Update README version history for v0.9rc3

- Improved documentation with optimization levels
- Updated version history
- Fixed aspect ratio in PDF output

.gitignore

@@ -10,3 +10,4 @@ wheels/
 .venv
 
 *.pdf
+*.png

LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2024 Kessinen
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

README.md

@@ -0,0 +1,127 @@
+# PDF Musical Score Cleaner
+
+A command-line tool for processing and cleaning scanned musical score PDFs. This tool helps you extract, deskew, optimize, and recompile PDF files while maintaining high quality and readability of musical notation.
+
+## Features
+
+- **Page Extraction**: Extract individual pages from PDF files
+- **Deskewing**: Automatically correct page rotation using staff line detection
+- **White Space Trimming**: Remove excess white space around the musical content
+- **PNG Optimization**: Optimize PNG files using optipng (if installed)
+- **Modular Processing**: Process your files step by step or all at once
+- **High Quality Output**: Preserve image quality throughout the process
+- **Professional PDF Output**: Generate A4-sized PDFs with proper borders and centered content
+
+## Installation
+
+1. Ensure you have Python 3.12+ installed
+2. Install uv (recommended) or pip
+3. Clone this repository:
+   ```bash
+   git clone <repository-url>
+   cd notes_cleaner
+   ```
+4. Install dependencies:
+   ```bash
+   uv sync
+   ```
+
+5. (Optional) Install optipng for additional PNG optimization:
+   ```bash
+   # Ubuntu/Debian
+   sudo apt-get install optipng
+   
+   # macOS
+   brew install optipng
+
+   # Arch Linux
+   sudo pacman -Sy optipng
+   ```
+
+## Usage
+
+The tool provides several commands that can be run independently:
+
+### Extract Pages
+```bash
+./pdf_cleaner.py extract input.pdf
+```
+Extracts all pages from the input PDF to a temporary directory.
+
+### Deskew Pages
+```bash
+./pdf_cleaner.py deskew
+```
+Automatically detects and corrects page rotation by analyzing staff lines.
+
+### Optimize Pages
+```bash
+./pdf_cleaner.py optimize [--level {1,2,3}]
+```
+Processes pages with different optimization levels:
+- Level 1: Only trims excess white space
+- Level 2: Trims white space and converts to 1-bit monochrome
+- Level 3: All optimizations + PNG optimization (requires optipng)
+
+Default level is 1 if not specified.
+
+### Create Final PDF
+```bash
+./pdf_cleaner.py finalize output.pdf
+```
+Combines all processed pages into a final PDF with proper A4 sizing, borders, and centered content.
+
+### Typical Workflow
+```bash
+./pdf_cleaner.py extract input.pdf  # Extract pages
+./pdf_cleaner.py deskew            # Correct rotation
+./pdf_cleaner.py optimize          # Remove white space and optimize
+./pdf_cleaner.py finalize output.pdf  # Create final PDF
+```
+
+## How It Works
+
+1. **Extraction**: Uses pdf2image to convert PDF pages to high-quality PNG images
+2. **Deskewing**: 
+   - Applies morphological operations to enhance horizontal lines
+   - Uses Hough transform to detect staff lines
+   - Calculates and corrects rotation based on detected lines
+3. **Optimization**:
+   - Detects content boundaries and removes excess white space
+   - Optionally runs optipng for additional file size reduction
+4. **Finalization**: 
+   - Combines processed images into a professional A4-sized PDF
+   - Adds configurable borders around content
+   - Centers content on each page while maintaining aspect ratio
+
+## Dependencies
+
+- click: Command line interface
+- opencv-python: Image processing and deskewing
+- numpy: Numerical operations
+- pdf2image: PDF to image conversion
+- img2pdf: Image to PDF conversion
+- optipng (optional): PNG file optimization
+
+## Contributing
+
+Contributions are welcome! Please feel free to submit a Pull Request.
+
+## License
+
+This project is licensed under the MIT License - see the [LICENSE](LICENSE) file for details.
+
+## Version History
+
+- v0.9rc3: Third release candidate
+  - Switched to parallel processing in optimization
+  - Removed unused imports and variables
+  - Added `ruff` as a development dependency
+- v0.9rc2: Second release candidate
+  - Improved documentation
+  - Added optimization level descriptions
+  - Fixed aspect ratio in PDF output
+- v0.9rc1: First release candidate with full functionality
+  - Professional PDF output with A4 sizing and borders
+  - Complete image processing pipeline
+  - Configurable settings

pdf_cleaner.py

@@ -1,157 +1,414 @@
-#!/usr/bin/env python3
+#!/usr/bin/env -S uv run
+
 import os
+import subprocess
+import click
 import cv2
+import img2pdf
 import numpy as np
 from pdf2image import convert_from_path
-import img2pdf
+
+import settings
 from PIL import Image
-import argparse
+import concurrent.futures
 
