Update dependency date-fns to v4

[renovate]

This PR contains the following updates:

Package	Change	Age	Adoption	Passing	Confidence
date-fns	3.6.0 -> 4.1.0

---

Release Notes

date-fns/date-fns (date-fns)

v4.1.0

Compare Source

This release adds time zone support to format functions (that I somehow missed when working on the feature) and fixes a few bugs.

Make sure also upgrade TZDate to v1.0.2 as it includes a bunch of critical bug fixes.

Fixed

• Fixed internal constructFrom throwing an exception on null arguments. While null isn't allowed, the functions should rather return Invalid Date or NaN in such cases. See #​3885.

Added

• Added missing time zone support to format, formatISO, formatISO9075, formatRelative and formatRFC3339. See #​3886.

v4.0.0

Compare Source

I have great news! First, ten years after its release, date-fns finally gets first-class time zone support.

Another great news is that there aren't many breaking changes in this release. All of them are type-related and will affect only those explicitly using internal date-fns types. Finally, it has been less than a year since the last major release, which is an improvement over the previous four years between v2 and v3. I plan on keeping the pace and minimizing breaking changes moving forward.

Read more about the release in the announcement blog post.

- Sasha @​kossnocorp

Added

• Added time zones support via @date-fns/tz's TZDate class and tz helper function. See its README for the details about the API.

• All relevant functions now accept the context in option, which allows to specify the time zone to make the calculations in. If the function also returns a date, it will be in the specified time zone:

  import { addDays, startOfDay } from "date-fns";
  import { tz } from "@​date-fns/tz";
  
  startOfDay(addDays(Date.now(), 5, { in: tz("Asia/Singapore") }));
  //=> "2024-09-16T00:00:00.000+08:00"

  In the example, addDays will get the current date and time in Singapore and add 5 days to it. startOfDay will inherit the date type and return the start of the day in Singapore.

Changed

• The function arguments, as well as Interval's start and end, now can be of different types, allowing you to mix UTCDate, TZDate, Date, and other extensions, as well as primitives (strings and numbers).

  The functions will normalize these values, make calculations, and return the result in the same type, preventing any bugs caused by the discrepancy. If passed, the type will be inferred from the context in option or the first encountered argument object type. The Interval's start and end will be considered separately, starting from start.

  In the given example, the result will be in the TZDate as the first argument is a number, and the start takes precedence over the end.

  clamp(Date.now(), {
    start: new TZDate(start, "Asia/Singapore"),
    end: new UTCDate(),
  });
  //=> TZDate

• BREAKING: This release contains a bunch of types changes that should not affect the library's expected usage. The changes are primarily internal and nuanced, so rather than listing them here, I recommend you running the type checker after the upgrade. If there are unfixable problems, please open an issue.

• BREAKING: The package now is ESM-first. The CommonJS is still support and It should not affect most users, but it might break in certains environments. If you encounter any issues, please report them.

Fixed

• Fixed CDN build compatibility with jQuery and other tools that expose $ by properly wrapping the code in an IIFE.

---

Configuration

📅 Schedule: Branch creation - "* 0-4,22-23 * * 1-5,* * * * 0,6" (UTC), Automerge - At any time (no schedule defined).

🚦 Automerge: Disabled by config. Please merge this manually once you are satisfied.

♻ Rebasing: Whenever PR becomes conflicted, or you tick the rebase/retry checkbox.

🔕 Ignore: Close this PR and you won't be reminded about this update again.

---

• If you want to rebase/retry this PR, check this box

---

This PR was generated by Mend Renovate. View the repository job log.

[louise-davies]

The test x-axis tab > min and max fields > date-time values > initialises correctly with falsy values is failing - rather than defaulting to 1970-01-01 00:00:00 it's now an empty string. This feels like acceptable behaviour so just need to update the test to check for empty string not UNIX Epoch

[louise-davies]

Need to wait for date-io/date-fns to update as well - seems to work fine but it complains about non-overlapping ranges...