The tests are passing an explicit `hour12` option, which disables any
`hourCycle` option and any `u-hc` Unicode extension. Therefore passing
input locales with `u-hc` is invalid.
```js
let locale = "en-u-hc-h24";
let hcDefault = new Intl.DateTimeFormat(locale, { hour: "numeric" }).resolvedOptions().hourCycle;
assert.sameValue(hcDefault, "h24", "hour-cycle through Unicode extension");
let hc24 = new Intl.DateTimeFormat(locale, { hour: "numeric", hour12: false }).resolvedOptions().hourCycle;
// Incorrect assertion, because |hc24| uses |hour12|, which disables any
// hour-cycle option and the hour-cycle is determined from the locale "en".
// That means |hc24| will be "h23".
assert.sameValue(hc24, hcDefault);
```
Time zone identifiers supported by ICU, but which aren't valid IANA
identifiers.
Implementations not supporting ECMA-402 are theoretically allowed to
support these identifiers, so the tests have to go into the "intl402"
directory.
These tests were incorrect, in checking for a RangeError when only one of
the era/eraYear fields were given. From CalendarResolveFields:
"The operation throws a *TypeError* exception if the properties of
_fields_ are internally inconsistent within the calendar or insufficient
to identify a unique instance of _type_ in the calendar."
Built-in non-ISO calendars require either monthCode/day, or month/day plus
some form of year specification.
This adds test coverage for each of the categories listed in
https://github.com/tc39/proposal-temporal/issues/2664, of which some must
currently reside in the test/intl402/ folders.
We'll do this for now, then separately work on migrating all of the tests
that require a non-ISO8601 calendar but aren't dependent on it being any
particular calendar.
* Add Tests for ECMA402 PR811
Add tests to check the order of option readings and output
keys in resolvedOptions of Intl.NumberFormat and PluralRules.
* Address reveiw feedback
Hard code the list of property to be inspect for GetOption
Use compareArray
* Update test/intl402/NumberFormat/constructor-option-read-order.js
Co-authored-by: Richard Gibson <richard.gibson@gmail.com>
* Update test/intl402/NumberFormat/constructor-option-read-order.js
Co-authored-by: Richard Gibson <richard.gibson@gmail.com>
* Update test/intl402/NumberFormat/prototype/resolvedOptions/return-keys-order-default.js
Co-authored-by: Richard Gibson <richard.gibson@gmail.com>
* Update test/intl402/PluralRules/constructor-option-read-order.js
Co-authored-by: Richard Gibson <richard.gibson@gmail.com>
* Update test/intl402/PluralRules/constructor-option-read-order.js
Co-authored-by: Richard Gibson <richard.gibson@gmail.com>
* Fix intl402/PluralRules/prototype/resolvedOptions/return-keys-order-default.js
To test all options
* Add more tests
---------
Co-authored-by: Richard Gibson <richard.gibson@gmail.com>
Test the explicit setting of ignorePunctuation in the option bag
reflect to the value in resolvedOptions() under Thai and
other locales. (Thai is a special case)
Test the behavior of the compare sync with the value of
resolvedOptions().ignorePunctuation
Note the monkeypatch of getPossibleInstantsFor in test/built-ins/Temporal/
TimeZone/prototype/getInstantFor/argument-builtin-calendar-no-array-
iteration.js.
Other than that, all the tests are basically identical.
This commit verifies that ISO strings with sub-minute offsets cannot
be parsed into time zone identifiers. This was a change introduced in
the recently-merged tc39/proposal-temporal#2607, but tests for this case
were missing from #3862 (the tests for that PR).
I noticed in codecov results on an unrelated PR that this case wasn't
being tested, so fixing that mistake now.
