<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 3.2//EN">
<!-- $Id: tutorial.html,v 1.8 2001/03/09 13:21:23 neumann Exp $ -->
<HTML>
<HEAD>
<TITLE>XOTcl - Tutorial</TITLE>
<META NAME="AUTHOR" CONTENT="Gustaf Neumann and Uwe Zdun">
<META NAME="DOCNUMBER" CONTENT="1.1.0">
<META NAME="CREATED" CONTENT="19991111;20552300">
<META NAME="CHANGEDBY" CONTENT="Gustaf ">
<META NAME="CHANGED" CONTENT="19991111;22461500">
</HEAD>
<BODY BGCOLOR="#ffffff">
<TABLE COLS=2 WIDTH=100% BORDER=0 CELLPADDING=2 CELLSPACING=0 BGCOLOR="#000055" STYLE="page-break-before: always">
<TR>
<TD WIDTH=75%>
<P><FONT COLOR="#ffffff"><FONT FACE="Arial, Helvetica"><FONT SIZE=6>XOTcl
- Tutorial - Index </FONT></FONT></FONT>
</P>
</TD>
<TD>
<IMG SRC="logo-100.jpg" NAME="Graphic1" ALIGN=RIGHT WIDTH=102 HEIGHT=42 BORDER=0></TD>
</TR>
</TABLE>
<p align=right>Version: 1.1.0</p>
<UL>
<LI><P STYLE="margin-bottom: 0in"><A HREF="#introduction">Introduction
</A>
</P>
<UL>
<LI><P STYLE="margin-bottom: 0in"><A HREF="#langOverview">Language Overview </A>
</P>
<LI><P STYLE="margin-bottom: 0in"><A HREF="#soccerClub"> Introductory Overview Example: Soccer Club</A>
</P>
</UL>
<LI><P STYLE="margin-bottom: 0in"><A HREF="#object_class_system">Object
and Class System </A>
</P>
<LI><P STYLE="margin-bottom: 0in"><A HREF="#basic">Basic
Functionalities</A>
</P>
<UL>
<LI><P STYLE="margin-bottom: 0in"><A HREF="#object">Objects </A>
</P>
<UL>
<LI><P STYLE="margin-bottom: 0in"><A HREF="#data_on_obj">Data on
objects </A>
</P>
<LI><P STYLE="margin-bottom: 0in"><A HREF="#obj_methods">Methods
on Objects</A>
</P>
<LI><P STYLE="margin-bottom: 0in"><A HREF="#obj_info">Information
on Objects</A>
</P>
</UL>
<LI><P STYLE="margin-bottom: 0in"><A HREF="#classes">Classes </A>
</P>
<UL>
<LI><P STYLE="margin-bottom: 0in"><A HREF="#class_instance">Creating
Classes and deriving Instances</A>
</P>
<LI><P STYLE="margin-bottom: 0in"><A HREF="#class_methods">Methods
in Classes</A>
</P>
<LI><P STYLE="margin-bottom: 0in"><A HREF="#class_info">Information
on Classes</A>
</P>
<LI><P STYLE="margin-bottom: 0in"><A HREF="#class_inheritance">Inheritance</A>
</P>
<LI><P STYLE="margin-bottom: 0in"><A HREF="#class_destroy">Destruction
of Classes</A>
</P>
<LI><P STYLE="margin-bottom: 0in"><A HREF="#class_method_chaining">Method
Chaining</A>
</P>
</UL>
<LI><P STYLE="margin-bottom: 0in"><A HREF="#class_dynamics">Dynamic
Class and Superclass Relationships</A>
</P>
<LI><P STYLE="margin-bottom: 0in"><A HREF="#meta-classes">Meta-Classes</A>
</P>
<LI><P STYLE="margin-bottom: 0in"><A HREF="#destroy-logic">Create, Destroy, and Recreate Scheme</A>
</P>
</UL>
<LI><P STYLE="margin-bottom: 0in"><A HREF="#interceptors">Message
Interception Techniques</A></P>
<UL>
<LI><P STYLE="margin-bottom: 0in"><A HREF="#filter">Filters</A>
</P>
<LI><P STYLE="margin-bottom: 0in"><A HREF="#mixins">Mixin Classes</A>
</P>
<LI><P STYLE="margin-bottom: 0in"><A HREF="#callstack_info">Callstack Information</A>
</P>
<LI><P STYLE="margin-bottom: 0in"><A HREF="#precedence order">Precedence Order</A>
</UL>
<LI><P STYLE="margin-bottom: 0in"><A HREF="#nesting">Class Nesting
and Dynamic Object Aggregations</A>
</P>
<LI><P STYLE="margin-bottom: 0in"><A HREF="#assertions">Assertions</A>
</P>
<LI><P STYLE="margin-bottom: 0in"><A HREF="#additional-functionalities">Additional
Functionalities</A>
</P>
<UL>
<LI><P STYLE="margin-bottom: 0in"><A HREF="#abstract-classes">Abstract
Classes</A>
</P>
<LI><P STYLE="margin-bottom: 0in" ><A HREF="#parameter">Parameter</A>
</P>
<LI><P STYLE="margin-bottom: 0in"><A HREF="#autonames">Automatic Name Creation</A>
</P>
<LI><P STYLE="margin-bottom: 0in"><A HREF="#meta-data">Meta-Data</A>
</P>
</UL>
<LI><P STYLE="margin-bottom: 0in"><A HREF="#cext">Integrating XOTcl Programs with C Extensions (such as TK)</A>
<LI><P STYLE="margin-bottom: 0in"><A HREF="#references">References</A>
</P>
</UL>
<!-- PAGE BREAK -->
<TABLE COLS=2 WIDTH=100% BORDER=0 CELLPADDING=2 CELLSPACING=0 BGCOLOR="#000055">
<TR>
<TD WIDTH=75%>
<P><A NAME="introduction"></A><FONT COLOR="#ffffff"><FONT FACE="Arial, Helvetica"><FONT SIZE=6>Introduction
</FONT></FONT></FONT>
</P>
</TD>
<TD>
<IMG SRC="logo-100.jpg" NAME="Graphic2" ALIGN=RIGHT WIDTH=102 HEIGHT=42 BORDER=0></TD>
</TR>
</TABLE>
<H2><A NAME="langOverview"></A> <BR>Language Overview
</H2>
<P>XOTcl <a href="#xotcl">[Neumann and Zdun 2000a]</a> is an extension
