Take previously added/removed nodes into account when fetching child …

[Moerphy]

Description
When when a node has two child nodes that would violate the core constraints (i.e. two inline nodes as children of the Editor) this can cause other nodes to be deleted that do not violate any core constraints.

Example
Input:

<editor>
    <inline>one</inline>
    <block>two</block>
    <inline>three</inline>
    <block>four</block>
</editor>

Normalized result before this change:

<editor>
    <block>four</block>
</editor>

Normalized result after this change:

<editor>
    <block>two</block>
    <block>four</block>
</editor>

See adapted unit tests for more examples.

Context
Disclaimer: I only know very little about Slates internal, so my understanding here might not be 100% correct.

Looks to me like the current logic for iterating children in the normalization uses two sets of variables node & i and currentNode & n. I think node & i are referring to the state of the editor when normalizeNode was entered and currentNode& n are referring the "current" state, taking into account transforms from an earlier iteration.
The transforms that are executed when a constraint violation is detected are always using n to construct the path to the transformed node, but the constraint was checked on the child node at index i instead, which leads to the wrong nodes being transformed.

Checks

• The new code matches the existing patterns and styles.
• The tests pass with yarn test.
• The linter passes with yarn lint. (Fix errors with yarn fix.)
• The relevant examples still work. (Run examples with yarn start.)
• You've added a changeset if changing functionality. (Add one with yarn changeset add.)

[changeset-bot]

🦋 Changeset detected

Latest commit: 8640c58

The changes in this PR will be included in the next version bump.

This PR includes changesets to release 1 package

Name	Type
slate	Patch

Not sure what this means? Click here to learn what changesets are.

Click here if you're a maintainer who wants to add another changeset to this PR

[dylans]

Looks reasonable to me @Moerphy . We'll land it. If it breaks something else we can always revert.