fix/1273: Fix child node length checks

[BoardJames]

Fixes a typo in the previous pull request #2482 (review)

[codecov]

Codecov Report

Merging #2532 into master will not change coverage.
The diff coverage is 0%.

@@          Coverage Diff           @@
##           master   #2532   +/-   ##
======================================
  Coverage    30.1%   30.1%           
======================================
  Files         174     174           
  Lines        5288    5288           
  Branches      905     905           
======================================
  Hits         1592    1592           
  Misses       3133    3133           
  Partials      563     563

Impacted Files	Coverage Δ
blocks/library/freeform/old-editor.js	1.51% <0%> (ø)	⬆️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update f3427c8...eca539c. Read the comment docs.

[aduth]

Presumably this won't work if the caret is within an empty element which is the fourth child of its body? Wondering if this is something we'd want to use recursion or loops, like in Editable:

gutenberg/blocks/editable/index.js

Lines 247 to 263 in ec49b3d

	isStartOfEditor() {
	const range = this.editor.selection.getRng();
	if ( range.startOffset !== 0 || ! range.collapsed ) {
	return false;
	}
	const start = range.startContainer;
	const body = this.editor.getBody();
	let element = start;
	while ( element !== body ) {
	const child = element;
	element = element.parentNode;
	if ( element.firstChild !== child ) {
	return false;
	}
	}
	return true;
	}

Or @iseulde's suggestion at #2482 (comment)

[BoardJames]

@aduth can you give an example of the content that could produce such an effect?

For example if I artifiically set the content to contain a span like:
tinyMCE.activeEditor.setContent('<p><span></span><br/></p>');
then the span is automatically removed.

The closest I can get to the situation you describe is by toggling on 'bold' in which case the first backspace removes the bold formatting (a behaviour I think is actually good).

[aduth]

Okay, on second glance, I see you're testing against the specific pattern of <p><br data-mce-bogus></p>.

What are your thoughts on using tinymce.DOM.isEmpty instead, which accounts for many other types of content? (Edit: I see you responded at #2482 (comment))

https://github.com/tinymce/tinymce/blob/cb7bba98a85b635e09283578b35d8c9436af77db/src/core/src/main/js/dom/DOMUtils.js#L1460-L1533