feat(format-po-gettext): respect Plural-Forms header #2010
Conversation
|
The latest updates on your projects. Learn more about Vercel for Git ↗︎
|
size-limit report 📦
|
|
Might have to warn or hard error on msgid ""
msgstr ""
"Language: fr\\n"
"Plural-Forms: nplurals=3; plural=(n == 0 || n == 1) ? 0 : n != 0 && n % 1000000 == 0 ? 1 : 2;\\n"
#. js-lingui:icu=%7B0%2C+plural%2C+one+%7B%7Bcount%7D+day%7D+other+%7B%7Bcount%7D+days%7D%7D&pluralize_on=0
msgid "{count} day"
msgid_plural "{count} days"
msgstr[0] "{count} jour"
msgstr[1] "{count} jours"will result in runtime errors: |
|
Also some uncaught runtime errors if the locale isn't found in the cldr plurals lib. Probably needs a cleaner error message, and also specifically for |
|
@garikkh could you take it over and finish this PR? |
|
Most likely I will have some time this week to take over this PR. |
|
Had a few fixes here: timofei-iatsenko/js-lingui@feature/respect-plural-forms-header...garikkh:js-lingui:feature/respect-plural-forms-header tested the PO files my TMS generates, looks good. One thing to note is that because the library uses Another note: I considered using either https://www.npmjs.com/package/cldr or https://www.npmjs.com/package/cldr-data instead of having to copy/paste a json file, but both were just as silly to use for this case. Do you want me to make a new PR to lingui or to your branch? |
|
@timofei-iatsenko @garikkh is there anything we can do to help move this PR forward?
💯 I also feel that committing these rules in raw format might not be the best approach. It might be more efficient and manageable to use a dependency for this instead. |
|
@andrii-bodnar I linked a PR I made with @timofei-iatsenko fork as the base branch, just so it was a tiny bit neater. Wasn't sure if you guys wanted a separate PR.
Agreed but the one being used here doesn't export the JSON and the maintainer doesn't want to add it to their exports. I don't remember the exact issues with the other two, so I could check again. |
|
More details: cldr-data is a great repository, but the plurals data is difficult to parse. each language has plurals defined like this: "cs": {
"pluralRule-count-one": "i = 1 and v = 0 @integer 1",
"pluralRule-count-few": "i = 2..4 and v = 0 @integer 2~4",
"pluralRule-count-many": "v != 0 @decimal 0.0~1.5, 10.0, 100.0, 1000.0, 10000.0, 100000.0, 1000000.0, …",
"pluralRule-count-other": " @integer 0, 5~19, 100, 1000, 10000, 100000, 1000000, …"
},And not exactly great to parse and extract actual samples from.
> cldr.extractPluralClasses("cs")
[ 'one', 'few', 'many', 'other' ]
> cldr.extractPluralRuleFunction("cs")(1)
'one'
> cldr.extractPluralRuleFunction("cs")(100)
'other'But other than that only provides raw xml for data: <pluralRules locales="cs sk">
<pluralRule count="one">i = 1 and v = 0 @integer 1</pluralRule>
<pluralRule count="few">i = 2..4 and v = 0 @integer 2~4</pluralRule>
<pluralRule count="many">v != 0 @decimal 0.0~1.5, 10.0, 100.0, 1000.0, 10000.0, 100000.0, 1000000.0, …</pluralRule>
<pluralRule count="other"> @integer 0, 5~19, 100, 1000, 10000, 100000, 1000000, …</pluralRule>
</pluralRules>And for the forms: <pluralRanges locales="cs pl sk">
<pluralRange start="one" end="few" result="few"/>
<pluralRange start="one" end="many" result="many"/>
<pluralRange start="one" end="other" result="other"/>
<pluralRange start="few" end="few" result="few"/>
<pluralRange start="few" end="many" result="many"/>
<pluralRange start="few" end="other" result="other"/>
<pluralRange start="many" end="one" result="one"/>
<pluralRange start="many" end="few" result="few"/>
<pluralRange start="many" end="many" result="many"/>
<pluralRange start="many" end="other" result="other"/>
<pluralRange start="other" end="one" result="one"/>
<pluralRange start="other" end="few" result="few"/>
<pluralRange start="other" end="many" result="many"/>
<pluralRange start="other" end="other" result="other"/>
</pluralRanges> |
Description
Types of changes
Fixes # (issue)
Checklist