Normative: Add firstDayOfWeek and consider "-u-fw"

[FrankYFTang]

Also change the return value of getWeekInfo() to use string instead of integer.

[FrankYFTang]

@anba

[anba]

Question: Do we need to do anything else to clarify that the value of the extension keyword can change the outcome of WeekInfoOfLocale, [...]

I think it should be clarified in more detail → tc39/proposal-intl-locale-info#30.

[anba]

This is also changing [[FirstDay]] and [[Weekend]] from numeric values to string values. I thought we intentionally used numeric values to:

1. Not tie weekday names to a specific calendar.
2. Match Temporal, which also uses integers from 1-7, cf. ToISODayOfWeek and CalendarDateDayOfWeek.

So we have to decide if it's more important to keep the values returned from Intl.Locale.prototype.getWeekInfo() consistent with the Unicode extension values for u-fw or with the Temporal API.

See also tc39/proposal-intl-locale-info#73 for another case where Unicode extension values and values returned from Intl.Locale won't match.

---

Edit: Also mentioned in https://github.com/tc39/proposal-intl-locale-info/issues/68#issuecomment-1574243778.

[FrankYFTang]

I totally agree we need to discuss the issue about the return value (string vs number) more. We may need to treat the value returned by Intl.Locale.prototype.firstDayOfWeek and Intl.Locale.prototype.getWeekInfo() differently.

[FrankYFTang]

Another ref https://www.unicode.org/reports/tr35/tr35-31/tr35-dates.html#Week_Data

<!ELEMENT firstDay EMPTY >
<!ATTLIST firstDay day (sun | mon | tue | wed | thu | fri | sat) #REQUIRED>
<!ATTLIST firstDay territories NMTOKENS #REQUIRED>

[FrankYFTang]

Related Temporal spec text https://tc39.es/proposal-temporal/#sec-temporal-toisodayofweek https://tc39.es/proposal-temporal/#sup-temporal.calendar.prototype.dayofweek https://tc39.es/proposal-temporal/#sec-temporal-calendardatedayofweek

Related spec to be considered to change https://tc39.es/proposal-intl-locale-info/#sec-Intl.Locale.prototype.getWeekInfo https://tc39.es/proposal-intl-locale-info/#sec-week-info-of-locale

[FrankYFTang]

UTS spec text https://www.unicode.org/reports/tr35/#UnicodeFirstDayIdentifier

[FrankYFTang]

https://tc39.es/ecma262/#sec-week-day Basically we will have three way to represent Sunday

• Date - Sunday = 0
• Temporal- Sunday = 7
• Intl.Locale.firstDayOfWeek / "-u-fw-sun" in locale / {firstDayOfWeek: "sun"} => "sun"

so the question is which representation should be used for Intl.Locale.getWeekInfo().firstDay and Intl.Locale.getWeekInfo().weekend

[FrankYFTang]

TG2 agree that we should keep using 1-7 for value inside getWeekInfo() sync w/ Temporal TG2 also suggest keep firstDay (not change to firstDayOfWeek for the return object of getWeekInfo()) since it is in number instead of string form. Revert that part of the change in this PR after that.

[FrankYFTang]

This is also changing [[FirstDay]] and [[Weekend]] from numeric values to string values.

Just to make it clear. that part of changes is already reverted and removed from this PR after our last TG2 meeting.

[FrankYFTang]

@anba I added some more changes to make it sure if we have -u-fw- we will override it. Reviewers could you please take another look and give me feedback before TC39 in July 11?

[FrankYFTang]

hum... one issue need to be clearify by Mark in https://unicode-org.atlassian.net/browse/CLDR-16839

[FrankYFTang]

The changes to WeekInfoOfLocale may be invalid - depending on the answer on https://unicode-org.atlassian.net/browse/CLDR-16839

[FrankYFTang]

v8 tracking bug https://bugs.chromium.org/p/v8/issues/detail?id=14159 v8 impl cl https://chromium-review.googlesource.com/c/v8/v8/+/4670213

[FrankYFTang]

The PR in the current state does not reach consensus in TC39 during July 12, 2023 meeting

I put down 4 new proposals in https://github.com/tc39/proposal-intl-locale-info/issues/68#issuecomment-1633352897 Let's discuss the option there

[FrankYFTang]

The topic has no enough time to be discussed during last TG2 but a big group of people share their opion within 10 minutes of the official end of TG2 and favor

Option C (Also mentioned in the TC39)

I. Intl.Locale.prototype.firstDayOfWeek returns: <<undefined, 1, 2, 3, 4, 5, 6, 7>>

2. {firstDayOfWeek} for new Intl.Locale() accepts: <<undefined, "mon", "tue", "wed", "thu", "fri", "sat", "sun", 0, 1, 2, 3, 4, 5, 6, 7>> Default: undefined where 0, 7, "sun" all repesent Sunday in option reading

I will change this PR to reflect that opinion soon and discuss that in TG1

[FrankYFTang]

PTAL

[FrankYFTang]

This PR is ready to review. Be aware it handle the "fw" keyword in the locale and return firstDayOfWeek in the set of <1 .. 7> The impact of ca = "iso8601" is NOT handled in this PR (For issue https://github.com/tc39/proposal-intl-locale-info/issues/30 ) For the ca issue, it is waiting for the resolution of https://unicode-org.atlassian.net/browse/CLDR-16866 which should come out soon. So that part will be a different normative PR after the merge of this one.

[FrankYFTang]

Please review and give me your feedback before the TC39 Sept 2023 in 9/26. I would like to get consensus to merge that in TG1

[FrankYFTang]

test https://github.com/tc39/test262/pull/3924

[FrankYFTang]

Consensus reach in 2023-09-26 TC39 meeting.

[trflynn89]

Hello - I'm implementing this change over in Serenity's LibJS and had a question. Do the AO definitions of WeekdayToNumber and WeekdayToString seem backwards? Currently WeekdayToNumber returns a string, and WeekdayToString returns a number, which seems pretty confusing to me.

[trflynn89]

It also seems like it's possible to hit the Assert: Should not reach here. step in WeekdayToNumber if we provide an invalid value as a -u-fw extension.

For example, if the user provides en-u-fw-100, then in step 34 of Intl.Locale:

34. Let r be ! ApplyUnicodeExtensionToTag(tag, opt, relevantExtensionKeys).

Then r.[[fw]] will have a value of 100. We will pass that to WeekdayToNumber and fail that assertion.

[FrankYFTang]

That is a good point. File a different issue https://github.com/tc39/proposal-intl-locale-info/issues/78 Let's move the discussion to that issue.