-def deskew(image):
+class Settings:
-    """Deskew the image using contour detection and rotation."""
+    TRIM_PADDING_PIXELS = 20
-    # Create a copy for processing while keeping original quality
+    MONOCHROME_THRESHOLD = 127
-    proc_image = image.copy()
+    OPTIPNG_OPTIMIZATION_LEVEL = 7
+    PDF_BORDER_SIZE = 50
 
-    # Convert to grayscale and blur
+def ensure_temp_dir():
-    gray = cv2.cvtColor(proc_image, cv2.COLOR_BGR2GRAY)
+    """Ensure temporary directory exists and return its path."""
-    blur = cv2.GaussianBlur(gray, (9, 9), 0)
-    
-    # Threshold the image
-    thresh = cv2.threshold(blur, 0, 255, cv2.THRESH_BINARY_INV + cv2.THRESH_OTSU)[1]
-    
-    # Find all contours
-    contours, _ = cv2.findContours(thresh, cv2.RETR_LIST, cv2.CHAIN_APPROX_SIMPLE)
-    
-    if not contours:
-        return image
-    
-    # Find largest contour
-    contour = max(contours, key=cv2.contourArea)
-    
-    # Find minimum area rectangle
-    rect = cv2.minAreaRect(contour)
-    angle = rect[-1]
-    
-    # Adjust angle to be between -45 and 45 degrees
-    while angle < -45:
-        angle += 90
-    while angle > 45:
-        angle -= 90
-    
-    # Only rotate if the angle is significant enough
-    if abs(angle) < 0.5:  # Skip tiny rotations
-        return image
-        
-    # Rotate the image
-    (h, w) = image.shape[:2]
-    center = (w // 2, h // 2)
-    M = cv2.getRotationMatrix2D(center, angle, 1.0)
-    rotated = cv2.warpAffine(image, M, (w, h), 
-                            flags=cv2.INTER_CUBIC,
-                            borderMode=cv2.BORDER_REPLICATE)
-    
-    return rotated
-
-def adjust_levels(image):
-    """Adjust image levels for better contrast and clarity."""
-    # Convert to grayscale if not already
-    if len(image.shape) == 3:
-        gray = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)
-    else:
-        gray = image
-    
-    # Apply bilateral filter to preserve edges while reducing noise
-    denoised = cv2.bilateralFilter(gray, 9, 75, 75)
-    
-    # Create a background mask using morphological operations
-    kernel = cv2.getStructuringElement(cv2.MORPH_RECT, (15, 15))
-    background = cv2.morphologyEx(denoised, cv2.MORPH_DILATE, kernel)
-    
-    # Subtract background to normalize lighting
-    normalized = cv2.subtract(background, denoised)
-    
-    # Apply Gaussian blur to reduce noise while preserving edges
-    blurred = cv2.GaussianBlur(normalized, (3, 3), 0)
-    
-    # Use Otsu's thresholding for optimal binary threshold
-    _, binary = cv2.threshold(blurred, 0, 255, cv2.THRESH_BINARY + cv2.THRESH_OTSU)
-    
-    # Always ensure background is white and text is black
-    if cv2.countNonZero(binary) < binary.size / 2:
-        binary = cv2.bitwise_not(binary)
-    
-    return binary
-
-def process_pdf(input_path, output_path):
-    """Process a PDF file and save the cleaned version."""
-    print(f"Processing {input_path}...")
-    
-    # Convert PDF to images with high DPI to ensure minimum width of 2048px
-    pages = convert_from_path(input_path, dpi=300)
-    
-    # Create temporary directory for processed images
     temp_dir = "temp_processed_images"
     os.makedirs(temp_dir, exist_ok=True)
+    return temp_dir
 
-    # Process each page
+def get_temp_files():
-    temp_image_paths = []
+    """Get list of temporary PNG files in order."""
+    temp_dir = ensure_temp_dir()
+    files = [f for f in os.listdir(temp_dir) if f.endswith('.png')]
+    files.sort()  # Ensure correct page order
+    return [os.path.join(temp_dir, f) for f in files]
+
+@click.group()
+def cli():
+    """PDF cleaning toolbox for musical scores."""
+    pass
+
+@cli.command()
+@click.argument('input_pdf', type=click.Path(exists=True))
+def extract(input_pdf):
+    """Extract pages from PDF to temporary directory."""
+    temp_dir = ensure_temp_dir()
+    
+    # Convert PDF to images
+    print(f"Extracting pages from {input_pdf}...")
+    pages = convert_from_path(input_pdf, dpi=400)
+    
+    # Save each page
     for i, page in enumerate(pages):
-        print(f"Processing page {i+1}/{len(pages)}")
+        # Convert to grayscale
+        page = page.convert('L')
         
