5<a<50 (was:Additions to C - range checking)

[Jack Jansen]

In article <155 at daisy.warwick.UUCP> cudcv at daisy.warwick.ac.uk (Rob McMahon) writes:
>In article <523 at ccird1.UUCP> rb at ccird1.UUCP (Rex Ballard) writes:
>>
>>if (5<a<50) do_something(a);
>>
>...
>>
>>Is this impossible to parse?
>
>This was in BCPL, and was a GOOD IDEA.

Yes, it is nice from a users standpoint, but if you want to define
the semantics of this operator in a reasonable way, it becomes
difficult. The only way out I see is to let expressions have *two*
values: an ordinary one, and a success/failure indication. Then you
can let the value of 'a<b' be 'b', and the operator return 'success'
when a is less than b. I remember seeing something about a language that
had such semantics, can't remember which one, however.
-- 
	Jack Jansen, jack at mcvax.UUCP
	The shell is my oyster.