fix(v2): i18n translation extractor should handle JSX formatting edge cases better

[slorber]

Motivation

Jest has this JSX content that would not be extracted properly by the i18n translation extractor:

          <Translate>
            {`Find out what's new with Jest.
- Follow[Jest](https://twitter.com/fbjest) on Twitter.
- Subscribe to the[Jest blog](/blog/).
- Look at the[changelog](https://github.com/facebook/jest/blob/master/CHANGELOG.md).`}
          </Translate>

This is become Babel AST parses this as 3 childrens: [empty text node, JSX expression, empty text node]

In such case we'll filter out the empty text nodes to focus mainly on the important node containing the translations

Have you read the Contributing Guidelines on pull requests?

yes

Test Plan

unit tests

[netlify]

✔️ Deploy preview for docusaurus-2 ready!

🔨 Explore the source changes: 4bbd4e2

🔍 Inspect the deploy logs: https://app.netlify.com/sites/docusaurus-2/deploys/5fd7af84078a610008fc93a3

😎 Browse the preview: https://deploy-preview-3920--docusaurus-2.netlify.app

[github-actions]

Size Change: +19 B (0%)

Total Size: 152 kB

ℹ️ View Unchanged

Filename	Size	Change
website/build/blog/2017/12/14/introducing-docusaurus/index.html	20.7 kB	0 B
website/build/docs/introduction/index.html	180 B	0 B
website/build/index.html	5.8 kB	-1 B
website/build/main.********.js	108 kB	+20 B (0%)
website/build/styles.********.css	17.6 kB	0 B

_{compressed-size-action}

[github-actions]

⚡️ Lighthouse report for the changes in this PR:

Category	Score
🔴 Performance	45
🟢 Accessibility	99
🟢 Best practices	100
🟢 SEO	100
🟢 PWA	95

Lighthouse ran on https://deploy-preview-3920--docusaurus-2.netlify.app/classic/