-        # Ensure minimum width of 2048px
+        output_path = os.path.join(temp_dir, f"page_{i:03d}.png")
-        width, height = page.size
-        scale = max(1, 2048 / width)
-        if scale > 1:
-            new_width = int(width * scale)
-            new_height = int(height * scale)
-            page = page.resize((new_width, new_height), Image.Resampling.LANCZOS)
         
-        # Convert PIL Image to OpenCV format
+        # Save initial version
-        opencv_image = cv2.cvtColor(np.array(page), cv2.COLOR_RGB2BGR)
+        page.save(output_path, "PNG", optimize=False)
         
-        # Process the image
+        # Trim whitespace
-        deskewed = deskew(opencv_image)
+        trim_whitespace(output_path)
-        cleaned = adjust_levels(deskewed)
         
-        # Convert to PIL Image
+        # Reload the trimmed image
-        pil_image = Image.fromarray(cleaned)
+        page = Image.open(output_path)
         
-        # Save temporarily as 1-bit PNG
+        # Calculate new height maintaining aspect ratio
-        temp_path = os.path.join(temp_dir, f"page_{i:03d}.png")
+        width = 2048
-        pil_image.save(temp_path, "PNG", optimize=False)
+        ratio = width / page.width
-        temp_image_paths.append(temp_path)
+        height = int(page.height * ratio)
         
-    # Save processed images as PDF with high quality settings
+        # Resize using Lanczos
-    print("Saving cleaned PDF...")
+        page = page.resize((width, height), Image.Resampling.LANCZOS)
-    a4_width_mm = 210
-    a4_height_mm = 297
-    layout_fun = img2pdf.get_layout_fun((img2pdf.mm_to_pt(a4_width_mm),
-                                       img2pdf.mm_to_pt(a4_height_mm)))
         
-    with open(output_path, "wb") as f:
+        # Save final version
-        f.write(img2pdf.convert(temp_image_paths,
+        page.save(output_path, "PNG", optimize=False)
-                              layout_fun=layout_fun,
+        print(f"Saved page {i+1}/{len(pages)}")
-                              with_pdfrw=True))
     
-    # Clean up temporary files
+    print(f"Extracted {len(pages)} pages to {temp_dir}/")
-    for temp_path in temp_image_paths:
-        os.remove(temp_path)
-    os.rmdir(temp_dir)
 
-    print(f"Saved cleaned PDF to {output_path}")
+def trim_whitespace(image_path: str, is_final: bool = False) -> bool:
+    """Remove white space from around the image.
     
-def main():
+    Handles both RGB and RGBA images, treating transparent pixels as white.
-    parser = argparse.ArgumentParser(description='Clean and straighten image-based PDFs')
-    parser.add_argument('input_pdf', help='Path to the input PDF file')
-    parser.add_argument('output_pdf', help='Path for the output PDF file')
     
-    args = parser.parse_args()
+    Args:
+        image_path: Path to the image file
+        is_final: Whether this is the final operation on the image
     
-    if not os.path.exists(args.input_pdf):
+    Returns:
-        print(f"Error: Input file '{args.input_pdf}' does not exist")
+        bool: True if successful, False otherwise
+    """
+    try:
+        # Open image with PIL
+        image = Image.open(image_path)
+        
+        # If image has transparency, flatten it first
+        if image.mode == 'RGBA':
+            # Create a white background
+            background = Image.new('RGB', image.size, 'white')
+            # Paste using alpha channel as mask
+            background.paste(image, mask=image.split()[3])
+            image = background
+            
+        # Convert to grayscale
+        image = image.convert('L')
+        
+        # Threshold to make all light pixels white and everything else black
+        # This helps with finding content bounds
+        image = image.point(lambda x: 255 if x > 250 else 0)
+        
+        # Invert so content is white on black background
+        image = Image.eval(image, lambda x: 255 - x)
+        
+        # Get the bounding box of content (now white pixels)
+        bbox = image.getbbox()
+        if not bbox:
+            print(f"Warning: No content found in {image_path}")
+            return False
+        
+        # Add padding
+        padding = settings.TRIM_PADDING_PIXELS
+        width, height = image.size
+        x1, y1, x2, y2 = bbox
+        x1 = max(0, x1 - padding)
+        y1 = max(0, y1 - padding)
+        x2 = min(width, x2 + padding)
+        y2 = min(height, y2 + padding)
+        
+        # Open original image again and crop it using the bounds
+        original = Image.open(image_path)
+        cropped = original.crop((x1, y1, x2, y2))
+        
+        # Save the cropped image, optimizing only if this is the final operation
+        cropped.save(image_path, "PNG", optimize=is_final)
+        return True
+        
+    except Exception as e:
+        print(f"Error processing {image_path}: {str(e)}")
+        return False
+
+@cli.command()
+def deskew():
+    """Deskew all pages in temporary directory."""
+    temp_files = get_temp_files()
+    if not temp_files:
+        print("No pages found in temporary directory. Run 'extract' first.")
         return
     
-    process_pdf(args.input_pdf, args.output_pdf)
+    for file_path in temp_files:
+        print(f"Deskewing {os.path.basename(file_path)}...")
         
