cd /tmp/workspace && python3 << 'EOF' import pdfplumber import os pdf_path = "/tmp/workspace/attachments/3f003029-1f5d-48ee-ac5e-4ccfd6a1d1d5/chap 51.pdf" print(f"File exists: {os.path.exists(pdf_path)}") with pdfplumber.open(pdf_path) as pdf: print(f"Total pages: {len(pdf.pages)}") # Extract first 5 pages to see structure for i, page in enumerate(pdf.pages[:5]): print(f"\n{'='*60}") print(f"PAGE {i+1}") print('='*60) text = page.extract_text() if text: print(text[:3000]) else: print("[No text extracted]") EOF

python3 << 'EOF' import pdfplumber from pypdf import PdfReader pdf_path = "/tmp/workspace/attachments/3f003029-1f5d-48ee-ac5e-4ccfd6a1d1d5/chap 51.pdf" reader = PdfReader(pdf_path) print(f"Total pages: {len(reader.pages)}") # Try to extract images info with pdfplumber.open(pdf_path) as pdf: for i, page in enumerate(pdf.pages[:3]): print(f"Page {i+1} - images: {len(page.images)}, width: {page.width}, height: {page.height}") # Try with different settings text = page.extract_text(x_tolerance=3, y_tolerance=3) print(f" Text with tolerance: {repr(text[:200]) if text else 'None'}") EOF

python3 << 'EOF' import subprocess import os # Convert PDF pages to images using PIL/pypdf pdf_path = "/tmp/workspace/attachments/3f003029-1f5d-48ee-ac5e-4ccfd6a1d1d5/chap 51.pdf" output_dir = "/tmp/workspace/cleft_pages" os.makedirs(output_dir, exist_ok=True) # Use pypdf to extract embedded images from each page from pypdf import PdfReader from PIL import Image import io reader = PdfReader(pdf_path) print(f"Total pages: {len(reader.pages)}") # Extract images from pages for page_num, page in enumerate(reader.pages): images = page.images for img_num, img in enumerate(images): img_data = img.data try: pil_img = Image.open(io.BytesIO(img_data)) out_path = f"{output_dir}/page_{page_num+1:02d}_img{img_num}.png" pil_img.save(out_path) print(f"Saved page {page_num+1} img {img_num}: {pil_img.size} -> {out_path}") except Exception as e: print(f"Page {page_num+1} img {img_num} error: {e}") print("Done extracting images") EOF

python3 << 'EOF' from PIL import Image import os output_dir = "/tmp/workspace/cleft_pages" batch_dir = "/tmp/workspace/cleft_batches" os.makedirs(batch_dir, exist_ok=True) # Create 5 batches: pages 1-6, 7-11, 12-16, 17-22, 23-28 batches = [ (range(1, 7), "batch1_pages1to6.png"), (range(7, 12), "batch2_pages7to11.png"), (range(12, 17), "batch3_pages12to16.png"), (range(17, 23), "batch4_pages17to22.png"), (range(23, 29), "batch5_pages23to28.png"), ] for page_range, fname in batches: images = [] for p in page_range: img_path = f"{output_dir}/page_{p:02d}_img1.png" if os.path.exists(img_path): img = Image.open(img_path) # Resize to reasonable width for analysis ratio = 900 / img.width new_h = int(img.height * ratio) img = img.resize((900, new_h), Image.LANCZOS) images.append(img) print(f" Added page {p}: {img.size}") if images: total_h = sum(i.height for i in images) combined = Image.new('RGB', (900, total_h), (255, 255, 255)) y = 0 for img in images: combined.paste(img, (0, y)) y += img.height out_path = f"{batch_dir}/{fname}" combined.save(out_path, optimize=True, quality=85) file_size = os.path.getsize(out_path) / 1024 / 1024 print(f"Saved {fname}: {combined.size}, {file_size:.1f} MB") print("Done creating batches") EOF

python3 << 'EOF' from PIL import Image import os, base64, io output_dir = "/tmp/workspace/cleft_pages" small_dir = "/tmp/workspace/cleft_small" os.makedirs(small_dir, exist_ok=True) # Create small versions for each page for p in range(1, 29): img_path = f"{output_dir}/page_{p:02d}_img1.png" if os.path.exists(img_path): img = Image.open(img_path) # Resize to 600px wide ratio = 600 / img.width new_h = int(img.height * ratio) img = img.resize((600, new_h), Image.LANCZOS).convert('L') # grayscale out_path = f"{small_dir}/page_{p:02d}.jpg" img.save(out_path, quality=70) size = os.path.getsize(out_path) print(f"Page {p}: {size//1024}KB -> {out_path}") print("Done") EOF