The point is; If I run npm run extract the language which is default(en) has all messages "Missing" (because it is non-sense to give translation to these as well if that is the source-of-truth).
Catalog statistics is used to determine whether there are some non-translated messages therefore if You use it for any kind of automation (as a test) it wouldn't pass unless you translate the source messages as well.
I am asking if according to your opinion is this good practice or I am missing some point where you simply set "sourceLocale": "en" and problem solves? (because this isn't included in your, nor official example)
I didn't understand your question. Can you provide code example, because it seems you use different terminology
I am trying your example code (github.com/stereobooster/react-lin...).
The point is; If I run
npm run extract
the language which is default(en) has all messages "Missing" (because it is non-sense to give translation to these as well if that is the source-of-truth).Yes they missing, because there is no
"sourceLocale": "en"
, otherwise they would be there. What is your question?Catalog statistics is used to determine whether there are some non-translated messages therefore if You use it for any kind of automation (as a test) it wouldn't pass unless you translate the source messages as well.
I am asking if according to your opinion is this good practice or I am missing some point where you simply set
"sourceLocale": "en"
and problem solves? (because this isn't included in your, nor official example)No
"sourceLocale": "en"
is not a solution. You can ask author for this feature