to the object-oriented scripting language OTcl <a
href="#otcl">[Wetherall and Lindblad 1995]</a> which itself extends
Tcl <a href="#tcl">[Ousterhout 1990]</a> (Tool Command Language) with
object-orientation. XOTcl is a value-added replacement for OTcl and
does not require OTcl to compile.</P>
<P>XOTcl runs in the <TT>tclsh</TT> and provides a few extension
commands. These are offered in a Tcl namespace ::xotcl, and need to be
imported into the current namespace. All Tcl-commands
remain available (and are also applicable on the extension
constructs).
</P>
<P>A central property of Tcl is, that it uses strings solely for the
representation of data. Internally it uses an dynamic type system with
automatic conversion (which enables efficient type handling). For
that reason all components (e.g. written in C) once integrated in Tcl
automatically fit together and the components can be reused in
unpredicted situations without change. The evolving <EM>component
frameworks</EM> provide a high degree of code reuse, rapid
application development, and ease of use. The application developer
may concentrate on the application task solely, rather than investing
efforts in fitting components together. Therefore, in certain
applications scripting languages like Tcl are very useful for a fast
and high-quality development of software (see <a
href="#ousterhout">[Ousterhout 1998]</a> for more details).
</P>
<P>Tcl is equipped with appropriate functionalities for the easy
gluing of components, like dynamic typing, dynamic extensibility, and
read/write introspection. OTcl is an object-oriented extension to Tcl,
which encourages a Tcl-like programming style and is composed of
language constructs with properties similar to Tcl. It offers an
object-orientation with encapsulation of data and operation without
protection mechanisms and single and multiple inheritance.
Furthermore it enables to change the relationships dynamically, offers
read/write introspection, has a three level class system based on
meta-classes and offers method chaining. These abilities are
integrated in XOTcl with only slight changes to OTcl visible to the
programmer.
</P>
<P>The XOTcl extension aims at complexity and adaptability issues that
may occur in context of large (object-oriented) software structures
and in the context of component glueing. In particular we added the
following support:
</P>
<UL>
<LI><P STYLE="margin-bottom: 0in"><I>Filters</I> as a means of
abstractions over method invocations to implement large program
structures, like design patterns.
</P>
<LI><P STYLE="margin-bottom: 0in"><I>Mixin Classes</I>, as a
means to give an object or a classes' instances access to several different supplemental
classes, which may be changed dynamically.
</P>
<LI><P STYLE="margin-bottom: 0in"><I>Dynamic Object Aggregations</I>,
to provide dynamic aggregations through nested namespaces.
</P>
<LI><P STYLE="margin-bottom: 0in"><I>Nested Classes</I>, to reduce
the interference of independently developed program structures.
</P>
<LI><P STYLE="margin-bottom: 0in"><I>Assertions</I>, to reduce the
interface and the reliability problems caused by dynamic typing and,
therefore, to ease the combination of components.
</P>
<LI><P><I>Meta-data and Automatic Documentation</I>, to enhance self-documentation of objects
and classes.
</P>
</UL>
<!-- PAGE BREAK -->
<P STYLE="margin-bottom: 0in"><BR>
</P>
<P ALIGN=CENTER STYLE="margin-bottom: 0in"><A NAME="features"></A><A NAME="1176"></A>
  
</P>
<P ALIGN=CENTER STYLE="margin-bottom: 0in"><STRONG>Figure 1:</STRONG>
Language Extensions of XOTcl
</P>
<CENTER>
<TABLE WIDTH=455 BORDER=0 CELLPADDING=2 CELLSPACING=0>
<COL WIDTH=451>
<TR>
<TD WIDTH=451>
<P><IMG SRC="features.gif" NAME="Graphic3" ALIGN=BOTTOM WIDTH=451 HEIGHT=378 BORDER=0></P>
</TD>
</TR>
</TABLE>
</CENTER>
<H2><A NAME="soccerClub"></A> <BR>Introductory Overview Example: Soccer Club
</H2>
<p>
To give you an impression of the language before we go into the
details of the language construct, we present in this section a
simple, introductory example. It shall demonstrate the basic
language constructs on the example of a soccer club (the full code can
be found in the <tt>xotcl/src/scripts/soccerClub.xotcl</tt> file. All
the characters in this example are fictitious, and any resemblance to
actual persons, living or deceased, is coincidental.
