linnig@skvax1.csc.ti.com (Mike Linnig) (03/11/89)
John Gateley's (Gateley@tilde.csc.ti.com) challenge was to implement the
following lisp construct in Ada...
----------------------------------------------------------------------
-- Scheme CL
-- (define print (defun print (x)
-- (lambda (x) (typecase x
-- (cond (integer 1)
-- ((integer? x) 1) (real 2)
-- ((real? x) 2) (complex 3)
-- ((complex? x) 3) (vector 4)))
-- ((vector? x) 4))))
----------------------------------------------------------------------
package dynamic_types is
TYPE real IS NEW integer; -- not how we would really do it
TYPE complex IS NEW integer; -- but good enuf for an example.
TYPE vector IS NEW integer;
Type Dynamic_Type_options is (integer_t,real_t,complex_t,vector_t);
type Dynamic_Type(which: Dynamic_Type_options := integer_t) IS record
CASE which IS
WHEN integer_t => I: integer;
WHEN real_t => R: real;
WHEN complex_t => C: complex;
WHEN vector_t => V: vector;
END CASE;
END RECORD;
FUNCTION print(x: Dynamic_Type) return integer;
END dynamic_types;
----------------------------------------------------------------------
WITH text_io;
PACKAGE BODY dynamic_types IS
FUNCTION print(x: Dynamic_Type) return integer is
BEGIN
CASE x.which IS
WHEN integer_t => return (1);
WHEN real_t => return (2);
WHEN complex_t => return (3);
WHEN vector_t => return (4);
END CASE;
END print;
END Dynamic_Types;
----------------------------------------------------------------------
WITH dynamic_types,text_io;
USE dynamic_types;
PROCEDURE lisp_test IS
PACKAGE int_io IS NEW text_io.integer_io(integer); -- I hate this.
tmp1 : dynamic_type := (which=>real_t, r=> 3); -- init to real type
BEGIN
text_io.put("the type number of tmp1 is :");
int_io.put(print(tmp1));
text_io.new_line;
tmp1 := (which=>complex_t, c => 99);
text_io.put("the NEW type number of tmp1 is :");
int_io.put(print(tmp1));
text_io.new_line;
END lisp_test;
----------------------- SAMPLE OUTPUT ---------------------------
the type number of tmp1 is : 2
the NEW type number of tmp1 is : 3
----------------------------------------------------------------------
I think this answers John Gateley's (Gateley@tilde.csc.ti.com) dynamic
typing challenge. I didn't understand the first challenge; if it involved
something similar to The C Language construct allowing pointers to
procedures (and calling them) that can be handled too.
Mike Linnig,
Texas Instruments
Geez, we work for the same company!mark@aoa.UUCP (Som Hane) (03/20/89)
In article <8903101703.AA04518@ti.com> linnig@skvax1.csc.ti.com (Mike Linnig) writes: >John Gateley's (Gateley@tilde.csc.ti.com) challenge was to implement the >following lisp construct in Ada... > > ---------------------------------------------------------------------- >-- Scheme CL >-- (define print (defun print (x) >-- (lambda (x) (typecase x >-- (cond (integer 1) >-- ((integer? x) 1) (real 2) >-- ((real? x) 2) (complex 3) >-- ((complex? x) 3) (vector 4))) >-- ((vector? x) 4)))) > ---------------------------------------------------------------------- >package dynamic_types is > > [ Ada code deleted ] > >I think this answers John Gateley's (Gateley@tilde.csc.ti.com) dynamic >typing challenge. I didn't understand the first challenge; if it involved >something similar to The C Language construct allowing pointers to >procedures (and calling them) that can be handled too. > Let me start off by saying that I am a Lisp ( and C ) programmer, and have only been doing Ada for 1 year, so I may be totally off base, but... I think you may have missed the point. In the CL example above the type of "x" could be set or changed dynamically; in your Ada example it was ( and had to be ? ) set at compile time. In addition, in the CL example the range of choices in the typecase itself could be modified dynamically to be the members of an arbitrary set of types ( some of which could themselves have been dynamically created ). I do not see at all how your example can be modified to do any of these things... so I think the challenge has not been met. Mark Reynolds ...!BBN.COM!bbn!aoa!mark mark%aoa.uucp@BBN.COM
scott@shuksan.UUCP (Scott Moody) (03/25/89)
I think the real issue is not to make Ada into something it isn't. It isn't the greatest dynamic prototyping language, but can still be used to prototype. Aside from the Ada Guru test of providing in Ada similar dynamic capabilities that other languages offer, I think we are missing the real issue of "design" and "requirements". Once the prototyper has developed a nifty lisp/smalltalk/ai program, take the "design/requirements" that the prototype generated and code it up in your favorite type safe language (Ada). A "design/requirements" shouldn't rely on dynamic typing (but an implementation may use these feature especially in a prototype). -- Scott Moody