Wrong RegExp to determine the markdown links.

[ulivz]

Step to reproduce

When the markdown file contains typescript like:

- Type: `{ [path: string]: Object }`

We'll get following error:

File is not found: [path: string]: Object }` (/path/to/foo.md:86:12)

[cdauth]

I'm also running into this. Here are some more cases where this goes wrong:

• [`{ [markerId: number]: Marker }`](./types#marker) leads to File is not found: [`{ [markerId: number]: Marker }`](./types#marker) (it detects the whole link but fails)
• (`{ [fieldName: string]: string }`): The filled out form fields of the marker. By default, this is a null-prototype object to avoid prototype pollution. Have a look at [marker/line data](./advanced#marker-line-data) for more details. leads to File is not found: [fieldName: string]: string }`): The filled out form fields of the marker. By default, this is a null-prototype object to avoid prototype pollution. Have a look at [marker/line data](./advanced#marker-line-data) for more details. (Somehow it thinks the whole line is a link.)

[cdauth]

After thinking about whether I can write an easy fix for this, I came to the conclusion that with the current approach of check-md this is not possible. Markdown is far too complex to parse it using a simple regular expression. Consider the following examples:

[`[test](test)`](test)
[\[](test) [\]](test)
`[test` [test](test) `test](test)`
``test`[test`` [test](test) ``test`](test)``

In my opinion, a proper markdown parser would need to be used.

As a workaround, I'm writing my documentation like this for now:

- Type: <code>{ &#91;path: string&#93;: Object }</code>