-if __name__ == "__main__":
+        # Read image
-    main()
+        image = cv2.imread(file_path)
+        
+        # Convert to grayscale
+        gray = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)
+        
+        # Apply threshold to get binary image
+        _, binary = cv2.threshold(gray, 127, 255, cv2.THRESH_BINARY)
+        
+        # Create a rectangular kernel that's wider than it is tall
+        # This helps detect horizontal lines
+        kernel_length = np.array(binary).shape[1]//80
+        horizontal_kernel = cv2.getStructuringElement(cv2.MORPH_RECT, (kernel_length, 1))
+        
+        # Detect horizontal lines
+        horizontal_lines = cv2.erode(binary, horizontal_kernel, iterations=3)
+        horizontal_lines = cv2.dilate(horizontal_lines, horizontal_kernel, iterations=3)
+        
+        # Use probabilistic Hough transform to detect line segments
+        line_segments = cv2.HoughLinesP(
+            cv2.bitwise_not(horizontal_lines),
+            rho=1,
+            theta=np.pi/180,
+            threshold=100,
+            minLineLength=binary.shape[1]//4,  # Lines must be at least 1/4 of image width
+            maxLineGap=20
+        )
+        
+        if line_segments is not None and len(line_segments) > 0:
+            # Calculate angles of detected line segments
+            angles = []
+            for line in line_segments:
+                x1, y1, x2, y2 = line[0]
+                if x2 - x1 == 0:  # Avoid division by zero
+                    continue
+                angle = np.degrees(np.arctan2(y2 - y1, x2 - x1))
+                # Only consider angles that are close to horizontal
+                if abs(angle) < 20:
+                    angles.append(angle)
+            
+            if angles:
+                # Use median angle to avoid outliers
+                median_angle = np.median(angles)
+                
+                # Only rotate if the angle is significant but not too large
+                if 0.5 < abs(median_angle) < 20:
+                    height, width = image.shape[:2]
+                    center = (width/2, height/2)
+                    rotation_matrix = cv2.getRotationMatrix2D(center, median_angle, 1.0)
+                    rotated = cv2.warpAffine(image, rotation_matrix, (width, height),
+                                           flags=cv2.INTER_CUBIC,
+                                           borderMode=cv2.BORDER_REPLICATE)
+                    
+                    # Save rotated image
+                    cv2.imwrite(file_path, rotated)
+                    print(f"  Rotated by {median_angle:.2f} degrees")
+                else:
+                    print("  No significant rotation needed")
+            else:
+                print("  No valid horizontal lines found")
+        else:
+            print("  No line segments detected")
+
+def convert_to_monochrome(image_path: str, is_final: bool = False) -> bool:
+    """Convert image to 1-bit monochrome.
+    
+    Handles RGBA images by converting transparent pixels to white before thresholding.
+    
+    Args:
+        image_path: Path to the image file
+        is_final: Whether this is the final operation on the image
+    
+    Returns:
+        bool: True if successful, False otherwise
+    """
+    try:
+        # Open image with PIL
+        image = Image.open(image_path)
+        
+        # Convert to RGBA if not already
+        if image.mode != 'RGBA':
+            image = image.convert('RGBA')
+        
+        # Get the image data as a list of pixels
+        data = image.getdata()
+        
+        # Create new image data, replacing transparent pixels with white
+        new_data = []
+        for item in data:
+            # If pixel is transparent (alpha < 128), make it white
+            if item[3] < 128:
+                new_data.append((255, 255, 255, 255))
+            else:
+                new_data.append(item)
+        
+        # Create new image with modified data
+        image.putdata(new_data)
+        
+        # Convert to grayscale
+        image = image.convert('L')
+        
+        # Convert to 1-bit using threshold
+        image = image.point(lambda x: 255 if x > settings.MONOCHROME_THRESHOLD else 0, '1')
+        
+        # Save the monochrome image, optimizing only if this is the final operation
+        image.save(image_path, "PNG", optimize=is_final)
+        return True
+        
+    except Exception as e:
+        print(f"Error converting to monochrome {image_path}: {str(e)}")
+        return False
+
+def check_optipng_installed() -> bool:
+    """Check if optipng is installed."""
+    try:
+        result = subprocess.run(['optipng', '-v'], capture_output=True, text=True)
+        return result.returncode == 0
+    except FileNotFoundError:
+        return False
+
+def run_optipng(file_path: str) -> bool:
+    """Run optipng on a file with error handling.
+    
+    Returns:
+        bool: True if successful, False otherwise
+    """
+    try:
+        subprocess.run(
+            ['optipng', f'-o{settings.OPTIPNG_OPTIMIZATION_LEVEL}', file_path],
+            capture_output=True,
+            text=True,
+            check=True
+        )
+        return True
+    except subprocess.CalledProcessError as e:
+        print(f"Error optimizing {file_path}: {e.stderr}")
+        return False
+    except Exception as e:
+        print(f"Unexpected error optimizing {file_path}: {str(e)}")
