auto OCR pdf files in current directory

pdf_auto_ocr.sh

# File name: pdf_auto_ocr.sh
# Author: aaronedev | https://github.com/aaronedev
# Date created: 2025-07-15 10:54:00
# Date modified: 2025-07-15 10:54:22
# ------
# Copyright 2025

set -euo pipefail

# --- Configuration ---
# Colors for output
RED='\033[0;31m'
GREEN='\033[0;32m'
YELLOW='\033[1;33m'
BLUE='\033[0;34m'
NC='\033[0m' # No Color

# Default values
DEFAULT_LANG="eng"
DEFAULT_JOBS=$(nproc)
LOG_FILE="ocr_processing.log"
TEMP_DIR="/tmp/pdf_ocr_$"
FORCE_YES=false

# --- Helper Functions ---

# Function to print colored output
print_status() {
	echo -e "${BLUE}[INFO]${NC} $1"
}

print_success() {
	echo -e "${GREEN}[SUCCESS]${NC} $1"
}

print_warning() {
	echo -e "${YELLOW}[WARNING]${NC} $1"
}

print_error() {
	echo -e "${RED}[ERROR]${NC} $1"
}

# Function to show help message
show_help() {
	echo "PDF OCR Script v2.0"
	echo
	echo "Usage: $0 -l <LANG> [options]"
	echo
	echo "Options:"
	echo "  -l, --lang <LANG>    (Required) Specify the language for OCR (e.g., 'eng', 'deu')."
	echo "  -j, --jobs <NUM>     Set the maximum number of parallel jobs. Defaults to the number of CPU cores ($DEFAULT_JOBS)."
	echo "  -y, --yes            Assume 'yes' to all prompts and run non-interactively."
	echo "  -h, --help           Show this help message."
	echo "  -v, --version        Show version information."
	echo
	echo "This script automatically:"
	echo "  1. Finds all PDF files in the current directory (excluding existing '*_OCR.pdf' files)."
	echo "  2. Performs OCR with the specified language."
	echo "  3. Runs jobs in parallel for efficiency."
	echo "  4. Outputs files with an '_OCR' suffix."
	echo
	echo "Requirements:"
	echo "  - ocrmypdf, tesseract, poppler (pdfinfo, pdftotext)"
	echo
	echo "Install on Arch Linux:"
	echo "  sudo pacman -S ocrmypdf tesseract-data-eng poppler"
}

# Function to check dependencies
check_dependencies() {
	local missing_deps=()
	command -v ocrmypdf >/dev/null 2>&1 || missing_deps+=("ocrmypdf")
	command -v tesseract >/dev/null 2>&1 || missing_deps+=("tesseract")
	command -v pdfinfo >/dev/null 2>&1 || missing_deps+=("poppler (pdfinfo)")
	command -v pdftotext >/dev/null 2>&1 || missing_deps+=("poppler (pdftotext)")
	command -v nproc >/dev/null 2>&1 || missing_deps+=("nproc (coreutils)")

	if [ ${#missing_deps[@]} -ne 0 ]; then
		print_error "Missing dependencies: ${missing_deps[*]}"
		exit 1
	fi
}

# Function to check if tesseract language data is available
check_tesseract_lang() {
	local lang_code="$1"
	if ! tesseract --list-langs 2>/dev/null | grep -q "^$lang_code$"; then
		print_error "Tesseract language data for '$lang_code' not found."
		print_error "Please install the appropriate tesseract data package (e.g., 'tesseract-data-$lang_code')."
		exit 1
	fi
}

# Function to detect if PDF already has text
has_text() {
	local pdf_file="$1"
	# Consider a PDF to have text if pdftotext can extract more than 100 bytes of non-whitespace characters.
	if [ "$(pdftotext "$pdf_file" - 2>/dev/null | tr -d '[:space:]' | wc -c)" -gt 100 ]; then
		return 0 # Has text
	else
		return 1 # No text
	fi
}

# Function to process a single PDF
# This function is designed to be run in a subshell for parallel execution
process_pdf() {
	local pdf_file="$1"
	local lang_code="$2"
	local base_name
	local output_file

	base_name=$(basename "$pdf_file" .pdf)
	output_file="${base_name}_OCR.pdf"

	# This function's output is redirected to the main log file,
	# so we use standard echo for logging progress.
	echo "Starting OCR for: $pdf_file with language '$lang_code'"

