin/bash

#depends
[[ $# != 1 ]] && { echo "useage: $(basename $0) pdf_file" >&2; exit 10; }
hash pdftoppm pdfunite tesseract || sudo apt install -y poppler-utils tesseract-ocr{,-pol}

#tmp
tmp=$(mktemp -d)
trap "rm -rf $tmp" EXIT

#pdf->png
echo "Resterizing ..." >&2
pdftoppm -png "$1" "$tmp/${1%.pdf}" 2>&1
echo "OCRing ..." >&2
#png->pdf
imgs=("$tmp/${1%.pdf}"*.png)
for img in "${imgs[@]}"; do
        echo -en "Page: $((++i))/${#imgs[@]}\r" >&2
        tesseract -l pol --psm 1 --oem 1 "$img" "${img%.png}" pdf 1>/dev/null 2>&1
done
#concat pdfs
echo "Concating ..." >&2
pdfunite "$tmp/${1%.pdf}"*.pdf "${1%.pdf}".copy.pdf 2>&1