+        return False
+
+@cli.command()
+@click.option('--level', type=click.IntRange(1, 3), default=1,
+              help='Optimization level: 1=trim, 2=monochrome, 3=full with optipng')
+def optimize(level):
+    """Optimize images with specified level of processing.
+    
+    Optimization Levels:
+        1: Only trim whitespace
+        2: Level 1 + convert to 1-bit monochrome
+        3: Level 2 + optipng optimization
+    """
+    temp_files = get_temp_files()
+    if not temp_files:
+        print("No pages found in temporary directory. Run 'extract' first.")
+        return
+    
+    opt_level = settings.OptimizationLevel(level)
+    
+    total_files = len(temp_files)
+    successful = {
+        'trim': 0,
+        'monochrome': 0,
+        'optipng': 0
+    }
+    
+    print(f"Processing {total_files} images at optimization level {level}...")
+    
+    # Step 1: Convert to monochrome if level >= 2
+    if int(opt_level) >= int(settings.OptimizationLevel.MONOCHROME):
+        print("\nConverting to monochrome...")
+        with concurrent.futures.ThreadPoolExecutor() as executor:
+            futures = {executor.submit(convert_to_monochrome, file_path, (opt_level == settings.OptimizationLevel.MONOCHROME)): file_path for file_path in temp_files}
+            for future in concurrent.futures.as_completed(futures):
+                file_path = futures[future]
+                try:
+                    if future.result():
+                        successful['monochrome'] += 1
+                        print(f"Converted {os.path.basename(file_path)}")
+                    else:
+                        print(f"Failed to convert {os.path.basename(file_path)}")
+                except Exception as exc:
+                    print(f"{os.path.basename(file_path)} generated an exception: {exc}")
+    
+    # Step 2: Always trim whitespace
+    print("\nTrimming whitespace from images...")
+    with concurrent.futures.ThreadPoolExecutor() as executor:
+        futures = {executor.submit(trim_whitespace, file_path, (opt_level == settings.OptimizationLevel.TRIM)): file_path for file_path in temp_files}
+        for future in concurrent.futures.as_completed(futures):
+            file_path = futures[future]
+            try:
+                if future.result():
+                    successful['trim'] += 1
+                    print(f"Trimmed {os.path.basename(file_path)}")
+                else:
+                    print(f"Failed to trim {os.path.basename(file_path)}")
+            except Exception as exc:
+                print(f"{os.path.basename(file_path)} generated an exception: {exc}")
+    
+    # Step 3: Run optipng if level = 3
+    if opt_level == settings.OptimizationLevel.FULL:
+        if check_optipng_installed():
+            print("\nOptimizing PNG files with optipng...")
+            with concurrent.futures.ThreadPoolExecutor() as executor:
+                futures = {executor.submit(run_optipng, file_path): file_path for file_path in temp_files}
+                for future in concurrent.futures.as_completed(futures):
+                    file_path = futures[future]
+                    try:
+                        if future.result():
+                            successful['optipng'] += 1
+                            print(f"Optimized {os.path.basename(file_path)}")
+                        else:
+                            print(f"Failed to optimize {os.path.basename(file_path)}")
+                    except Exception as exc:
+                        print(f"{os.path.basename(file_path)} generated an exception: {exc}")
+        else:
+            print("\nNote: optipng not found. Skipping PNG optimization.")
+    
+    # Print summary
+    print("\nOptimization complete!")
+    print(f"Successfully trimmed: {successful['trim']}/{total_files} images")
+    if int(opt_level) >= int(settings.OptimizationLevel.MONOCHROME):
+        print(f"Successfully converted to monochrome: {successful['monochrome']}/{total_files} images")
+    if opt_level == settings.OptimizationLevel.FULL and check_optipng_installed():
+        print(f"Successfully optimized with optipng: {successful['optipng']}/{total_files} images")
+
+@cli.command()
+@click.argument('output_pdf', type=click.Path())
+def finalize(output_pdf):
+    """Combine processed pages into final PDF and clean up."""
+    temp_dir = ensure_temp_dir()
+    temp_files = get_temp_files()
+    if not temp_files:
+        print("No pages found in temporary directory. Run 'extract' first.")
+        return
+    
+    print(f"Combining {len(temp_files)} pages into {output_pdf}...")
+    
+    # A4 size in millimeters
+    A4_WIDTH_MM = 210
+    A4_HEIGHT_MM = 297
+    
+    # Convert to PDF with border
+    with open(output_pdf, "wb") as f:
+        f.write(img2pdf.convert(
+            temp_files,
+            with_pdfrw=True,
+            layout_fun=img2pdf.get_layout_fun(
+                pagesize=(img2pdf.mm_to_pt(A4_WIDTH_MM), img2pdf.mm_to_pt(A4_HEIGHT_MM)),
+                border=(settings.PDF_BORDER_SIZE,) * 4,  # Same border size for all sides (top, right, bottom, left)
+                fit=img2pdf.FitMode.into
+            )
+        ))
+    
+    # Clean up temporary files
+    print("Cleaning up temporary files...")
+    for file_path in temp_files:
+        os.remove(file_path)
+    os.rmdir(temp_dir)
+    
+    print("PDF created successfully and temporary files removed!")
+
+if __name__ == '__main__':
+    cli()

