Fix slugify bug affecting headings containing special characters

[michaelhogg]

🔴 Problem

As described in issue #17, Markdown heading text containing special characters doesn't get correctly slugified, and so the links from the TOC to the headings don't work.

🔬 Investigation

By default, this extension uses the slugify() function provided by markdown-it-table-of-contents:

• https://github.com/cmaas/markdown-it-table-of-contents/blob/v0.4.3/index.js#L2

  const slugify = (s) => encodeURIComponent(String(s).trim().toLowerCase().replace(/\s+/g, '-'));

However, that function's output is incompatible with VSCode's slugifier ❌

🟢 Solution

This PR copies the exact logic from VSCode's slugifier:

• https://github.com/microsoft/vscode/blob/1.108.2/extensions/markdown-language-features/src/slugify.ts#L51-L53
• https://github.com/microsoft/vscode/blob/1.108.2/extensions/markdown-language-features/src/slugify.ts#L60-L63

to ensure that the slugification logic is accurate.

ℹ️ Example

The following example heading text:

This example heading & contains — special → characters 🌈

is slugified by markdown-it-table-of-contents's slugify() function like this: ❌

this-example-heading-%26-contains-%E2%80%94-special-%E2%86%92-characters-%F0%9F%8C%88

This PR fixes the slugification logic, so that that example heading text is now slugified to:

this-example-heading--contains--special--characters-

which is identical to the slug generated by VSCode's slugifier ✅