Update Dependencies

[hobbescodes]

Description

The main purpose of this PR is updating dependencies (minus storybook related deps) as well as peer dependencies that downstream apps must rely on. A few other changes were made in order to resolve some underlying issues:

• Omit item prop in itemProps on Combobox as it is handled internally
• Updated Splitter props and logic to handle breaking changes from the dependency updates as well as provide better type safety
• Updated rendering logic for Icon to prevent recreation of component classes on every render
• Updated logic for useBreakpoint to derive state during render and prevent cascading renders due to synchronous setStates within effects

Test Steps

1. Verify that logic is sound
2. Verify that all components render as expected in storybook
3. Verify that tests pass
4. Verify that components are bundled properly by testing out downstream usage

[changeset-bot]

🦋 Changeset detected

Latest commit: dd587c7

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

This PR includes changesets to release 1 package

Name	Type
@omnidev/sigil	Minor

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

[pkg-pr-new]

Open in StackBlitz

npm i https://pkg.pr.new/omnidotdev/sigil/@omnidev/sigil@178

commit: dd587c7

[hobbescodes]

We should normalize a pattern for this across the board at some point. A lot of components that use a trigger have different compositional patterns atm.

[hobbescodes]

ts-pattern tip: If we match on the entire section, and then derive the arms based on destructured props a few cool things happen:

1. Since it knows sectionType can only be either panel or resizeTrigger you can see below that the exhaustive pattern still works properly. I thought this was pretty cool. You can increase the scope of this (i.e. match on multiple props in the same arm) and it will still provide type safety if you miss a combination.

2. The entire section is passed to the return handler and its properties are narrowed in terms of type inference! Which is why we don't need to do any casting for i.e. id in the resizeTrigger match arm, because it is inferred already based on the matching sectionType.

Cool stuff!

[hobbescodes]

This is the only line added to this file, the rest was formatting. Happy to dig into this, just didnt want to mess with this file at all as it is used throughout every component. If there is indeed issue here though with this generalized context helper as a whole, it could provide major perf impact in downstream apps, so may be worth digging into in the future!