#!/usr/bin/env python3 from pathlib import Path from zipfile import ZipFile import xml.etree.ElementTree as ET import shutil, re, json, hashlib, os from PIL import Image, ImageOps, ImageDraw, ImageFont NS = { 'a':'http://schemas.openxmlformats.org/drawingml/2006/main', 'p':'http://schemas.openxmlformats.org/presentationml/2006/main', 'r':'http://schemas.openxmlformats.org/officeDocument/2006/relationships', 'rel':'http://schemas.openxmlformats.org/package/2006/relationships', } def natural_key(p): s=str(p) return [int(t) if t.isdigit() else t for t in re.split(r'(\d+)', s)] def safe_name(s): s = re.sub(r'[\\/:*?"<>|\s]+', '_', s).strip('_') return s[:80] or 'ppt' def parse_rels(z, rel_path): rels={} if rel_path not in z.namelist(): return rels root=ET.fromstring(z.read(rel_path)) for rel in root.findall('rel:Relationship', NS): rid=rel.attrib.get('Id') target=rel.attrib.get('Target') typ=rel.attrib.get('Type','') rels[rid]=(target,typ) return rels def target_to_zip_path(slide_path, target): if target.startswith('/'): return target.lstrip('/') base=Path(slide_path).parent return str((base / target).as_posix()).replace('ppt/slides/../','ppt/') def iter_slides(z): return sorted([n for n in z.namelist() if re.match(r'ppt/slides/slide\d+\.xml$', n)], key=natural_key) def extract_text(root): vals=[] for t in root.findall('.//a:t', NS): if t.text and t.text.strip(): vals.append(t.text.strip()) return vals def image_size(path): try: with Image.open(path) as im: return im.size except Exception: return (0,0) def make_thumb(img_path, size=(240,170)): try: im=Image.open(img_path).convert('RGB') im.thumbnail(size, Image.LANCZOS) canvas=Image.new('RGB', size, 'white') x=(size[0]-im.width)//2; y=(size[1]-im.height)//2 canvas.paste(im,(x,y)) return canvas except Exception: return Image.new('RGB', size, '#eeeeee') def draw_wrapped(draw, text, xy, font, fill, max_width, line_spacing=3, max_lines=4): x,y=xy; line=''; lines=[] for ch in text: test=line+ch if draw.textbbox((0,0), test, font=font)[2] <= max_width: line=test else: if line: lines.append(line) line=ch if line: lines.append(line) if len(lines)>max_lines: lines=lines[:max_lines] lines[-1]=lines[-1][:-1]+'…' for ln in lines: draw.text((x,y), ln, font=font, fill=fill) y += font.size + line_spacing return y def main(): project=Path('/Users/bot1/Volumes/root_for_ai/AI工作区/文博IP_PPT_公司文创开发能力展示_20260604_2014') sources=[] for p in (project/'source').rglob('*.pptx'): sources.append(p) sources=sorted(sources, key=lambda p: str(p)) outroot=project/'work'/'incoming_extract_20260605' outroot.mkdir(parents=True, exist_ok=True) summary=[] font=None try: font=ImageFont.truetype('/System/Library/Fonts/PingFang.ttc', 14) font2=ImageFont.truetype('/System/Library/Fonts/PingFang.ttc', 18) except Exception: font=ImageFont.load_default(); font2=ImageFont.load_default() for ppt in sources: name=safe_name(ppt.stem) out=outroot/name media_out=out/'media' slide_out=out/'slides' media_out.mkdir(parents=True, exist_ok=True); slide_out.mkdir(parents=True, exist_ok=True) deck={'file':str(ppt), 'slides':[]} with ZipFile(ppt) as z: for idx, sp in enumerate(iter_slides(z), start=1): root=ET.fromstring(z.read(sp)) texts=extract_text(root) rel_path='ppt/slides/_rels/'+Path(sp).name+'.rels' rels=parse_rels(z, rel_path) rids=[] for blip in root.findall('.//a:blip', NS): rid=blip.attrib.get('{%s}embed'%NS['r']) or blip.attrib.get('{%s}link'%NS['r']) if rid and rid not in rids: rids.append(rid) imgs=[] for n,rid in enumerate(rids, start=1): if rid not in rels: continue target,typ=rels[rid] if 'image' not in typ: continue zp=target_to_zip_path(sp,target) if zp not in z.namelist(): continue ext=Path(zp).suffix or '.bin' data=z.read(zp) h=hashlib.sha1(data).hexdigest()[:10] fn=f'slide{idx:03d}_{n:02d}_{h}{ext}' fp=media_out/fn if not fp.exists(): fp.write_bytes(data) w,hgt=image_size(fp) imgs.append({'file':str(fp.relative_to(project)), 'w':w, 'h':hgt, 'zip_path':zp}) deck['slides'].append({'slide':idx,'texts':texts,'images':imgs}) # markdown text md=[] for s in deck['slides']: md.append(f"## Slide {s['slide']}") if s['texts']: md.extend([f"- {t}" for t in s['texts']]) else: md.append('- [no text]') md.append(f"images: {len(s['images'])}\n") (out/'slide_text.md').write_text('\n'.join(md), encoding='utf-8') (out/'manifest.json').write_text(json.dumps(deck, ensure_ascii=False, indent=2), encoding='utf-8') # contact sheet per deck tiles=[] for s in deck['slides']: # choose up to 4 largest imgs imgs=sorted(s['images'], key=lambda x:x['w']*x['h'], reverse=True)[:4] tile=Image.new('RGB',(560,260),'#f7f3ed') d=ImageDraw.Draw(tile) d.text((10,8), f"{name} | P{s['slide']}", font=font2, fill='#4b2e20') text=' / '.join(s['texts'][:8])[:120] draw_wrapped(d, text, (10,36), font, '#222222', 250, max_lines=6) x0=280; y0=34 for j,img in enumerate(imgs): thumb=make_thumb(project/img['file'], (130,95)) x=x0+(j%2)*138; y=y0+(j//2)*104 tile.paste(thumb,(x,y)); d.rectangle([x,y,x+129,y+94], outline='#d0c2b0') tiles.append(tile) if tiles: cols=2; rows=(len(tiles)+cols-1)//cols sheet=Image.new('RGB',(cols*560, rows*260),'white') for i,tile in enumerate(tiles): sheet.paste(tile, ((i%cols)*560,(i//cols)*260)) sheet_path=out/'contact_sheet.jpg' sheet.save(sheet_path, quality=88) summary.append({'name':name,'ppt':str(ppt),'slides':len(deck['slides']),'images':sum(len(s['images']) for s in deck['slides']),'out':str(out.relative_to(project))}) (outroot/'summary.json').write_text(json.dumps(summary, ensure_ascii=False, indent=2), encoding='utf-8') print(json.dumps(summary, ensure_ascii=False, indent=2)) if __name__=='__main__': main()