[
https://issues.apache.org/jira/browse/GROOVY-4018?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15450287#comment-15450287
]
Steve Tekell commented on GROOVY-4018:
--------------------------------------
I don't see how this is covered by GROOVY-3309. Does the solution for !(NaN >
0) somehow make NaN false?
If you want to merge the tickets because it makes sense for development, that's
fine, but there needs to be a condition added there.
NaN should be false
assert !(Double.NaN)
> Make the Groovy truth value of NaN be false
> -------------------------------------------
>
> Key: GROOVY-4018
> URL: https://issues.apache.org/jira/browse/GROOVY-4018
> Project: Groovy
> Issue Type: Improvement
> Components: groovy-runtime
> Reporter: Steve Tekell
>
> Similar to null and empty-string, have Groovy evaluate NaN as false rather
> than true.
> def x = a?.b?.calcX(...)
> if (x) {
> // now I have a valid non-null number
> }
> This is how JavaScript treats NaN and like other Groovy design choices, it
> just makes sense.
> Discussed in this thread
> http://old.nabble.com/Shouldn't-the-Groovy-truth-value-of-NaN-be-false--tt27348256.html
> That is, if NaN was false the follow should work
> assert !(Double.NaN)
> assert !((Double.NaN as Boolean) == true)
> assert ((Double.NaN as Boolean) == false)
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
