#!/usr/bin/env python3 """ SUPERSEDED by src/video/assemble/pdf.ts — bug fixes land in the TS port only; this script is kept as the porting reference. See SUPERSEDED.md. Extract slides from a PDF file as images and generate slides.json. Converts each PDF page to a JPEG image using pdf2image (wraps pdftoppm). Also copies images to images/scene_N_frame.jpg for the downstream pipeline. Usage: # Basic extraction with equal duration python extract_pdf_slides.py \ --pdf "slides.pdf" \ --output-dir "projects/{slug}/slides" \ --output-json "projects/{slug}/analysis/slides.json" \ --total-duration 120 # Custom DPI python extract_pdf_slides.py \ --pdf "slides.pdf" \ --output-dir "projects/{slug}/slides" \ --output-json "projects/{slug}/analysis/slides.json" \ --dpi 300 # Dry-run (check PDF page count without converting) python extract_pdf_slides.py \ --pdf "slides.pdf" --dry-run """ import argparse import json import os import shutil import sys def convert_pdf_to_images( pdf_path: str, output_dir: str, dpi: int = 200, ) -> list[str]: """Convert PDF pages to JPEG images. Returns list of output paths.""" try: from pdf2image import convert_from_path except ImportError: print("Error: pdf2image not installed. Run: pip install pdf2image") sys.exit(1) os.makedirs(output_dir, exist_ok=True) print(f" Converting PDF at {dpi} DPI...") images = convert_from_path(pdf_path, dpi=dpi, fmt="jpeg") output_paths = [] for i, image in enumerate(images, start=1): filename = f"slide_{i:03d}.jpg" filepath = os.path.join(output_dir, filename) image.save(filepath, "JPEG", quality=95) output_paths.append(filepath) return output_paths def get_pdf_page_count(pdf_path: str) -> int: """Get page count without full conversion.""" try: from pdf2image import pdfinfo_from_path info = pdfinfo_from_path(pdf_path) return info.get("Pages", 0) except Exception: # Fallback: convert first page only to check try: from pdf2image import convert_from_path images = convert_from_path(pdf_path, first_page=1, last_page=1) return len(images) if images else 0 except Exception: return 0 def generate_slides_json( image_paths: list[str], pdf_path: str, total_duration: float | None = None, default_per_slide: float = 8.0, ) -> dict: """Generate slides.json with equal durations.""" num_slides = len(image_paths) if total_duration and total_duration > 0: per_slide_duration = total_duration / num_slides else: per_slide_duration = default_per_slide total_duration = per_slide_duration * num_slides slides = [] current_time = 0.0 for i, image_path in enumerate(image_paths, start=1): slides.append({ "slide": i, "timestamp": f"{int(current_time // 60):02d}:{current_time % 60:05.2f}", "timestamp_seconds": round(current_time, 2), "duration": round(per_slide_duration, 2), "image": os.path.basename(image_path), "status": "detected", }) current_time += per_slide_duration return { "source_type": "pdf", "source_file": os.path.abspath(pdf_path), "source_duration": round(total_duration, 2), "total_slides": num_slides, "slides": slides, "settings": {}, } def extract_single_page( pdf_path: str, page_num: int, output_dir: str, dpi: int = 200, ) -> str: """Re-extract a single page from PDF after nano-pdf edit. page_num is 1-based. Overwrites existing slide image. Returns the output file path. """ try: from pdf2image import convert_from_path except ImportError: print("Error: pdf2image not installed. Run: pip install pdf2image") sys.exit(1) os.makedirs(output_dir, exist_ok=True) images = convert_from_path( pdf_path, dpi=dpi, fmt="jpeg", first_page=page_num, last_page=page_num, ) if not images: print(f"Error: Could not extract page {page_num} from {pdf_path}") sys.exit(1) filename = f"slide_{page_num:03d}.jpg" filepath = os.path.join(output_dir, filename) images[0].save(filepath, "JPEG", quality=95) print(f" Re-extracted page {page_num} -> {filepath}") return filepath def copy_to_images_dir(image_paths: list[str], project_dir: str) -> list[str]: """Copy slide images to images/ as scene_N_frame.jpg for downstream pipeline.""" images_dir = os.path.join(project_dir, "images") os.makedirs(images_dir, exist_ok=True) copied = [] for i, src in enumerate(image_paths, start=1): dst = os.path.join(images_dir, f"scene_{i}_frame.jpg") shutil.copy2(src, dst) copied.append(dst) return copied def main(): parser = argparse.ArgumentParser( description="Extract slides from a PDF file as images" ) parser.add_argument("--pdf", required=True, help="Path to PDF file") parser.add_argument("--output-dir", help="Directory for extracted slide images") parser.add_argument("--output-json", help="Path for slides.json output") parser.add_argument("--total-duration", type=float, help="Total video duration in seconds (divided equally among slides)") parser.add_argument("--dpi", type=int, default=200, help="Render DPI for PDF pages (default: 200)") parser.add_argument("--dry-run", action="store_true", help="Preview page count without converting") args = parser.parse_args() pdf_path = os.path.abspath(args.pdf) if not os.path.exists(pdf_path): print(f"Error: PDF not found: {pdf_path}") sys.exit(1) print(f"\n{'='*60}") print("PDF Slide Extraction") print(f"{'='*60}") print(f"PDF: {pdf_path}") # Dry run: just show page count if args.dry_run: count = get_pdf_page_count(pdf_path) print(f"Pages: {count}") if args.total_duration: per_slide = args.total_duration / max(count, 1) print(f"Duration: {args.total_duration}s total, {per_slide:.1f}s per slide") else: print(f"Duration: {8.0 * count}s total (8s default per slide)") print("\nDry run complete.") return # Resolve output directories output_dir = args.output_dir if not output_dir: print("Error: --output-dir is required (not in dry-run mode)") sys.exit(1) os.makedirs(output_dir, exist_ok=True) # Convert PDF to images image_paths = convert_pdf_to_images(pdf_path, output_dir, dpi=args.dpi) print(f" Extracted {len(image_paths)} slides to {output_dir}") # Generate slides.json slides_data = generate_slides_json( image_paths, pdf_path, total_duration=args.total_duration, ) # Write slides.json output_json = args.output_json if output_json: os.makedirs(os.path.dirname(os.path.abspath(output_json)), exist_ok=True) with open(output_json, "w") as f: json.dump(slides_data, f, indent=2) print(f" slides.json: {output_json}") # Copy to images/ directory (infer project dir from output-dir) # output-dir is typically projects/{slug}/slides project_dir = os.path.dirname(output_dir) if os.path.basename(output_dir) == "slides": copied = copy_to_images_dir(image_paths, project_dir) print(f" Copied {len(copied)} images to {os.path.join(project_dir, 'images')}") print(f"\n{'='*60}") print(f"Extracted {len(image_paths)} slides") if args.total_duration: per_slide = args.total_duration / len(image_paths) print(f"Duration: {args.total_duration}s total, {per_slide:.1f}s per slide") else: print(f"Duration: {8.0 * len(image_paths)}s total (8s default per slide)") print(f"{'='*60}\n") if __name__ == "__main__": main()