</p>
<p>
In XOTcl we do not have to provide a file description as a comment,
but we can use the <tt>@</tt> object, which is used generally to
provide any kind of information, meta-data, and documentation on a
running program. Here, we just give a file description. Then the <tt>
makeDoc.xotcl</tt> tool can automatically document the program file for
us.
</p>
<PRE STYLE="margin-bottom: 0.2in">
@ @File {
description {
This is a simple introductory example for the language XOTcl.
It demonstrates the basic language constructs on the example of
a soccer club.
}
}
</PRE>
<p>
All things and entities in XOTcl are objects, a special kind of objects
are classes. These define common properties for other objects. For a
soccer club, we firstly require a common class for all kinds of members.
</p>
<p>
Common to all members is that they have a name. Common properties
defined across all instances of a class are called 'parameter' in
XOTcl. In this example the instance variable <tt>name</tt> will be
initialized by default with an empty string.
</p>
<PRE STYLE="margin-bottom: 0.2in">
Class ClubMember -parameter {{name ""}}
</pre>
<p>
A special club member is a <tt>Player</tt>. Derived classes can be
build with inheritance (specified through
<tt>superclass</tt>). Players may have a <tt>playerRole</tt> (defaults
to NONE).
</p>
<PRE STYLE="margin-bottom: 0.2in">
Class Player -superclass ClubMember -parameter {{playerRole NONE}}
</pre>
<p>
Other club member types are trainers, player-trainers, and presidents:
</p>
<PRE STYLE="margin-bottom: 0.2in">
Class Trainer -superclass ClubMember
Class President -superclass ClubMember
</pre>
<p>
The PlayerTrainer uses multiple inheritances by being both a player
and a trainer:
</p>
<PRE STYLE="margin-bottom: 0.2in">
Class PlayerTrainer -superclass {Player Trainer}
</pre>
<p>
Now we define the SoccerTeam class:
</p>
<PRE STYLE="margin-bottom: 0.2in">
Class SoccerTeam -parameter {name location type}
</pre>
<p>
We may add a player. This is done by a method. Instance methods are
in XOTcl defined with <tt>instproc</tt>. All club members are aggregated in
the team (denoted by :: namespace syntax).
</p>
<PRE STYLE="margin-bottom: 0.2in">
SoccerTeam instproc newPlayer args {
# we use a unique autoname for the object to prevent name
# collisions, like <soccerteam>::player01, <soccerteam>::player02, ...
eval Player [self]::[my autoname player%02d] $args
}
</pre>
<p>
A player can be transfered to another team. The player object does
not change internally (e.g. the playerRole stays the same). Therefore we
<tt>move</tt> it to the destination team.
</p>
<PRE STYLE="margin-bottom: 0.2in">
SoccerTeam instproc transferPlayer {playername destinationTeam} {
# We use the aggregation introspection option <tt>children</tt> in order
# to get all club members
foreach player [my info children] {
# But we only remove matching playernames of type "Player". We do
# not want to remove another club member type who has the same
# name.
if {[$player istype Player] && [$player name] == $playername} {
# We simply 'move' the player object to the destination team.
# Again we use a unique autoname in the new scope
$player move [set destinationTeam]::[$destinationTeam autoname player%02d]
}
}
}
</pre>
<p>
Finally we define two convenience to print the members/players to
the stdout with puts.
</p>
<PRE STYLE="margin-bottom: 0.2in">
SoccerTeam instproc printMembers {} {
puts "Members of [my name]:"
foreach m [my info children] {puts " [$m name]"}
}
SoccerTeam instproc printPlayers {} {
puts "Players of [my name]:"
foreach m [my info children] {
if {[$m istype Player]} {puts " [$m name]"}
}
}
</pre>
<p>
Now let us build to example soccer team objects.
</p>
<PRE STYLE="margin-bottom: 0.2in">
SoccerTeam chelsea -name "Chelsea FC" -location "Chelsea"
SoccerTeam bayernMunich -name "F.C. Bayern München" -location "Munich"
</pre>
<p>
With <tt>addPlayer</tt> we can create new aggregated player objects
<p></p>
Let us start some years in the past, when "Franz Beckenbauer" was
still a player.
