refactor!: Seal the Dependency class

[spydon]

• Thanks for your contribution! Please replace this text with a description of what this PR is changing or adding and why, list any relevant issues, and review the contribution guidelines below.

---

• I’ve reviewed the contributor guide and applied the relevant portions to this PR.

Seals the Dependency class to make it easier to pattern match over the types.

[kevmoo]

changelog, please!

[kevmoo]

Also, we'll have to make this a breaking change to be paranoid

[spydon]

Also, we'll have to make this a breaking change to be paranoid

Since it had a private constructor before this change nobody could extend it, so I don't think we'd have to mark this as breaking?

EDIT: Ah, but they could have extended the subclasses...

[kevmoo]

Also, we'll have to make this a breaking change to be paranoid

Since it had a private constructor before this change nobody could extend it, so I don't think we'd have to mark this as breaking?

EDIT: Ah, but they could have extended the subclasses...

This is the annoying bit. It's like 99% likely to NOT be a problem...but...

[kevmoo]

@devoncarew @natebosch @jakemac53 – thoughts on breaking change or not here?

[spydon]

@kevmoo @natebosch @jakemac53 @devoncarew can we land this PR and #137 before moving the repo?

[kevmoo]

I'm waiting for a second opinion on sealing the classes being breaking...

[spydon]

I'm waiting for a second opinion on sealing the classes being breaking...

Isn't there some google stick you can poke them with? 😅

Right, now we get this of course:

Because pubspec_parse depends on json_serializable ^6.6.0 which depends on pubspec_parse ^1.0.0, pubspec_parse ^1.0.0 is required.

Imho I don't think we'd need to do a breaking change for this change, it's do incredibly unlikely that it would cause any breakage for anyone. I'll set it back to 1.3.1, and if you disagree I'll make it 2.0.0 again.

[devoncarew]

Imho I don't think we'd need to do a breaking change for this change, it's do incredibly unlikely that it would cause any breakage for anyone. I'll set it back to 1.3.1, and if you disagree I'll make it 2.0.0 again.

I guess the argument is that people couldn't have subclassed Dependency, they could have subclasses of SdkDependency, PathDependency, ..., but that it's unlikely that people would have actually done that? So technically a breaking change but one that's unlikely to break anyone in practice.

Do we know if the main users (a few of https://pub.dev/packages?q=dependency%3Apubspec_parse) are safe w/ this change?

[spydon]

@devoncarew I went through the two first pages (based on likes) and none of them extends any of the Dependency subclasses.

[natebosch]

Yeah I think it would be OK to land this without a major version change. We should switch to a feature version bump though.

[natebosch]

We will need to update the pubpsec too.

[spydon]

@natebosch bumped to 1.4.0

[kevmoo]

Thanks @spydon !