Support multi locales in one translation file

[shu-mutou]

zh is same as zh-cn, so maintain zh and change zh-cn as link to zh.

Also, added support for i18n/locales_not_for_build_local that keeps locales not for build as local setting for developer. Because building dashboard takes much time.

Ref: #4734 comment
Ref: #4763

[codecov]

Codecov Report

Merging #4814 into master will decrease coverage by 0.01%.
The diff coverage is n/a.

@@            Coverage Diff             @@
##           master    #4814      +/-   ##
==========================================
- Coverage   45.43%   45.42%   -0.02%     
==========================================
  Files         214      214              
  Lines        9977     9977              
  Branches       94       94              
==========================================
- Hits         4533     4532       -1     
- Misses       5180     5181       +1     
  Partials      264      264

Impacted Files	Coverage Δ
...p/backend/integration/metric/common/aggregation.go	89.09% <0%> (-1.82%)	⬇️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 998f3aa...f093ae8. Read the comment docs.

[floreks]

I am not sure if creating a symbolic link is a correct way of doing this. Every translation file in the header has information about target-language

dashboard/i18n/zh-cn/messages.zh-cn.xlf

Line 3 in 14ae7c1

<file source-language="en" datatype="plaintext" original="ng2.template" target-language="zh-cn">

If we create a symbolic link to a different language code, then target-language will not be correct.

[shu-mutou]

We can use zh-Hans representing Simplified Chinese and and zh-Hant for Traditional Chinese.
http://www.iana.org/assignments/lang-tags/zh-Hans
http://www.iana.org/assignments/lang-tags/zh-Hant

Firefox and Chrome don't seem to have zh-Hans and zh-Hant, but we can create them for translation work and link to them from other zh*s.

It has inconsistent in target-language but reasonable if do so, I think. npm builds dashboard for each languages with symbolic linked messages.xx.xlf file and the binary works perfectly as expected.

I'll investigate how not to build zh-hans and zh-hant that won't be used, and update document.

/hold

[shu-mutou]

/hold cancel

[hwdef]

This may not be the right way.
Although both mainland China and Singapore use simplified Chinese, some words are different.
Although Taiwan and Hong Kong both use traditional Chinese, they are also very different.
So we should use different files for each region

[shu-mutou]

/hold
@hwdef I see situations for zh locales. I'll revert zh* reconstructions. Thanks.

[shu-mutou]

@hwdef, how about zh and zh-cn ? Are there deference between them?

[hwdef]

@shu-mutou
zh represents all kinds of Chinese, it is better to make zh consistent with zh-cn, because zh-cn is the most used.

[shu-mutou]

@hwdef I got it. I'll make zh-cn as link to zh.

[shu-mutou]

@floreks Changes for this PR are coming back to first situation and inconsistency in target-language is back. But zh means "all kinds of Chinese" and its contents is same as zh-cn, as hwdef said, so it's reasonable doing so, I think.

[shu-mutou]

/hold cancel

[floreks]

I am not fully sure if keeping target locale that is different than configured locale code will not cause some issues during compiling the translated versions. Having multiple overlapping target languages might cause issues.

[shu-mutou]

@floreks No worry, I think. Dashboard is built with ng build --i18nLocale = [locale] and the Angular documentation says that the locale specified by this option will be set to the LOCALE_ID of the compiled file.
https://github.com/angular/angular/blob/master/aio/content/guide/i18n.md#i18n-pipes

In fact, LOCALE_ID in compiled JavaScript files for zh-cn that created as a symbolic link tozh in this PR is zh-cn instead of zh.
As same as zh-cn, compiled files for ja or fr that do not have target-language in xlf files have each LOCALE_ID specified with --i18nLocale option.

[floreks]

@shu-mutou makes sense.

/lgtm

[k8s-ci-robot]

[APPROVALNOTIFIER] This PR is APPROVED

This pull-request has been approved by: floreks, shu-mutou

The full list of commands accepted by this bot can be found here.

The pull request process is described here

Needs approval from an approver in each of these files:

• OWNERS [floreks,shu-mutou]

Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment