Path: utzoo!utgpu!jarvis.csri.toronto.edu!cs.utexas.edu!uwm.edu!bionet!ames!sgi!shinobu!odin!delrey!shap
From: shap@delrey.sgi.com (Jonathan Shapiro)
Newsgroups: comp.lang.c++
Subject: Re: unbounded static array member
Message-ID: <5104@odin.SGI.COM>
Date: 9 Mar 90 23:47:03 GMT
References: <11024@saturn.ucsc.edu>
Sender: news@odin.SGI.COM
Organization: Silicon Graphics, Inc., Mountain View, CA
Lines: 45

In article <11024@saturn.ucsc.edu> daniel@saturn.ucsc.edu (Daniel Edelson) writes:
>Is there or should there be a syntax for declaring a static array 
>member whose length is determined by the length of its initializer list?
>
>e.g.,
>
>	struct S 
>		static int array[];		// Error, bound required
>	};
>
>	int S::array[] = { 1, 2, 3, 0};

No.  Such an array is not (by definition) fixed size given only the
declaration to work from.  In C/C++, that's what pointers are for.

What IS needed is a way to define class objects that act like arrays.
These need a suitable constructor that is not now found in the
language for static initialization:

	class Foo;

	class FooArray {
		...
	public:

		FooArray[](int, const Foo *const);
	};


	FooArray myFooArray = { foo, foo, foo, foo, foo };

The claim is that we should model the array initialization syntax as a
pair of the form

	(length, pointer-to-base)

In the absence of this one cannot build things like static
bounds-checked arrays that are initialized without introducing extra
variables.

On the other hand, it's definitely a marginal problem, and this idea
is at best half-baked.

Jonathan Shapiro
Silicon Graphics