| |
| import os, sys, json, re, time, urllib.request |
| from huggingface_hub import HfApi |
|
|
| SPACE_ID = "bep40/V.AISTUDIO" |
| BASE_URL = "https://bep40-v-aistudio.static.hf.space" |
|
|
| def slugify(name): |
| s = ''.join(c for c in (name or '').lower() if c.isalnum() or c in '- ') |
| s = re.sub(r'-+', '-', s).strip('-') |
| return s or 'product' |
|
|
| def main(): |
| print("=== SEO Stub Generator ===") |
| api = HfApi() |
| |
| |
| print("[1/3] Downloading index.html...") |
| with urllib.request.urlopen(f"https://huggingface.co/spaces/{SPACE_ID}/resolve/main/index.html") as r: |
| html = r.read().decode('utf-8') |
| print(f" Size: {len(html):,} bytes") |
| |
| |
| print("[2/3] Extracting products...") |
| m = re.search(r'var\s+products\s*=\s*(\[[\s\S]*?\]);', html) |
| if not m: |
| print("ERROR: Cannot find products array") |
| return |
| products = json.loads(m.group(1)) |
| print(f" Found {len(products)} products") |
| |
| |
| print("[3/3] Uploading stubs...") |
| uploaded = 0 |
| for p in products: |
| slug = p.get('slug') or slugify(p.get('name', 'p')) |
| name = str(p.get('name', 'Product'))[:200] |
| brand = str(p.get('brand', 'V.AISTUDIO')) |
| img = p.get('image') or f"{BASE_URL}/logo/logo_600.png" |
| desc = re.sub(r'<[^>]+>', '', str(p.get('description', '')))[:160] |
| title = f"{name} | {brand}" |
| |
| stub = f'''<!DOCTYPE html> |
| <html lang="vi"> |
| <head> |
| <meta charset="UTF-8"> |
| <meta name="robots" content="noindex,follow"> |
| <title>{title}</title> |
| <meta name="description" content="{desc}"> |
| <link rel="canonical" href="{BASE_URL}/san-pham/{slug}/"> |
| <meta property="og:type" content="product"> |
| <meta property="og:title" content="{title}"> |
| <meta property="og:description" content="{desc}"> |
| <meta property="og:image" content="{img}"> |
| <meta property="og:url" content="{BASE_URL}/san-pham/{slug}/"> |
| <meta property="og:site_name" content="V.AISTUDIO"> |
| <meta name="twitter:card" content="summary_large_image"> |
| <script>window.location.replace("/?product={slug}");</script> |
| </head> |
| <body><h1><a href="/?product={slug}">{title}</a></h1></body> |
| </html>''' |
| |
| try: |
| api.upload_file( |
| path_or_fileobj=stub.encode('utf-8'), |
| path_in_repo=f"san-pham/{slug}/index.html", |
| repo_id=SPACE_ID, |
| repo_type="space", |
| commit_message=f"SEO: {name[:40]}" |
| ) |
| uploaded += 1 |
| if uploaded % 100 == 0: |
| print(f" {uploaded} stubs uploaded...") |
| except Exception as e: |
| pass |
| |
| print(f"\nDONE! {uploaded} stubs uploaded to {SPACE_ID}") |
|
|
| if __name__ == "__main__": |
| main() |
