[Patterns]: Support the blockTypes prop for patterns fetched from Pattern Directory

[ntsekouras]

What?

This PR adds support for the blockTypes prop
The changes here are from this PR: #32113 and is just here to move things faster in order to land this for 6.2.

props @ryelle 🙇

Without this change, the patterns fetched from Pattern Directory wouldn't use this prop when they were registered. This become more obvious with this issue for adding more Query Loop patterns and it affects the header, footer patterns that are now fetched from Pattern Directory here: #46017

Also noting that the added patterns from #46017, seem to still don't have the blockTypes prop in PD and they would need to be updated to include this.

Before

Screen.Recording.2023-02-02.at.11.35.59.AM.mov

After

Screen.Recording.2023-02-02.at.11.36.55.AM.mov

Testing Instructions

1. Insert a Heading block
2. Open the block switcher menu
3. Observe that there is a pattern suggested, coming from Pattern Directory

[github-actions]

Flaky tests detected in 3e2d823.
Some tests passed with failed attempts. The failures may not be related to this commit but are still reported for visibility. See the documentation for more information.

🔍 Workflow run URL: https://github.com/WordPress/gutenberg/actions/runs/4073154985
📝 Reported issues:

• [Flaky Test] should always expand single line selection #39787 in specs/editor/various/multi-block-selection.test.js
• [Flaky Test] should forward delete across blocks #44353 in specs/editor/various/multi-block-selection.test.js
• [Flaky Test] should handle Enter across blocks #39763 in specs/editor/various/multi-block-selection.test.js
• [Flaky Test] should insert block with the slash inserter when using multiple words #40301 in specs/editor/various/inserting-blocks.test.js
• [Flaky Test] should write over selection #39938 in specs/editor/various/multi-block-selection.test.js

[carolinan]

The feature with the transform works well.
I am unable to review the test.

In gutenberg_load_remote_block_patterns the viewport_width setting seems unrelated to enabling block types -but I don't see a reason to remove it if it speeds things up.

[ntsekouras]

In gutenberg_load_remote_block_patterns the viewport_width setting seems unrelated to enabling block types -but I don't see a reason to remove it if it speeds things up.

Yes, it is, but it was caught from @ryelle and I think it's land this too.

[ntsekouras]

This setting is not directly related to the main block_types change, but it's okay fix this too.

[Mamaduka]

I left a few minor notes, but the fix works as expected. PHPUnit tests are passing locally as well ✅

[Mamaduka]

Thank you, Nik. Updates look good 🚢

[Mamaduka]

I just cherry-picked this PR to the release/15.1 branch to get it included in the next release: f0c0581