Built-in Temporal parsers

[denexapp]

Hi, Temporal spec is stable ans supported for a while now in major browsers. Since some date parsers are included to nuqs, it will be cool to see built-in parsers for Temporal too. I'd personally be happy to see a built-in Temporal.PlainDate parser

[franky47]

Good idea, although I would not put it in the main bundle until Temporal is Baseline widely available, we could expose a shadcn registry item with those parsers instead:

Parser	Type	URL format
parseAsTemporalInstant	Temporal.Instant	ISO 8601 — 2026-05-26T10:00:00Z
parseAsTemporalInstantTimestamp	Temporal.Instant	Epoch milliseconds — 1748252400000
parseAsTemporalZonedDateTime	Temporal.ZonedDateTime	ISO + IANA bracket — 2026-05-26T10:00:00-07:00[America/Los_Angeles]
parseAsTemporalPlainDateTime	Temporal.PlainDateTime	ISO without offset — 2026-05-26T10:00:00
parseAsTemporalPlainDate	Temporal.PlainDate	YYYY-MM-DD — 2026-05-26
parseAsTemporalPlainTime	Temporal.PlainTime	HH:MM:SS — 10:00:00
parseAsTemporalPlainYearMonth	Temporal.PlainYearMonth	YYYY-MM — 2026-05
parseAsTemporalPlainMonthDay	Temporal.PlainMonthDay	--MM-DD — --05-26
parseAsTemporalDuration	Temporal.Duration	ISO 8601 duration — PT1H30M

That would solve a few problems at once:

• Keeping core bundle size lean & tree shaking (keep only what you need)
• Polyfill or not decision left to the dev (they would likely already have done that beforehand)
• Tweaking the serialisation format as needed

What do you think?

Note: I would like to merge #1188 first so we have a test harness for registry items.

[denexapp]

• I've never used shadcn stuff (shame on me), so I can't comment on that. I guess it's a good idea to test it in the registry before adding it to the bundle
• I've totally missed the part that Temporal is not fully supported in Safari yet, it makes sense to not include that to the main bundle yet
• Tweaking the serialization format - not sure if this level of flexebility is needed
• Non-ISO calendars support is not something I need personally, but I think it should be supported as well