</p>
<PRE STYLE="margin-bottom: 0.2in">
set fb [bayernMunich newPlayer -name "Franz Beckenbauer" \
-playerRole PLAYER]
</pre>
<p>
<tt>playerRole</tt> may not take any value. It may either be NONE, PLAYER,
or GOALY ... such rules may be given as assertions (here: an instinvar
gives an invariant covering all instances of a class). In XOTcl
the rules are syntactically identical to <tt>if</tt> statements:
</p>
<PRE STYLE="margin-bottom: 0.2in">
Player instinvar {
{[my playerRole] == "NONE" ||
[my playerRole] == "PLAYER" ||
[my playerRole] == "GOALY"}
}
</pre>
<p>
If we break the invariant and turn assertions checking on, we should
get an error message:
</p>
<PRE STYLE="margin-bottom: 0.2in">
$fb check all
if {[catch {$fb set playerRole SINGER} errMsg]} {
puts "CATCHED EXCEPTION: playerRole has either to be NONE, PLAYER, or TRAINER"
# turn assertion checking off again and reset to PLAYER
$fb check {}
$fb set playerRole PLAYER
}
</pre>
<p>
But soccer players may play quite different, orthogonal
roles. E.g. Franz Beckenbauer was also a singer (a remarkably bad
one). However, we can not simply add such orthogonal, extrinsic
extensions with multiple inheritance or delegation. Otherwise we
would have either to build a lot of unnecessary helper classes, like
PlayerSinger, PlayerTrainerSinger, etc., or we would have to build
such helper objects. This either leads to an unwanted combinatorial
explosion of class or object number
</p><p>
Here we can use a per-object mixin, which is a language construct
that expresses that a class is used as a role or as an extrinsic
extension to an object.
</p><p>
First we just define the Singer class.
</p>
<PRE STYLE="margin-bottom: 0.2in">
Class Singer
Singer instproc sing text {
puts "[my name] sings: $text, lala."
}
</pre>
<p>
Now we register this class as a per-object mixin on the player object:
</p>
<PRE STYLE="margin-bottom: 0.2in">
$fb mixin Singer
</pre>
<p>
And now Franz Beckenbauer is able to sing:
</p>
<PRE STYLE="margin-bottom: 0.2in">
$fb sing "lali"
</pre>
<p>
But Franz Beckenbauer has already retired. When a player retires, we
have an intrinsic change of the classification. He *is* not a player
anymore. But still he has the same name, is club member, and
is a singer (brrrrrr).
</p><p>
Before we perform the class change, we extend the Player class to
support it. I.e. the playerRole is not valid after class change
anymore (we unset the instance variable).
</p>
<PRE STYLE="margin-bottom: 0.2in">
Player instproc class args {
my unset playerRole
next
}
</pre>
<p>
Now we can re-class the player object to its new class (now Franz
Beckenbauer is President of Bayern Munich.
</p>
<PRE STYLE="margin-bottom: 0.2in">
$fb class President
# Check that the playerRole isn't there anymore.
if {[catch {$fb set playerRole} errMsg]} {
puts "CATCHED EXCEPTION: The player role doesn't exist anymore \
(as it should be after the class change)"
}
</pre>
<p>
But still Franz Beckenbauer can entertain us with what he believes
is singing:
</p>
<PRE STYLE="margin-bottom: 0.2in">
$fb sing "lali"
</pre>
<p>
Now we define some new players for Bayern Munich:
</p>
<PRE STYLE="margin-bottom: 0.2in">
bayernMunich newPlayer -name "Oliver Kahn" -playerRole GOALY
bayernMunich newPlayer -name "Giovanne Elber" -playerRole PLAYER
</pre>
<p>
If we enlist the players of Munich Franz Beckenbauer is not enlisted
anymore:
</p>
<PRE STYLE="margin-bottom: 0.2in">
bayernMunich printPlayers
</pre>
<p>
But as a president he still appears in the list of members:
</p>
<PRE STYLE="margin-bottom: 0.2in">
bayernMunich printMembers
</pre>
<p>
Now consider an orthonogal extension of a transfer list. Every
transfer in the system should be notified. But since the transfer
list is orthogonal to SoccerTeams we do not want to interfere with
the existing implementation at all. Moreover, the targeted kind of
extension has also to work on all subclasses of SoccerTeam. Firstly, we
just create the extension as an ordinary class:
</p>
<PRE STYLE="margin-bottom: 0.2in">
Class TransferObserver
TransferObserver instproc transferPlayer {pname destinationTeam} {
puts "Player '$pname' is transfered to Team '[$destinationTeam name]'"
next
}
</pre>
<p>
Now we can apply the class as a per-class mixin, which functions
exactly like a per-object mixin, but on all instances of a class and
its subclasses. The <tt>next</tt> primitive ensures that the original
method on <tt>SoccerTeam</tt> is called after notifying the transfer (with
puts to stdout):
</p>
<PRE STYLE="margin-bottom: 0.2in">
SoccerTeam instmixin TransferObserver
</pre>
<p>
If we perform a transfer of one of the players, he is moved to the new
club and the transfer is reported to the stdout:
</p>
<PRE STYLE="margin-bottom: 0.2in">
bayernMunich transferPlayer "Giovanne Elber" chelsea
</pre>
<p>
Finally we verify the transfer by printing the players:
</p>
<PRE STYLE="margin-bottom: 0.2in">
chelsea printPlayers
bayernMunich printPlayers
</pre>
<p>
<P><BR><BR>
</P>
<TABLE COLS=2 WIDTH=100% BORDER=0 CELLPADDING=2 CELLSPACING=0 BGCOLOR="#000055">
<TR>
<TD WIDTH=75%>
<P><A NAME="object_class_system"></A><FONT COLOR="#ffffff"><FONT FACE="Arial, Helvetica"><FONT SIZE=6>Object
and Class System </FONT></FONT></FONT>
</P>
</TD>
<TD>
<IMG SRC="logo-100.jpg" NAME="Graphic4" ALIGN=RIGHT WIDTH=102 HEIGHT=42 BORDER=0></TD>
</TR>
</TABLE>
<P>In XOTcl every object is associated with a class over the <TT>class</TT>
relationship. Classes are special objects with the purpose of
managing other objects. ``Managing'' means that a class controls the
creation and destruction of its instances and that it contains a
repository of methods (``instprocs'') accessible for the instances.