pyproject.toml

@@ -1,9 +1,19 @@
 [project]
 name = "notes-cleaner"
-version = "0.1.0"
+version = "0.9rc3"
 description = "Add your description here"
 readme = "README.md"
+license = "MIT"
+authors = [
+    { name = "Kessinen" }
+]
 requires-python = ">=3.12"
 dependencies = [
+    "click>=8.1.7",
     "pydantic>=2.10.3",
 ]
+
+[dependency-groups]
+dev = [
+    "ruff>=0.8.1",
+]

settings.py

@@ -0,0 +1,37 @@
+"""Configuration settings for the PDF Musical Score Cleaner."""
+
+import os
+import multiprocessing
+from enum import IntEnum
+
+class OptimizationLevel(IntEnum):
+    """Optimization levels for image processing.
+    
+    Levels:
+        1: Only trim whitespace
+        2: Trim + convert to 1-bit monochrome
+        3: All optimizations + optipng
+    """
+    TRIM = 1  # Only trim whitespace
+    MONOCHROME = 2  # Trim + convert to 1-bit monochrome
+    FULL = 3  # All optimizations + optipng
+
+# Number of parallel processes to use for operations that support parallelization
+# Defaults to number of CPU cores - 1, but never less than 1
+DEFAULT_PARALLEL_PROCESSES = max(1, multiprocessing.cpu_count() - 1)
+
+# Can be overridden by environment variable
+PARALLEL_PROCESSES = int(os.getenv('NOTES_CLEANER_PARALLEL_PROCESSES', DEFAULT_PARALLEL_PROCESSES))
+
+# Border size in pixels for the final PDF output
+PDF_BORDER_SIZE = 50
+
+# Optimization settings
+OPTIPNG_OPTIMIZATION_LEVEL = int(os.getenv('NOTES_CLEANER_OPTIPNG_LEVEL', '7'))  # 0-7, higher = better compression but slower
+TRIM_PADDING_PIXELS = int(os.getenv('NOTES_CLEANER_TRIM_PADDING', '20'))  # Padding around content after trimming
+
+# Image processing settings
+MONOCHROME_THRESHOLD = int(os.getenv('NOTES_CLEANER_MONO_THRESHOLD', '200'))  # 0-255, higher = more white
+
+# Temporary directory settings
+TEMP_DIR_PREFIX = 'notes_cleaner_'

uv.lock

@@ -10,16 +10,49 @@ wheels = [
     { url = "https://files.pythonhosted.org/packages/78/b6/6307fbef88d9b5ee7421e68d78a9f162e0da4900bc5f5793f6d3d0e34fb8/annotated_types-0.7.0-py3-none-any.whl", hash = "sha256:1f02e8b43a8fbbc3f3e0d4f0f4bfc8131bcb4eebe8849b8e5c773f3a1c582a53", size = 13643 },
 ]
 
