Fix SEO issues

[jacklin213]

Currently with HashRouter being used, Google bot won't actually crawl these pages. Possible ways of improvements is:

• Pre-Rendering into Static HTML Files
  • Such as React snap
• Using BrowserRouter instead of HashRouter.
  • GitHub pages has issues running with BrowserRouter out of the box as https://harrypotterspells.github.io/downloads will go the gh-pages server, the server will see there is no downloads project and return a 404
  • With some configuration, can use a trick to teach GitHub Pages to handle 404s by redirecting to index.html with a custom redirect parameter as shown here
• Static Rendering

[jacklin213]

Notes on React snap:

• Hasn't been updated in 2 years
• Only works with routing strategies using the HTML5 history API. No hash(bang) URLs. So have to migrate to BrowserRouter either way

Tried plugging it into the project (Added 17 red npm audit issues), was able to successfully run react-snap when site was using HashRouter (But only finds home page). Once migrated to BrowserRouter, it successfully detected 7 pages but only was able to pre-render downloads, commands, and spells.

No longer considering React snap a viable option. If SEO still is unable to index the sub-pages, will look into other pre-rendering solutions

[jacklin213]

Upon testing live URL in Google Search console, because the BrowserRouter hack relies on hitting a 404 first before serving 404.html which has the custom redirect logic, this method won't work as google won't index pages that return 404.

[jacklin213]

Didn't read through the README of spa-github-pages properly. The seo section clearly covers the above point and how you have to index the redirect route eg) https://harrypotterspells.github.io/?/spells as opposed to https://harrypotterspells.github.io/spells.

Though this still doesn't solve the problem that 404 is returned by Github Pages first before the correct route is returned.