Object-specific methods are called ``procs'', instance methods are
called ``instprocs''.
</P>
<P>The instance methods common to all objects are defined in the root
class <TT>Object</TT> (predefined or user-defined). Since a class is a
special (managing) kind of object it is managed itself by a special
class called ``meta-class'' (which manages itself). One interesting
aspect of meta-classes is that by providing a constructor
pre-configured classes can be derived. New user-defined meta-classes
can be derived from the predefined meta-class <TT>Class</TT> in order
to restrict or enhance the abilities of the classes that they manage.
Therefore meta-classes can be used to instantiate large program
structures, like some design patterns (see <a
href="#xotcl-filter">[Neumann and Zdun 1999a]</a> for more
details). The meta-class may hold the generic parts of the
structures. Since a meta-class is an entity of the program, it is
possible to collect these in pattern libraries for later reuse easily.
</P>
<P>XOTcl supports single and multiple inheritance. Classes are ordered
by the relationship <TT>superclass</TT> in a directed acyclic
graph. The root of the class hierarchy is the class <TT>Object</TT>.
A single object can be instantiated directly from this class. An
inherent problem of multiple inheritance is the problem of name
resolution, when for example two super-classes contain an instance
method with the same name. XOTcl provides an intuitive and unambiguous
approach for name resolution by defining the precedence order along a
linear ``<EM>next-path</EM>'' incorporating the class and mixin
hierarchies, which is modeled after CLOS. A method can invoke
explicitly the shadowed methods by the predefined command
<TT>next</TT>. When this command is executed a shadowed method is
``mixed into'' the execution of the current method. Method chaining
without explicit naming of the targeted method is very important for
languages supporting a dynamic class system, because one cannot always
predict which classes are currently participating in the inheritance
hierarchy at design time (often necessary in inheritance models, like
C++).
</P>
<P STYLE="margin-bottom: 0in">An important feature of all XOTcl
objects is the read/write introspection. The reading introspection
abilities of XOTcl are packed compactly into the <TT>info</TT>
instance method which is available for objects and classes. All
obtained information can be changed at run-time dynamically with
immediate effect. Unlike languages with a static class concept, XOTcl
supports dynamic class/superclass relationships. At any time the class
graph may be changed entirely using the <TT>superclass</TT> method, or
an object may change its class through the <TT>class</TT> method. This
feature can be used for an implementation of a life-cycle or other
intrinsic changes of object properties (in contrast to extrinsic
properties e.g. modeled through roles and implemented through
per-object and per-class mixins <a href="#xotcl-mixin">[Neumann and
Zdun 1999c]</a> ) . These changes can be achieved without loosing the
object's identity, its inner state, and its per-object behavior (procs
and per-object mixins).
</P>
<P ALIGN=CENTER STYLE="margin-bottom: 0in"><A NAME="features1"></A><A NAME="11761"></A>
  
</P>
<P ALIGN=CENTER STYLE="margin-bottom: 0in"><STRONG>Figure 2:</STRONG>
Object and Class System
</P>
<CENTER>
<TABLE WIDTH=469 BORDER=0 CELLPADDING=2 CELLSPACING=0>
<COL WIDTH=465>
<TR>
<TD WIDTH=465>
<P ALIGN=CENTER><IMG SRC="obj_class_system.gif" NAME="Graphic5" ALIGN=BOTTOM WIDTH=467 HEIGHT=144 BORDER=0></P>
</TD>
</TR>
</TABLE>
</CENTER>
<TABLE COLS=2 WIDTH=100% BORDER=0 CELLPADDING=2 CELLSPACING=0 BGCOLOR="#000055">
<TR>
<TD WIDTH=75%>
<P><A NAME="basic"></A><FONT COLOR="#ffffff"><FONT FACE="Arial, Helvetica"><FONT SIZE=6>Basic
Functionalities </FONT></FONT></FONT>
</P>
</TD>
<TD>
<IMG SRC="logo-100.jpg" NAME="Graphic6" ALIGN=RIGHT WIDTH=102 HEIGHT=42 BORDER=0></TD>
</TR>
</TABLE>
<H2><A NAME="object"></A> <BR>Objects
</H2>
<P>Initially XOTcl offers two new commands: <TT>Object</TT> and
<TT>Class</TT>. They represent hooks to the features of the language.
