ENH: use lychee for linkchecking

[mmcky]

This PR sets up lychee for more robust link checking and:

1. removes link-checking from the PR process (as a distraction)
2. uses lychee as a more robust solution to parse html
3. checks the live site gh-pages every day.
4. Posts an issue with any results (excluding 403 and 503)

Interestingly the only 503 results are all the AI company websites. They are blocking web traffic requests! Kind of ironic given how they have been trained.

This PR also fixes the canonical site address from python. to python-programming. that is embedded in the html.

---

Issues with checking md directly.

These types of links [lecture on broadcasting](broadcasting) get resolved to a web address by default which produces an error.

• perhaps should use html parsing instead of md parsing to avoid some of these minor syntax issues that Jupyter-book resolves at build time.

[github-actions]

🚀 Deployed on https://684278261ab4c10ad2529a14--epic-agnesi-957267.netlify.app