Updated Pack properties to use new Travertino dataclass-based API

[HalfWhitt]

This has been tested against the up-to-date Travertino with beeware/travertino#141, but can't be merged here until there's a new release of Travertino and Toga's dependency is updated.

PR Checklist:

• All new features have been tested
• All new features have been documented
• I have read the CONTRIBUTING.md file
• I will abide by the code of conduct

[freakboy3742]

@HalfWhitt I don't know if you've got any plans for more Travertino changes; the only thing that comes to mind for me was the issue about py3.13 support that I flagged during the review of the data class PR.

Once you're done, let me know and I'll cut a release to clear the path for this PR.

[HalfWhitt]

@freakboy3742 Well... I have started converting the rest of the tests to pytest. But that shouldn't affect anything outward-facing. Up to you whether you'd rather wait till that's done or not.

[HalfWhitt]

On second thought, I suppose this would be a good time to see about implementing list_property and/or composite_property, assuming those are still intended to happen. I only know about them from the commented-out sections. It looks like composite_property is supposed to function something like how the font parser already does, splitting a string into the relevant aliased properties, is that correct?

[freakboy3742]

On second thought, I suppose this would be a good time to see about implementing list_property and/or composite_property, assuming those are still intended to happen. I only know about them from the commented-out sections. It looks like composite_property is supposed to function something like how the font parser already does, splitting a string into the relevant aliased properties, is that correct?

That's correct. composite_property would be used by a font property, and list_property by font_family - with the obvious complication that font_family is then the first part of `font.

Those two definitions have been skipped because Pack has been able to survive without them - for the purposes of a "CSS-lite", it's fine to require the more explicit font syntax. They've only been stubbed out because they will be required for Colosseum.

I guess the question about whether to wait or not is up to you. They're not super-high priorities for me personally, because there's a totally acceptable workaround - but it they motivate you, then it's probably worth waiting so the next Travertino release is a really juicy one. And if someone (cough) were to use this as the start of getting Colosseum back on track... I also would not complain :-)

[HalfWhitt]

Alright yeah, hold off for now, I think I'll try my hand at them. And no promises about Colosseum, but I had in fact been thinking of such a thing...

[HalfWhitt]

Closed in favor of making this change as part of the Travertino migration (#3086).

[HalfWhitt]

Closed in favor of making this change as part of the Travertino migration (#3086).

Leaving a note here in case I or anyone else later comes by, spelunking through history: Pack's property syntax was indeed updated to the new descriptor API in that PR, but Pack wasn't actually decorated as a dataclass until #3215.