change to return false when a nullable argument is passed

[raccoonback]

I have used i18next, react-i18next.
However, when using the i18n object to determine the existence of the key, I found that passing one nullish factor(undefined or null, or '') results in the error below:

TypeError
Cannot read property 'forEach' of undefined

I think i18n.exist(undefined or null, or '') should return false.

Similar to mine, there will be many cases that deliver nullish factors, and I think it is obvious that nullish factors do not exist as keys in the resource.

Checklist

• only relevant code is changed (make a diff before you submit the PR)
• run tests npm run test
• tests are included

Checklist (for documentation change)

• only relevant documentation part is changed (make a diff before you submit the PR)
• motivation/reason is provided

[coveralls]

Coverage increased (+0.01%) to 93.216% when pulling f9f3817 on raccoonback:check-nullable-args-in-translator-exist into b371a5c on i18next:master.

[jamuhl]

wouldn't it be cleaner to add a check in the exists function itself than in the reused resolve? Like it's already done for the translate here: https://github.com/i18next/i18next/blob/master/src/Translator.js#L87

[raccoonback]

@jamuhl

I like your opinion,
but wouldn't it be better to deal with common within the resolve() function because this error is from the resolve() function?

because nullable values(undefined or null, or '') can be passed to keys parameter elsewhere using the resolve() function`.

[jamuhl]

resolve is internal...currently translate, exist are "public" functions - as translate handles the return value for nullish already it seems logical for me to let exists handle it itself too

[raccoonback]

@jamuhl
Yes, I understand. I'll change it.

[raccoonback]

@jamuhl
I've reflected all the reviews. Please check it again.

[raccoonback]

@jamuhl
I've reflected the your review. Please check it one more. 😃

[adrai]

lgtm

[jamuhl]

lgtm

[jamuhl]

Guess this could be released with next version...?