Editorial: Add note to ref to UTS35 about weekinfo

[FrankYFTang]

Address #30 by deferring to UTS35
and invalidate tc39/ecma402#950

[ben-allen]

Making sure I understand the broader matter correctly: the current wording will respect -u-ca, but because "rg" is not in [[RelevantExtensionKeys]], -u-rg will not be taken into account when determining the returned Record's value?

I wonder if this note on the whole might be clearer if the specific extensions that affect the Record's value were listed in the note, and if it were explicitly stated that https://www.unicode.org/reports/tr35/tr35-dates.html#first-day-overrides determines which of the components of _locale_ have priority.

[anba]

[[RelevantExtensionKeys]] for Intl.Locale doesn't have the same semantics as [[RelevantExtensionKeys]] for Intl service constructors. I guess it's now finally time to rename [[RelevantExtensionKeys]] for Intl.Locale → tc39/ecma402#914.

[FrankYFTang]

My intention of this note is to delegate the whole algorithm to UTS35 instead of specify a half baked algorithm that we need to keep updating the algorithm if UTS35 change later.

[anba]

The link target https://www.unicode.org/reports/tr35/tr35-dates.html#first-day-overrides doesn't exist.

[FrankYFTang]

???? do you have internet problem? it is a good link

[anba]

The anchor "#first-day-overrides" doesn't exist.

The relevant source code:

<h4>First Day Overrides</h4>

This must instead be:

<h4><a name="first-day-overrides" href="#first-day-overrides">First Day Overrides</a></h4>

for anchors to work.

[FrankYFTang]

But

<script src="./js/anchor.min.js"></script>

in the beginning of https://www.unicode.org/reports/tr35/tr35-dates.html add the id for the <h4>

see https://www.w3schools.com/html/html_id.asp
and https://www.bryanbraun.com/anchorjs/

[anba]

Ah thanks, that's explains it. With JS disabled, the "#first-day-overrides" anchor doesn't work, whereas all other anchors I've tried do work.

[anba]

UTS 35 has validity requirements which are not implemented in ICU4C.

Relevant specifications for rg and sd, including their validity requirements:

• https://unicode.org/reports/tr35/tr35.html#RegionOverride
• https://unicode.org/reports/tr35/tr35.html#UnicodeSubdivisionIdentifier
• https://unicode.org/reports/tr35/tr35.html#Unicode_Subdivision_Codes
• https://unicode.org/reports/tr35/tr35.html#Validity_Data

UTS 35 contains examples for validity checks in the "Unit preference overrides" section:

• https://unicode.org/reports/tr35/tr35-info.html#Unit_Preferences_Overrides

Is it a normative requirement to implement the validity checks when referencing UTS 35's "First Day Overrides"?

---

Results for some locales and the returned week info objects from the V8 implementation:

Locale	Week Info	Correctness	Comment
en	{firstDay: 7, weekend: [6, 7], minimalDays: 1}	Correct	Missing region, add US per likely-subtags.
en-US	{firstDay: 7, weekend: [6, 7], minimalDays: 1}	Correct	Expected week info for US.
en-DE	{firstDay: 1, weekend: [6, 7], minimalDays: 4}	Correct	Expected week info for DE.
en-DE-u-rg-uszzzz	{firstDay: 7, weekend: [6, 7], minimalDays: 1}	Correct	Region override to use US.
en-DE-u-rg-aazzzz	{firstDay: 1, weekend: [6, 7], minimalDays: 1}	Incorrect	Region AA doesn't exist, so region override is invalid and should be ignored per UTS 35.
en-DE-u-rg-001zzz	{firstDay: 1, weekend: [6, 7], minimalDays: 1}	Incorrect	001 is a macro region, so region override is invalid and should be ignored per UTS 35.

[FrankYFTang]

UTS 35 has validity requirements which are not implemented in ICU4C.

Relevant specifications for rg and sd, including their validity requirements:

• https://unicode.org/reports/tr35/tr35.html#RegionOverride
• https://unicode.org/reports/tr35/tr35.html#UnicodeSubdivisionIdentifier
• https://unicode.org/reports/tr35/tr35.html#Unicode_Subdivision_Codes
• https://unicode.org/reports/tr35/tr35.html#Validity_Data

UTS 35 contains examples for validity checks in the "Unit preference overrides" section:

• https://unicode.org/reports/tr35/tr35-info.html#Unit_Preferences_Overrides

Is it a normative requirement to implement the validity checks when referencing UTS 35's "First Day Overrides"?

Results for some locales and the returned week info objects from the V8 implementation:

v8 is on ICU 74-2 and not updated to ICU 75-1 yet. Some of the UTS35 is implemented in 75-1. I need to check are the two issues you mentioned due to 74-2 vs 75-1 or the trunk is not correct yet.

[FrankYFTang]

UTS 35 has validity requirements which are not implemented in ICU4C.

could you file an ICU ticket about what is missing? I am not quite sure I understand what is missing there now. Comparing the current v8 behavior is not a good idea since I added some of that latest UTS35 support into 75-1 but v8 is still on 74-2 so far.

[ben-allen]

I don't like putting this in a note, since it's required rather than recommended behaviour. How about defining Step 3 as something like the following:

1. Let r be a Record with fields as defined by Table 2, with field values calculated from locale per the specifications in UTS 35's Week Data and the algorithm given in First Day Overrides.

on edit: I believe the suggested text would make this a normative change rather than an editorial change.

[anba]

UTS 35 has validity requirements which are not implemented in ICU4C.

could you file an ICU ticket about what is missing? I am not quite sure I understand what is missing there now.

Sure. Filed as https://unicode-org.atlassian.net/browse/ICU-22854.

[anba]

Hmm, I just saw this in https://tc39.es/ecma402/#sec-language-tags (emphasis mine):

All structurally valid language tags are appropriate for use with the APIs defined by this specification, but implementations are not required to use Unicode Common Locale Data Repository (CLDR) data for validating them;

This looks like another problem when directly referencing to UTS 35 for week info processing, because UTS 35 does require validation. Maybe you can sidestep this issue by adding a note that validation as performed in UTS 35 is not required for ECMA-402?

[FrankYFTang]

LGTM