	# Perform OCR. Redirect all output to the log file.
	if ocrmypdf --language "$lang_code" --output-type pdfa --optimize 1 --jpeg-quality 85 --png-quality 85 "$pdf_file" "$output_file"; then
		local original_size
		local ocr_size
		original_size=$(du -h "$pdf_file" | cut -f1)
		ocr_size=$(du -h "$output_file" | cut -f1)
		echo "SUCCESS: OCR for $pdf_file. Size: $original_size -> $ocr_size. Output: $output_file"
		return 0
	else
		echo "ERROR: OCR failed for $pdf_file."
		return 1
	fi
}

# Function to cleanup temporary files
cleanup() {
	if [ -d "$TEMP_DIR" ]; then
		rm -rf "$TEMP_DIR"
	fi
}

# --- Main Execution ---
main() {
	local lang=""
	local max_jobs="$DEFAULT_JOBS"

	# Parse command-line arguments
	while [ $# -gt 0 ]; do
		case "$1" in
		-l | --lang)
			lang="$2"
			shift 2
			;;
		-j | --jobs)
			if ! [[ "$2" =~ ^[0-9]+$ ]] || [ "$2" -eq 0 ]; then
				print_error "Invalid number of jobs: $2. Must be a positive integer."
				exit 1
			fi
			max_jobs="$2"
			shift 2
			;;
		-y | --yes)
			FORCE_YES=true
			shift 1
			;;
		-h | --help)
			show_help
			exit 0
			;;
		-v | --version)
			show_help # Version is at the top of help
			exit 0
			;;
		*)
			print_error "Unknown option: $1"
			show_help
			exit 1
			;;
		esac
	done

	# Language parameter is mandatory
	if [ -z "$lang" ]; then
		print_error "Language not specified. Use -l <LANG>."
		show_help
		exit 1
	fi

	check_dependencies
	check_tesseract_lang "$lang"

	# Setup temp dir and cleanup trap
	mkdir -p "$TEMP_DIR"
	trap cleanup EXIT SIGINT SIGTERM

	# Initialize log file
	echo "OCR Processing Log - $(date)" >"$LOG_FILE"
	print_status "Full output is being logged to $LOG_FILE"

	local pdf_files=()
	mapfile -t pdf_files < <(find . -maxdepth 1 -name "*.pdf" -type f ! -name "*_OCR.pdf" | sort)

	if [ ${#pdf_files[@]} -eq 0 ]; then
		print_warning "No PDF files found to process."
		exit 0
	fi

	local files_to_process=()
	if [ "$FORCE_YES" = true ]; then
		files_to_process=("${pdf_files[@]}")
		print_status "Found ${#files_to_process[@]} PDF(s) to process."
	else
		print_status "Found ${#pdf_files[@]} PDF file(s):"
		for pdf in "${pdf_files[@]}"; do
			if has_text "$pdf"; then
				echo "  - $(basename "$pdf") (Warning: Already has text)"
			else
				echo "  - $(basename "$pdf")"
			fi
		done
		echo
		read -p "Process all ${#pdf_files[@]} file(s)? (Y/n): " -n 1 -r
		echo
		if [[ ! $REPLY =~ ^[Nn]$ ]]; then
			files_to_process=("${pdf_files[@]}")
		else
			print_status "Processing cancelled by user."
			exit 0
		fi
	fi

	if [ ${#files_to_process[@]} -eq 0 ]; then
		print_status "No files selected for processing."
		exit 0
	fi

	local total_start_time
	total_start_time=$(date +%s)

	export -f process_pdf
	export LOG_FILE

	# Run jobs in parallel and log their output
	printf "%s\0" "${files_to_process[@]}" |
		xargs -0 -P "$max_jobs" -I {} bash -c "process_pdf '{}' '$lang'" >>"$LOG_FILE" 2>&1 &

	local xargs_pid=$!
	print_status "Started parallel processing (PID: $xargs_pid). Monitoring progress..."

	# Simple progress indicator
	while ps -p $xargs_pid >/dev/null; do
		echo -n "."
		sleep 2
	done
	echo

	wait $xargs_pid

	local total_end_time
	local total_duration
	total_end_time=$(date +%s)
	total_duration=$((total_end_time - total_start_time))

	# Grep log file for success/error counts
	local processed
	local failed
	processed=$(grep -c "SUCCESS: OCR for" "$LOG_FILE" || true)
	failed=$(grep -c "ERROR: OCR failed for" "$LOG_FILE" || true)

	echo "========================================"
	print_success "Processing summary:"
	print_status "  - Successfully processed: $processed"
	[ "$failed" -gt 0 ] && print_warning "  - Failed files: $failed"
	print_status "  - Total time: ${total_duration}s"
	print_status "  - A detailed log is available at: $LOG_FILE"
	echo "========================================"
}

# Script entry point
if [[ "${BASH_SOURCE[0]}" == "${0}" ]]; then
	main "$@"
fi