+[[package]]
+name = "click"
+version = "8.1.7"
+source = { registry = "https://pypi.org/simple" }
+dependencies = [
+    { name = "colorama", marker = "platform_system == 'Windows'" },
+]
+sdist = { url = "https://files.pythonhosted.org/packages/96/d3/f04c7bfcf5c1862a2a5b845c6b2b360488cf47af55dfa79c98f6a6bf98b5/click-8.1.7.tar.gz", hash = "sha256:ca9853ad459e787e2192211578cc907e7594e294c7ccc834310722b41b9ca6de", size = 336121 }
+wheels = [
+    { url = "https://files.pythonhosted.org/packages/00/2e/d53fa4befbf2cfa713304affc7ca780ce4fc1fd8710527771b58311a3229/click-8.1.7-py3-none-any.whl", hash = "sha256:ae74fb96c20a0277a1d615f1e4d73c8414f5a98db8b799a7931d1582f3390c28", size = 97941 },
+]
+
+[[package]]
+name = "colorama"
+version = "0.4.6"
+source = { registry = "https://pypi.org/simple" }
+sdist = { url = "https://files.pythonhosted.org/packages/d8/53/6f443c9a4a8358a93a6792e2acffb9d9d5cb0a5cfd8802644b7b1c9a02e4/colorama-0.4.6.tar.gz", hash = "sha256:08695f5cb7ed6e0531a20572697297273c47b8cae5a63ffc6d6ed5c201be6e44", size = 27697 }
+wheels = [
+    { url = "https://files.pythonhosted.org/packages/d1/d6/3965ed04c63042e047cb6a3e6ed1a63a35087b6a609aa3a15ed8ac56c221/colorama-0.4.6-py2.py3-none-any.whl", hash = "sha256:4f1d9991f5acc0ca119f9d443620b77f9d6b33703e51011c16baf57afb285fc6", size = 25335 },
+]
+
 [[package]]
 name = "notes-cleaner"
-version = "0.1.0"
+version = "0.9rc2"
 source = { virtual = "." }
 dependencies = [
+    { name = "click" },
     { name = "pydantic" },
 ]
 
+[package.dev-dependencies]
+dev = [
+    { name = "ruff" },
+]
+
 [package.metadata]
-requires-dist = [{ name = "pydantic", specifier = ">=2.10.3" }]
+requires-dist = [
+    { name = "click", specifier = ">=8.1.7" },
+    { name = "pydantic", specifier = ">=2.10.3" },
+]
+
+[package.metadata.requires-dev]
+dev = [{ name = "ruff", specifier = ">=0.8.1" }]
 
 [[package]]
 name = "pydantic"
@@ -74,6 +107,31 @@ wheels = [
     { url = "https://files.pythonhosted.org/packages/df/c3/b15fb833926d91d982fde29c0624c9f225da743c7af801dace0d4e187e71/pydantic_core-2.27.1-cp313-none-win_arm64.whl", hash = "sha256:45cf8588c066860b623cd11c4ba687f8d7175d5f7ef65f7129df8a394c502de5", size = 1882983 },
 ]
 
