fix: truncate blogpost properly with truncate-html

[endiliey]

Motivation

Fix #719

Instead of naively cutting down markdown at certain length and then render it to HTML, we render markdown to HTML first and then truncate it (keeping html tag safe) using https://www.npmjs.com/package/truncate-html

Have you read the Contributing Guidelines on pull requests?

Yes

Test Plan

Consider this blogpost file

---
title: Docusaurus
author: Endilie
authorURL: https://github.com/endiliey
authorFBID: 100000251103620
authorTwitter: endiliey
---

Hi, this is just a test blog post. Please be a nice person. Hehehehehe

[Docusaurus](https://github.com/facebook/Docusaurus)
[Prettier](https://github.com/prettier/prettier)
[Jest](https://github.com/facebook/jest)
[ReasonML](https://github.com/reasonml)
[Reason React](https://github.com/reasonml/reason-react)

Lorem ipsum Lorem ipsum Lorem ipsum Lorem ipsum Lorem ipsum 
Lorem ipsum Lorem ipsum Lorem ipsum Lorem ipsum Lorem ipsum 

Before
Notice how the URL got cut naively, resulting on broken link like below
[ReasonML](<a href="https://github.com/reaso">https://github.com/reaso</a>

After
Notice how the URL is preserved
<a href="https://github.com/reasonml">ReasonML</a> <a href="https://github.com/reasonml/reason-react">Reason React</a>

[docusaurus-bot]

Deploy preview for docusaurus-preview ready!

Built with commit 6bc6d95

https://deploy-preview-880--docusaurus-preview.netlify.com

[JoelMarcey]

cc @chenglou

[JoelMarcey]

I have a slight concern that the project hasn't been updated in 4 months, but if this works for now, then I think it can solve some headaches.

[endiliey]

@JoelMarcey i'm betting on the package 6,937 downloads per week credibility 👍

Edit: Either way, i think at least it is better than previous logic (see above test plan). If it needs an update then we can consider forking it