fix: Restore   behavior from v0.1.27

[karfau]

Might be considered a breaking change.

fixes #57

[brodybits]

Thanks. This would definitely be breaking for anything that does not expect \u00a0 characters for input with &nbsp characters.

[karfau]

I think it's a question of definition.
This was the behavior of 0.1.27 changed in 0.2.0 - 0.3.0(so not a semver issue since major is still 0),
Now it "breaks" it "back" so I guess we could call it 0.4.0.
or if you say it was an unintended regression we could include it in 0.3.1.

Your decision.

[brodybits]

Approved to be merged after PR #77 - it would be nice for me to validate that the Stryker dashboard is updated with this change.

I do have one non-blocking suggestion.

[brodybits]

Suggested change

	var dom = new DOMParser().parseFromString('<div>&nbsp;&copy;&nbsp&copy</div>','text/html');
	var dom = new DOMParser().parseFromString('<div>&copy;&nbsp;&copy;</div>>','text/html');

(simplify the test string and fix the indentation)

Or do we need to keep the "&nbsp&copy" part to test something? If so, I would kinda favor separate test strings, with separate test assertions, to check the   behavior and to test with "&nbsp&copy".

I do think it would be better to break this test script into smaller test cases, and potentially multiple files, at some point.

TODO: The assertion would have to be updated as well. Unfortunately the GitHub suggested changes feature does not work for multiple lines if some lines were deleted in the middle.

[karfau]

I agree that the tests are not ideal, that's why I raised #72 to discuss the options.

I would avoid to mix this into other PRs to not loose track.

[brodybits]

👍

[brodybits]

I went ahead with the merge, and the change does now show up on the Stryker dashboard. Thanks!