+[[package]]
+name = "ruff"
+version = "0.8.1"
+source = { registry = "https://pypi.org/simple" }
+sdist = { url = "https://files.pythonhosted.org/packages/95/d0/8ff5b189d125f4260f2255d143bf2fa413b69c2610c405ace7a0a8ec81ec/ruff-0.8.1.tar.gz", hash = "sha256:3583db9a6450364ed5ca3f3b4225958b24f78178908d5c4bc0f46251ccca898f", size = 3313222 }
+wheels = [
+    { url = "https://files.pythonhosted.org/packages/a2/d6/1a6314e568db88acdbb5121ed53e2c52cebf3720d3437a76f82f923bf171/ruff-0.8.1-py3-none-linux_armv6l.whl", hash = "sha256:fae0805bd514066f20309f6742f6ee7904a773eb9e6c17c45d6b1600ca65c9b5", size = 10532605 },
+    { url = "https://files.pythonhosted.org/packages/89/a8/a957a8812e31facffb6a26a30be0b5b4af000a6e30c7d43a22a5232a3398/ruff-0.8.1-py3-none-macosx_10_12_x86_64.whl", hash = "sha256:b8a4f7385c2285c30f34b200ca5511fcc865f17578383db154e098150ce0a087", size = 10278243 },
+    { url = "https://files.pythonhosted.org/packages/a8/23/9db40fa19c453fabf94f7a35c61c58f20e8200b4734a20839515a19da790/ruff-0.8.1-py3-none-macosx_11_0_arm64.whl", hash = "sha256:cd054486da0c53e41e0086e1730eb77d1f698154f910e0cd9e0d64274979a209", size = 9917739 },
+    { url = "https://files.pythonhosted.org/packages/e2/a0/6ee2d949835d5701d832fc5acd05c0bfdad5e89cfdd074a171411f5ccad5/ruff-0.8.1-py3-none-manylinux_2_17_aarch64.manylinux2014_aarch64.whl", hash = "sha256:2029b8c22da147c50ae577e621a5bfbc5d1fed75d86af53643d7a7aee1d23871", size = 10779153 },
+    { url = "https://files.pythonhosted.org/packages/7a/25/9c11dca9404ef1eb24833f780146236131a3c7941de394bc356912ef1041/ruff-0.8.1-py3-none-manylinux_2_17_armv7l.manylinux2014_armv7l.whl", hash = "sha256:2666520828dee7dfc7e47ee4ea0d928f40de72056d929a7c5292d95071d881d1", size = 10304387 },
+    { url = "https://files.pythonhosted.org/packages/c8/b9/84c323780db1b06feae603a707d82dbbd85955c8c917738571c65d7d5aff/ruff-0.8.1-py3-none-manylinux_2_17_i686.manylinux2014_i686.whl", hash = "sha256:333c57013ef8c97a53892aa56042831c372e0bb1785ab7026187b7abd0135ad5", size = 11360351 },
+    { url = "https://files.pythonhosted.org/packages/6b/e1/9d4bbb2ace7aad14ded20e4674a48cda5b902aed7a1b14e6b028067060c4/ruff-0.8.1-py3-none-manylinux_2_17_ppc64.manylinux2014_ppc64.whl", hash = "sha256:288326162804f34088ac007139488dcb43de590a5ccfec3166396530b58fb89d", size = 12022879 },
+    { url = "https://files.pythonhosted.org/packages/75/28/752ff6120c0e7f9981bc4bc275d540c7f36db1379ba9db9142f69c88db21/ruff-0.8.1-py3-none-manylinux_2_17_ppc64le.manylinux2014_ppc64le.whl", hash = "sha256:b12c39b9448632284561cbf4191aa1b005882acbc81900ffa9f9f471c8ff7e26", size = 11610354 },
+    { url = "https://files.pythonhosted.org/packages/ba/8c/967b61c2cc8ebd1df877607fbe462bc1e1220b4a30ae3352648aec8c24bd/ruff-0.8.1-py3-none-manylinux_2_17_s390x.manylinux2014_s390x.whl", hash = "sha256:364e6674450cbac8e998f7b30639040c99d81dfb5bbc6dfad69bc7a8f916b3d1", size = 12813976 },
+    { url = "https://files.pythonhosted.org/packages/7f/29/e059f945d6bd2d90213387b8c360187f2fefc989ddcee6bbf3c241329b92/ruff-0.8.1-py3-none-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:b22346f845fec132aa39cd29acb94451d030c10874408dbf776af3aaeb53284c", size = 11154564 },
+    { url = "https://files.pythonhosted.org/packages/55/47/cbd05e5a62f3fb4c072bc65c1e8fd709924cad1c7ec60a1000d1e4ee8307/ruff-0.8.1-py3-none-musllinux_1_2_aarch64.whl", hash = "sha256:b2f2f7a7e7648a2bfe6ead4e0a16745db956da0e3a231ad443d2a66a105c04fa", size = 10760604 },
+    { url = "https://files.pythonhosted.org/packages/bb/ee/4c3981c47147c72647a198a94202633130cfda0fc95cd863a553b6f65c6a/ruff-0.8.1-py3-none-musllinux_1_2_armv7l.whl", hash = "sha256:adf314fc458374c25c5c4a4a9270c3e8a6a807b1bec018cfa2813d6546215540", size = 10391071 },
+    { url = "https://files.pythonhosted.org/packages/6b/e6/083eb61300214590b188616a8ac6ae1ef5730a0974240fb4bec9c17de78b/ruff-0.8.1-py3-none-musllinux_1_2_i686.whl", hash = "sha256:a885d68342a231b5ba4d30b8c6e1b1ee3a65cf37e3d29b3c74069cdf1ee1e3c9", size = 10896657 },
+    { url = "https://files.pythonhosted.org/packages/77/bd/aacdb8285d10f1b943dbeb818968efca35459afc29f66ae3bd4596fbf954/ruff-0.8.1-py3-none-musllinux_1_2_x86_64.whl", hash = "sha256:d2c16e3508c8cc73e96aa5127d0df8913d2290098f776416a4b157657bee44c5", size = 11228362 },
+    { url = "https://files.pythonhosted.org/packages/39/72/fcb7ad41947f38b4eaa702aca0a361af0e9c2bf671d7fd964480670c297e/ruff-0.8.1-py3-none-win32.whl", hash = "sha256:93335cd7c0eaedb44882d75a7acb7df4b77cd7cd0d2255c93b28791716e81790", size = 8803476 },
+    { url = "https://files.pythonhosted.org/packages/e4/ea/cae9aeb0f4822c44651c8407baacdb2e5b4dcd7b31a84e1c5df33aa2cc20/ruff-0.8.1-py3-none-win_amd64.whl", hash = "sha256:2954cdbe8dfd8ab359d4a30cd971b589d335a44d444b6ca2cb3d1da21b75e4b6", size = 9614463 },
+    { url = "https://files.pythonhosted.org/packages/eb/76/fbb4bd23dfb48fa7758d35b744413b650a9fd2ddd93bca77e30376864414/ruff-0.8.1-py3-none-win_arm64.whl", hash = "sha256:55873cc1a473e5ac129d15eccb3c008c096b94809d693fc7053f588b67822737", size = 8959621 },
+]
+
 [[package]]
 name = "typing-extensions"
 version = "4.12.2"