isnum("%pi") returns %T, but isnum("%e") returns %F, and isnum("%pi+1") returns (and strtod() policy)
Reported by Samuel GOUGEON (@sgougeon)
Originally assigned to Paul BIGNIER
-- Bug description --
isnum("%pi") returns %T, but isnum("%e") returns %F, and isnum("%pi+1") returns also %F. A clear policy should be set and kept: Either the argument is always interpreted (evstr()) before being num-tested, or it is never interpreted.
Else, why do Neper should loose with his %e against Archimedes with his %pi ?
And what about isnum("%i") for Cardan, Ferrari & al. ?
-- Scilab error message --
none
-- How to reproduce the bug --
--> help isnum
--> a=isnum("%pi"), b=isnum("pi+1"), c=isnum("%e")