This section discusses both of them in detail and shows how they
function in the context of XOTcl. Note, that even if most of this is
compatible to OTcl, a few changes occur. For this reason, this
section is no introduction to plain OTcl. The <TT>Object</TT> command
provides access to the <TT>Object</TT> class, which holds the common
features of all objects, and allows us to define new objects. Objects
are always instances of classes, therefore, objects defined with the
<TT>Object</TT> command are (initially) instances of the <TT>Object</TT>
class. But since they have no user-defined type, they may be referred
to as <EM>singular objects</EM>. As all other objects they may be
specialized by object-operations and -data.
</P>
<P>The object command has the following syntax:
</P>
<PRE STYLE="margin-bottom: 0.2in"><em> Object objName ?args?</em></PRE><P>
A command of this form is a short-cut for a message to the <TT>create</TT>
instance method (forwarded automatically by the <TT>unknown</TT>
mechanism, which is invoked every time the message dispatch system
discovers an unknown message):
</P>
<PRE STYLE="margin-bottom: 0.2in"><em> Object create objName ?args?</em></PRE><P>
It creates a new object of type <TT>Object</TT> with the name <TT>objName</TT>
(in fact it invokes a <TT>create</TT> call on the <TT>Object</TT> class).
<TT>objName</TT> becomes a new command, which allows us to access the
created object. Similar to the <TT>Object</TT> command it may be
used like a normal Tcl-command (using sub-commands to access the
object's methods). Therefore, this form of access is called
<EM>object-command</EM> approach. A simple example is an object which
holds the information of a kitchen. It is created by:
</P>
<PRE STYLE="margin-bottom: 0.2in"> Object kitchen</PRE><P>
An object creation calls the constructor <TT>init</TT> of the
object's class. The destruction of an object is handled by the
<TT>destroy</TT> instance method. The general syntax of <TT>destroy
</TT>is:
</P>
<PRE STYLE="margin-bottom: 0.2in"><em> objName destroy</em></PRE><P>
E.g. the kitchen object is destroyed by:
</P>
<PRE STYLE="margin-bottom: 0.2in"> kitchen destroy</PRE><P>
To invoke a user-defined destruction process, it is possible to
overload this instance method in every class derived from object.
</P>
<H3><A NAME="data_on_obj"></A>Data on Objects
</H3>
<P>The <TT>Object</TT> class provides a range of operations to manage
objects, including those to manipulate data-structures on the
objects. They are similiar to the same-named Tcl-commands:
</P>
<PRE><em> objName set varname ?value?
objName unset v1 ?v2 ... vn?
</em></PRE>
<P>
The <TT>set</TT> instance method with given <TT>value</TT> option
allows us to manipulate an object-variable's value or to create a new
one, if the variable <TT>varname</TT> does not exist on the object so
far. Without <TT>value</TT> option the <TT>set</TT> operation queries
the variable and returns it's value, if the variable exists,
otherwise it produces an error message. The <TT>unset</TT> operation
deletes one or optionally a set of variables from an object. For
example the <TT>kitchen</TT> object can store information on the
color of the wall-paper by:
</P>
<PRE STYLE="margin-bottom: 0.2in"> kitchen set wallPaperColor white</PRE><P>
Similar to Tcl-variables the object variables are dynamical; they
may be set at run-time when they are needed and unset when they
become obsolete. E.g. the persons in the kitchen may be stored in an
array. If there are no persons in the kitchen the array is deleted:
</P>
<PRE> # Peter enters the kitchen to cook
kitchen set persons(cook) Peter
...
# Marion enters the kitchen to take one of the seats
kitchen set persons(seat1) Marion
...
# Both Peter and Marion leave the kitchen
# the array is deleted by unset
kitchen unset persons</PRE><P>
Since XOTcl variables are internally realized through Tcl-variables
they may be treated like all Tcl-variables. For that reason they have
all Tcl-variable abilities, including the possibility to handle them
as lists or arrays (as seen in the last example). The <TT>array</TT>
command of Tcl is mapped to an XOTcl-command directly. An
object-oriented call to an object of the form
</P>
<PRE STYLE="margin-bottom: 0.2in"><em> objName array option ary args</em></PRE><P>
forwards its arguments to an <TT>array</TT> Tcl-command for the
object´s instance variable <TT>ary</TT>. It could be used like
the same-named Tcl-command, e.g. the command
</P>
<PRE STYLE="margin-bottom: 0.2in"> kitchen array names persons</PRE><P>
returns all indexes currently stored in the <TT>persons</TT> array.
</P>
<P>Similarly Tcl´s <TT>incr</TT> command is mapped to the
object system. A call with the syntax:
</P>
<PRE STYLE="margin-bottom: 0.2in"><em> objName incr varName ?value?</em></PRE><P>
increments <TT>varName</TT> with the given value (or without given
value with 1).
