geoff%boulder.csnet@CSNET-RELAY.ARPA (Geoffrey Clemm) (09/18/85)
I posted this on csnet a couple of times a while back and received no significant response. It was suggested that I send it directly to info-ada@eclb. If there is a better alternative solution to the one I am using, I would be very interested in hearing about it. --------------------------------------------------------------------- From postnews Sun Jul 7 17:19:36 1985 Subject: Implementing mutually dependent access types in separate packages Newsgroups: net.lang.ada Distribution: net * Problem : I have a large program with data structures whose implementations are all mutually dependent. For example, a "FileHeader" is a structure containing a field that is a pointer to a "Parameter", a "Parameter" is a structure containing a field that is a pointer to an "Item", and an "Item" is a structure containing a field that is a pointer back to a "FileHeader". (For a simple example, see section 3.8.1 of the ADA standard). The desireable implementation would consist of one package for each data structure, where a package performs the usual data abstraction principle of providing accessing functions while hiding the physical implementation of the data structure. In the above example, the package for "FileHeader" must provide (amoung other things) the type "FileHeaderPtr" (ACCESS FileHeader) and an accessing function "FileHeader_Parameter" that takes a FileHeaderPtr as an input argument and returns a ParameterPtr. Similarly for each of the other packages. Unfortunately, I can find no way of doing this in ADA. The obvious implementation generates a "USE" cycle that is illegal according to the language standard (not to mention being rejected by our compiler). The only method I can discover for doing this in ADA is to take all of the data structure definitions out of their respective packages and then clump them together into a single "implementation package". All the individual packages then "USE" this implementation package. This is very unfortunate for a variety of reasons. Not the least of these is that unless the ADA compiler is very smart (which ours isn't), changing the implementation of any data structure causes the re-compilation of all the data structures that USE the implementation package. Question : Can anyone come up with a way of leaving the implementation of each data structure in its own package, while providing the desired accessing functions and not violating the ADA standard ? Note: If one was not required to provide accessing functions of the form "FileHeader_Parameter(FileHeaderPtr) RETURN ParameterPtr", there of course would be no problem, but these accessing functions are critical. Geoffrey Clemm
BRYAN@SU-SIERRA.ARPA (Doug Bryan) (09/19/85)
Geoffrey, Submitted for your consideration: --------------------- generic type File is limited private; package Item_Pac is type Item is limited private; function Get_File (Of_Item : Item) return File; -- other operations... private type Item is record F : File; -- other data end record; end Item_Pac; package body Item_Pac is ... end Item_Pac; --------------------- generic type Item is limited private; package Parameter_Pac is type Parameter is limited private; function Get_Item (Of_Parameter : Parameter) return Item; -- other operations... private type Parameter is record I : Item; -- other data end record; end Parameter_Pac; package body Parameter_Pac is ... end Parameter_Pac; --------------------- with Item_Pac, Parameter_Pac; package File_Pac is type File is limited private; type Access_File is access File; package Items is new Item_Pac (File => Access_File); package Parameters is new Parameter_Pac (Item => Items.Item); function Get_Parameter (Of_File : File) return Parameters.Parameter; function Get_Parameter (Of_File : Access_File) return Parameters.Parameter; -- other operations... private type File is record P : Parameters.Parameter; -- other data end record; end File_Pac; package body File_Pac is ... end File_Pac; ------------------ Now the users need simply "with" File_Pac to get the operations on all three abstract data types. The types need not be limited. We used limited types to show that this method should work for any actual implementations of the types. what think you? Doug Bryan Geoff Mendal bryan@su-sierra Mendal%UMich-MTS.Mailnet@MIT-Multics.ARPA Stanford U. Lockheed Missiles & Space Company, Inc. -------
geoff@boulder.UUCP (Geoffrey M. Clemm) (09/23/85)
In article <8509191758.AA23499@UCB-VAX.ARPA> BRYAN@SU-SIERRA.ARPA (Doug Bryan)
suggests to use generic packages to separate the parts of a mutually recursive
data type into separate packages.
I agree that generics allow one to "un-nest" packages, but one of the main
reasons I wanted to separate the mutually recursive data type into separate
packages was to provide for re-compilation efficiency. In our compiler,
generics are just treated as fancy macros (as far as I can tell), which
means that a generic solution, i.e.
GENERIC PACKAGE a ... END a;
GENERIC PACKAGE b ... END b;
PACKAGE all IS
PACKAGE my_a IS NEW a;
PACKAGE my_b IS NEW b;
...
END all;
is no better than
PACKAGE all IS
PACKAGE my_a IS ... END my_a;
PACKAGE my_b IS ... END my_b;
...
END all;
The only way that I can get re-compilation efficiency (i.e. if I modify
one part of the mutually recursive data structure, I only have to recompile
the accessing functions for that part), would be to place them in separate
non-generic packages, i.e.
PACKAGE my_a IS ... END my_a;
PACKAGE my_b IS ... END my_b;
And that's what I don't think I can do in ADA.
Geoffrey Clemm
P.S. The reason I am concerned with recompilation efficiency is that
I am doing exploratory programming with a large system, half of which
is accessing functions to data structures that all end up being mutually
recursive. If I had an efficient validated ADA interpreter I wouldn't care ...
franka@mmintl.UUCP (Frank Adams) (09/24/85)
[Not food] One possibility would be to remove the access functions from the packages for the types, and put them in a separate package of their own. I know this isn't ideal, but it may be better than the alternatives. Sorry to send this on the news, but mail didn't make it. Frank Adams ihpn4!philabs!pwa-b!mmintl!franka Multimate International 52 Oakland Ave North E. Hartford, CT 06108