Fix zh-Hant - PM "afternoon" handled correctly

[jspeaker]

Fixed zh-Hant PM recognition to include 'afternoon' as it alphabetically comes before 'am' - fixes issue where AM and PM were both parsing as the AM datetime because "afternoon" was not being recognized as PM and therefore not adding 12 hours.

[rxaviers]

Thank you for your contribution. Please, could you include a test case that passes after the fix, but fails without it?

Thanks

[jspeaker]

It appears that you use QUnit on this project, and my familiarity is mostly in the Jasmine world… it would take me a bit to get set up to run said tests, I’m supposing, and I’ve got my hands full with the work I’m actively doing (that required this fix).

I’ll flag this email and if I get “free time,” then certainly.

Also, it appeared additionally that no tests have been created in the project’s date tests that specifically switch locales. It seems that the best bet would be to refactor the date tests to run the full set of assertions based on every supported locale rather than just the currently hardcoded “en”.

Thoughts?

Cheers

Jim Speaker

From: Rafael Xavier de Souza [mailto:[email protected]]
Sent: Saturday, September 12, 2015 11:29 AM
To: jquery/globalize
Cc: Jim Speaker
Subject: Re: [globalize] Fix zh-Hant - PM "afternoon" handled correctly (#508)

Thank you for your contribution. Please, could you include a test case that passes after the fix, but fails without it?

Thanks

—
Reply to this email directly or view it on GitHub #508 (comment) . https://github.com/notifications/beacon/AAyc15TuX3U8IClFvsBnS3mZECERHSGXks5oxGZlgaJpZM4F8TNc.gif

[rxaviers]

No problem. So, please could you provide me a couple of Dates where you see this problem happening for the zh-Hant locale?

[jspeaker]

FYI, as you can see, I had trouble with the check-in comment format on my pull request, tried many times, and finally threw up my hands.

Any time with the PM characters (下午), which are the same as “afternoon,” will cause this failure, as it will be evaluated as AM.

Globalize.parseDate("下午1:00", { time: "short" });

The following test fails with the original code, and passes with the updated code.

The failure is as follows:

Expected Date(Tue Sep 15 2015 01:00:00 GMT-0700 (Pacific Daylight Time)) to equal Date(Tue Sep 15 2015 13:00:00 GMT-0700 (Pacific Daylight Time)).

Original code:

  if ( hour12 && amPm === "pm" ) {

      date.setHours( date.getHours() + 12 );

  }

Updated code:

 if ( hour12 && ( amPm === "pm" || amPm === "afternoon" ) ) {

     date.setHours( date.getHours() + 12 );

 }

And, here’s the jasmine test for this zh-Hant time problem.

describe("Given globalization date module as zh-Hant", function () {

var language = "zh-Hant", actual, expected = new Date(2015, 8, 15, 13, 0, 0, 0);

beforeEach(function (done) {

$.when(

    $.getJSON("/úæʒ_Scripts/globalization/cldr/main/zh-Hant/numbers.json"),

    $.getJSON("/úæʒ_Scripts/globalization/cldr/supplemental/numberingSystems.json"),

    $.getJSON("/úæʒ_Scripts/globalization/cldr/main/zh-Hant/ca-gregorian.json"),

    $.getJSON("/úæʒ_Scripts/globalization/cldr/supplemental/timeData.json"),

    $.getJSON("/úæʒ_Scripts/globalization/cldr/supplemental/likelySubtags.json")

  )

  .then(function () {

    return [].slice.apply(arguments, [0]).map(function (result) {

      return result[0];

    });

  })

  .then(Globalize.load)

  .then(

    function () {

      Globalize(language);

      Globalize.locale(language);

      done();

    }

  );

});

describe("When parsing time", function() {

beforeEach(function () {

  actual = Globalize.parseDate("下午1:00", { time: "short" });

});



it("Then it should equal todays date with the correct PM time", function () {

  expect(actual).toEqual(expected);

});

});

});

Sorry I don’t have the time to dig into QUnit. I’m sure it’s no big deal, but I need to get back to work now.

Oh, and by the way, I absolutely love the Globalize library. I’m using globalize.validate with jquery.validate, and incorporated globalize into some of my custom validators – it’s a God send for the multi-lingual web application I’m working on. The timing of the 1.0 release was right on the money with my need for it.

Thank you very much!

Jim Speaker

From: Rafael Xavier de Souza [mailto:[email protected]]
Sent: Tuesday, September 15, 2015 3:24 AM
To: jquery/globalize
Cc: Jim Speaker
Subject: Re: [globalize] Fix zh-Hant - PM "afternoon" handled correctly (#508)

No problem. So, please could you provide me a couple of Dates where you see this problem happening for the zh-Hant locale?

—
Reply to this email directly or view it on GitHub #508 (comment) . https://github.com/notifications/beacon/AAyc1_igKf4yGZrjKkTam9JkBd1MG6Voks5ox-k6gaJpZM4F8TNc.gif

[jspeaker]

Oh, one more thing, this will also fix the same problem with zh (aka zh-Hans). The "afternoon" characters exist in both, though they are ofc different characters in zh-Hans.

[rxaviers]

Any time with the PM characters (下午), which are the same as “afternoon,” will cause this failure, as it will be evaluated as AM.

Excellent, that's what I needed. Thanks. Now I need to find time to add the unit test, so we merge this. Alternatively, if you want to give it a try, the unit tests should go here and we'll need to load data for the zh-Hant data analogous to this code and create an zhHant instance analogous to this code.

[rxaviers]

@jspeaker the implementation unfortunately isn't correct and needs considerable additions to behave correctly. Reviewing this, I found out that:

• afternoon is locale-dependent, i.e., the definition changes from locale to locale;
• afternoon is not the only keyword, there are also evening, midnight, night, and noon. Also they are numbered and occasionally there are two variants like afternoon1 and afternoon2.
• Not all afternoonss means it's +12. For example, for uz afternoon means it's from 11am and on (link). Therefore, assuming we can always +12 when we find afternoon is wrong.

Ref: http://www.unicode.org/reports/tr35/tr35-dates.html#Day_Period_Rule_Sets