</P>
<H3><A NAME="obj_methods"></A>Methods on Objects
</H3>
<P>Methods in XOTcl resemble Tcl-procedures. On objects one can define
object-specific methods, called procs. Instance methods which are
defined on classes are called instprocs. A new proc is defined using
the <TT>proc</TT> instance method of the class <TT>Object</TT>:
</P>
<PRE STYLE="margin-bottom: 0.2in"><em> objName proc name args body</em></PRE><P>
The arguments of the <TT>proc</TT> instance method specify the name,
the arguments as a Tcl-list, and the body of the new proc. All of them
must be given, only one of <TT>args</TT> and <TT>body</TT> may be
empty. An example proc would be a method to let persons enter the
kitchen:
</P>
<PRE> kitchen proc enter {name} {
[self] set persons($name) [clock seconds]
}</PRE><P>
Here the predefined <TT>self</TT> command is used in one of three
possible ways, which allow us to access useful information when working
with XOTcl-methods, these are in particular:
</P>
<UL>
<LI><P STYLE="margin-bottom: 0in"><TT>self</TT> - returns the name
of the object, which is currently in execution. This command is
similar to <TT>this</TT> in C++. It is automatically generated on
each object. If it is called from outside of a proc, it returns the
error message ``<TT>Can't find self</TT>''.
</P>
<LI><P STYLE="margin-bottom: 0in"><TT>self class</TT> - the self
command with a given argument <TT>class</TT> returns the name of the
class, which holds the currently executing instproc. Note, that this
may be different to the class of the current object. If it is called
from a proc it returns an empty string.
</P>
<LI><P><TT>self proc</TT> - the self command with a given argument
<TT>proc</TT> returns the name of the currently executing proc or
instproc.
</P>
</UL>
<p>The method <tt>enter</tt> can be written in XOTcl as well with
less syntactic overhead by using the predefined primitive <tt>my</tt>
instead of <tt>[self]</tt>:</p>
<PRE>
kitchen proc enter {name} {
my set persons($name) [clock seconds]
}</PRE><P>
<P>Note, that there is a difference to the realization of these
object informations to OTcl. XOTcl uses commands in order to make
XOTcl-methods compatible to Tcl-procedures and accessible via
namespace-paths. OTcl uses the three variables <TT>self</TT>, <TT>class</TT>
and <TT>proc</TT>, which are filled automatically with proper values
by the interpreter each time a method is called. To gain
compatibility a compilation option <TT>AUTOVARS</TT> is available
which provides these variables additionally (if the option is defined
when compiling XOTcl). The default is, that the option is turned off.
</P>
<P>Each XOTcl-method has its own scope for definition of local
variables for the executing method. In most cases when a method uses
object-variables, it is likely that the programmer wants to make one
or more of these variables part of the method's scope. Then the
Tcl-command for variable handling, like <TT>set</TT>, <TT>lindex</TT>,
<TT>array</TT>, ... work also on these variables. The
<TT>instvar</TT> instance method links a variable to the scope of
an executing method. It has the syntax:
</P>
<PRE STYLE="margin-bottom: 0.2in"><em> objName instvar v1 ?v2 ... vn?</em></PRE>
<P>
It makes the variables <tt>v1 ... vn</tt>, which must
be variables of the object, part of the current method's scope. A
special syntax is:
</P>
<PRE STYLE="margin-bottom: 0.2in"><em> {varName aliasName}</em></PRE>
<P>
for one of the variables. This gives the variable with the name
<TT>varName</TT> the alias <TT>aliasName</TT>. This way the variables
can be linked to the methods scope, even if a variable with that name
already exists in the scope. Now the <TT>enter</TT> method can be
adapted slightly and a <TT>leave</TT> method can be added, which uses
Tcl's <FONT FACE="courier, monospace">info</FONT> command to check
whether the named person is in the object's <TT>persons</TT> array. To
demonstrate the alias-syntax this is done with the <TT>persons</TT>
array and the alias <TT>p</TT>.
</P>
<PRE> kitchen proc enter {name} {
my instvar persons
set persons($name) [clock seconds]
}
kitchen proc leave {name} {
my instvar {persons p}
if {[info exists p($name)]} {
puts "$name leaves after [expr {[clock seconds]-$p($name)}] seconds"
unset p($name)
} else {
puts "$name is not in the room"
}
}</PRE><H3>
<A NAME="obj_info"></A>Information on Objects
</H3>
<P STYLE="margin-bottom: 0in">XOTcl offers reading and writing
introspection. The reading introspection abilities are packed
compactly into the <TT>info</TT> instance method which is available
for objects and classes (there are special info options for object
aggregations, nested classes, mixins, filters, meta-data and
assertions, which are explained separately in the following
sections). The <TT>info</TT> instance method's options, from the view
of an object, are summarized in the following table. They are
identically to the OTcl info options on objects.
</P>
<P ALIGN=CENTER STYLE="margin-bottom: 0in"><A NAME="table_oinfo"></A> 
</P>
<P ALIGN=CENTER STYLE="margin-bottom: 0in"><STRONG>The object info
options</STRONG></P>
<CENTER>
<TABLE BORDER=1>
<TR>
<TD>
<P ALIGN=LEFT><TT>objName info args method </TT>
</P>
</TD>
<TD>
<P ALIGN=LEFT>Returns the arguments of the specified method.</P>
</TD>
</TR>
<TR>
<TD>
<P ALIGN=LEFT><TT>objName info body method </TT>
</P>
</TD>
<TD>
<P ALIGN=LEFT>Returns the body of the specified method.</P>
</TD>
</TR>
<TR>
<TD>
<P ALIGN=LEFT><TT>objName info class ?classname? </TT>
</P>
</TD>
<TD>
<P ALIGN=LEFT>Returns the name
of the class of the current
object, if classname was not
specified. Otherwise it
returns 1 if classname matches
the object's class and 0 if
not.
