Stream: implementers
Topic: FHIRPath Issue
Grahame Grieve (Mar 10 2018 at 12:24):
@Bryn Rhodes @Ewout Kramer I have a problem with GF#15650. The problem is that toInteger() fails hard, and we do not require short circuit evaluation. So all uses of toInteger() require an iif().. that is an unexpected - and therefore bad - outcome. toInteger() should fail to null not a hard error
Bryn Rhodes (Mar 10 2018 at 15:15):
@Grahame Grieve I'd be happy to relax conversion failures to return null, but we'd also then have to have some way to tell whether a conversion is valid, an isInteger() in this case
Bryn Rhodes (Mar 10 2018 at 15:19):
Tracker logged: GF#15741
Grahame Grieve (Mar 11 2018 at 04:23):
but relaxing it actually removes the requirement to have isInteger() because you can do toInteger().exists()... but isInteger() is more natural, so I am fine with that
Grahame Grieve (Mar 11 2018 at 04:51):
also, we aere unclear in the spec whether .is() and .as() are defined
Bryn Rhodes (Mar 11 2018 at 21:35):
I thought there was a tracker for this already, but I couldn't find it, so I submitted one: GF#15745
Bryn Rhodes (Mar 11 2018 at 21:37):
In looking for that, I also found several deferred trackers that I updated to triaged.
Michael Lawley (Mar 14 2018 at 00:53):
heh - I was just looking for exactly this (definition of as())
Grahame Grieve (Mar 14 2018 at 00:54):
actually, it should be .ofType() not .as()
Last updated: Apr 12 2022 at 19:14 UTC
