Fix llms-full.txt generation hang and add option to disable it

[svekars]

• Switch from ProcessPoolExecutor to ThreadPoolExecutor for parallel markdown generation, fixing a deadlock that caused llms-full.txt to be partially generated and the build to hang in large projects like PyTorch
• Add new llm_generate_full theme option to allow large projects to skip llms-full.txt generation entirely while still producing individual .md files and llms.txt

  html_theme_options = {                                    
      "llm_generate_md": "true",                                                                                                              
      "llm_generate_full": "false",  # Skip llms-full.txt
  }                                                                                                                                           

Defaults to "true".

Test plan

• Build docs with llm_generate_full = "true" (default) — verify llms-full.txt is generated
• Build docs with llm_generate_full = "false" — verify llms-full.txt is not generated, individual .md files still are
• Build does not hang on large doc sets (e.g., PyTorch)

[netlify]

✅ Deploy Preview for pytorchsphinxtheme ready!

Name	Link
🔨 Latest commit	5b3b241
🔍 Latest deploy log	https://app.netlify.com/projects/pytorchsphinxtheme/deploys/69dd7a728b0fbd00080cec2e
😎 Deploy Preview	https://deploy-preview-245--pytorchsphinxtheme.netlify.app
📱 Preview on mobile	Toggle QR Code... Use your smartphone camera to open QR code link.

---

To edit notification comments on pull requests, go to your Netlify project configuration.