fix(v2): markdown title parser should ignore all forms of MDX import statements

[nam-hle]

Motivation

Currently the markdown title can be parsed after some kind of import statements. For example:

import Component1 from '@site/src/components/Component1';

import './styles.css';

# Markdown Title

Lorem Ipsum

This PR allow the most of import syntaxes.

import DefaultComponent from '@site/src/components/Component1';
import DefaultComponent2 from '../relative/path/Component2';
import * as EntireComponent from './relative/path/Component3';

import { Component4 } from "double-quote-module-name";
import { Component6 as AliasComponent6 } from "module-name";
import DefaultComponent8, { DefaultComponent81 } from "module-name";
import DefaultComponent9, * as EntireComponent9 from "module-name";

import _ from 'underscore';
import "module-name";

# Markdown Title

Lorem Ipsum

Have you read the Contributing Guidelines on pull requests?

Yes.

Test Plan

Already added test for the change.

Related PRs

From @slorber's suggestion in #4729.

(If this PR adds or changes functionality, please take some time to update the docs at https://github.com/facebook/docusaurus, and link to your PR here.)

[netlify]

[V2]

Built without sensitive environment variables with commit 9d24419

https://deploy-preview-4735--docusaurus-2.netlify.app

[slorber]

thanks for this :) looks great