# pptx_to_images_python_pptx.py from pathlib import Path from typing import List, Optional, Tuple from pptx import Presentation from pptx.enum.shapes import MSO_SHAPE_TYPE from PIL import Image, ImageDraw, ImageFont import textwrap import io import os def _load_font(preferred_size: int, bold: bool = False): """ Try to load a TTF for nicer rendering; fallback to default PIL font. """ # Try common system font names; adjust if missing candidates = [ "/usr/share/fonts/truetype/dejavu/DejaVuSans-Bold.ttf", "/usr/share/fonts/truetype/dejavu/DejaVuSans.ttf", "/Library/Fonts/Arial.ttf", "C:/Windows/Fonts/arial.ttf", ] for p in candidates: try: if Path(p).exists(): return ImageFont.truetype(str(p), preferred_size) except Exception: continue # fallback return ImageFont.load_default() def pptx_to_images_via_python_pptx( pptx_path: str, out_dir: str, size: Tuple[int, int] = (1280, 720), title_font_size: int = 44, bullet_font_size: int = 26, max_bullets: int = 6, right_image_width_ratio: float = 0.37 ) -> List[str]: """ Convert PPTX to PNG images using python-pptx + Pillow. - pptx_path: input .pptx - out_dir: directory where slide_{:03d}.png will be placed - size: (width, height) of output images - right_image_width_ratio: fraction of width reserved for images on right (0..1) Returns: list of generated image paths (in order). """ W, H = size out_dir_p = Path(out_dir) out_dir_p.mkdir(parents=True, exist_ok=True) pres = Presentation(pptx_path) slides = pres.slides title_font = _load_font(title_font_size, bold=True) bullet_font = _load_font(bullet_font_size, bold=False) generated: List[str] = [] for idx, slide in enumerate(slides, start=1): # Extract text and images title_text = "" bullet_lines: List[str] = [] slide_images: List[Path] = [] for shape in slide.shapes: # pictures if shape.shape_type == MSO_SHAPE_TYPE.PICTURE: try: img = shape.image blob = img.blob ext = img.ext # Save in-memory then to file img_name = out_dir_p / f"slide_{idx:03d}_img_{len(slide_images)+1}.{ext}" with open(img_name, "wb") as f: f.write(blob) slide_images.append(img_name) except Exception: # ignore if can't extract image continue # text frames elif shape.has_text_frame: text = shape.text.strip() if not text: continue # If we don't yet have a title, pick first text box as title if not title_text: # take first line as heading and remaining lines as bullets for convenience lines = [ln.strip() for ln in text.splitlines() if ln.strip()] if lines: title_text = lines[0] if len(lines) > 1: for l in lines[1:]: bullet_lines.extend([ln for ln in l.splitlines() if ln.strip()]) else: # append other text boxes as bullets (preserve internal line breaks) for l in text.splitlines(): if l.strip(): bullet_lines.append(l.strip()) # Fallback title if none found if not title_text: title_text = f"Slide {idx}" # Limit bullets bullet_lines = [b for b in bullet_lines if b] # clean if len(bullet_lines) > max_bullets: bullet_lines = bullet_lines[:max_bullets] # Post-process bullets: if bullets are short fragments, keep them but wrap into 1-2 sentences later in pipeline # Now render image with Pillow img = Image.new("RGB", (W, H), color=(255, 255, 255)) draw = ImageDraw.Draw(img) # Title area (top) padding = 36 title_h_area = int(H * 0.18) # Draw title centered (wrap if necessary) title_wrap_width = 40 title_lines = textwrap.wrap(title_text, width=title_wrap_width) # compute title start y y = padding for line in title_lines: bbox = draw.textbbox((0, 0), line, font=title_font) tw, th = bbox[2] - bbox[0], bbox[3] - bbox[1] x = (W - tw) // 2 draw.text((x, y), line, fill=(20, 20, 20), font=title_font) y += th + 6 # Body area (left) and image area (right) right_width = int(W * right_image_width_ratio) left_width = W - right_width - padding * 2 left_x = padding left_y = title_h_area bullet_wrap_chars = max(20, int(left_width / (bullet_font_size * 0.5))) # heuristic # Draw bullets bullet_y = left_y + 8 for b in bullet_lines: wrapped = textwrap.wrap(b, width=bullet_wrap_chars) # draw a small bullet circle bullet_dot_radius = 4 # draw wrapped lines for i, wline in enumerate(wrapped): if i == 0: # draw bullet circle on first line draw.ellipse((left_x, bullet_y + 6, left_x + bullet_dot_radius*2, bullet_y + 6 + bullet_dot_radius*2), fill=(34, 120, 200)) text_x = left_x + bullet_dot_radius*2 + 8 else: text_x = left_x + bullet_dot_radius*2 + 8 draw.text((text_x, bullet_y), wline, fill=(40,40,40), font=bullet_font) bbox = draw.textbbox((0, 0), wline, font=bullet_font) tw, th = bbox[2] - bbox[0], bbox[3] - bbox[1] bullet_y += th + 6 bullet_y += 4 # small gap after bullet # If slide has images, place first image on right area (scale to fit) if slide_images: img_path = slide_images[0] try: pil_img = Image.open(img_path).convert("RGB") # scale to fit inside right_width x available_height avail_w = right_width - padding avail_h = H - title_h_area - padding pil_w, pil_h = pil_img.size scale = min(avail_w / pil_w, avail_h / pil_h, 1.0) new_w = int(pil_w * scale) new_h = int(pil_h * scale) pil_img = pil_img.resize((new_w, new_h), Image.LANCZOS) # paste at right area centered vertically paste_x = W - new_w - padding paste_y = title_h_area + (avail_h - new_h) // 2 img.paste(pil_img, (paste_x, paste_y)) except Exception: pass # Save to file with zero-padded slide index out_path = out_dir_p / f"slide_{idx:03d}.png" img.save(out_path, format="PNG", optimize=True) generated.append(str(out_path)) return generated # Example usage if __name__ == "__main__": # adjust paths pptx_file = "Video_generation/ppt/photosynthesis-basics-1758447500.pptx" out_images_dir = "slides_images" imgs = pptx_to_images_via_python_pptx(pptx_file, out_images_dir, size=(1280,720)) print("Generated images:", imgs)