</P>
</TD>
</TR>
<TR>
<TD>
<P ALIGN=LEFT><TT>objName info commands ?pattern?</TT></P>
</TD>
<TD>
<P ALIGN=LEFT>Returns all
commands defined on the object
if <TT>pattern</TT> was not
specified. Otherwise it
returns all commands that
match the pattern.</P>
</TD>
</TR>
<TR>
<TD>
<P ALIGN=LEFT><TT>objName info default method arg var</TT></P>
</TD>
<TD>
<P ALIGN=LEFT>Returns 1 if the
argument <TT>arg</TT> of the
method <TT>method</TT> has a
default value, otherwise 0. If
the default value exists it is
stored in <TT>var</TT>.</P>
</TD>
</TR>
<TR>
<TD>
<P ALIGN=LEFT><TT>objName info procs ?pattern? </TT>
</P>
</TD>
<TD>
<P ALIGN=LEFT>Returns all
procs defined on the object if
<TT>pattern</TT> was not
specified, otherwise it
returns all procs that match
the pattern.</P>
</TD>
</TR>
<TR>
<TD>
<P ALIGN=LEFT><TT>objName info vars ?pattern? </TT>
</P>
</TD>
<TD>
<P ALIGN=LEFT>Returns all variables defined on the object if
<TT>pattern</TT>was not specified, otherwise it returns all
variables that match the pattern.</P>
</TD>
</TR>
</TABLE>
</CENTER>
<P><BR><BR>
</P>
<P>For example on the <TT>kitchen</TT> object
</P>
<PRE STYLE="margin-bottom: 0.2in"> kitchen info procs</PRE><P>
returns <TT>enter</TT> and <TT>leave</TT> as a Tcl-list since these
are the procs defined on the object.
</P>
<H2><A NAME="classes"></A>Classes
</H2>
<H3><A NAME="class_instance"></A>Creating Classes and deriving
Instances
</H3>
<P>There are different ways to create a class in XOTcl. They have in
common that they derive the new class from a meta-class. Initially the
<TT>Class</TT> command provides access to the meta-class
<TT>Class</TT>, which holds the features common to all classes. It
also allows one to derive new meta-classes. The common way to create a
new class is:
</P>
<PRE STYLE="margin-bottom: 0.2in"><em> Class ClassName ?args?</em></PRE>
<P>
Similar to the object short form, this is a short form of a call to
the <TT>create</TT> instance method of the meta-class <TT>Class</TT>,
which is also executed by the standard <TT>unknown</TT> mechanism.
This mechanism is always triggered when XOTcl does not know a method
called on an object. Supposed that there is no method with the name
<TT>ClassName</TT>, defined on the class-object of <TT>Class</TT>,
XOTcl looks up the method <TT>unknown</TT> (which is found on the
Class <TT>Object</TT>) and executes it. The standard unknown-mechanism
of XOTcl calls <TT>create</TT> with all arguments stepping one step
to the right; in the general case:
</P>
<PRE STYLE="margin-bottom: 0.2in"><em> Class create ClassName ?args?</em></PRE><P>
This may also be called directly. Besides the indirection when using
<TT>unknown</TT>, in most cases there is no difference in the action
performed: Firstly the memory is allocated, using the <TT>alloc</TT>
instance method; as the next step the constructor <TT>init</TT> is called
on the creating object, which is in this case the class-object of the
meta-class <TT>Class</TT>. In seldom cases the programmer may want to
suppress the <TT>init</TT> call. To do so the <TT>alloc</TT> instance
method may also be called directly:
</P>
<PRE><em>
Class alloc ClassName ?args?
</em></PRE>
<P>
As seen in the preceding section objects are created in the same way.
The difference was, that the command <TT>Object</TT>, which accesses
a class, instead of the command <TT>Class</TT>, which accesses a
meta-class, was used. The user-defined classes may also be used in
the same way to create new objects:
</P>
<PRE STYLE="margin-bottom: 0.2in"><em> ClassName objName ?args?</em></PRE>
<P>
Resembling the creation of classes this creates an object <TT>objName</TT>
of type <TT>ClassName</TT> using the <TT>unknown</TT> mechanism. That
means the <TT>create</TT> instance method of the class is called. If
there is no other instance method defined on the class-path so far
(which would mean, an user defined creation process is invoked), the
<TT>create</TT> instance method of the class <TT>Object</TT> is
invoked. This method is similar to the <TT>create</TT> method of the
meta-class <TT>Class</TT>. It firstly calls the <TT>alloc</TT>
instance method on its (of the <TT>Class</TT> class) which allocates
memory for the object, and makes it an instance of it's class.
Afterwards a call to the constructor <TT>init</TT> is invoked.
</P>
<P>Now we can specify the object for the kitchen by the class to
which it belongs. In this case a kitchen is an instance of a room.
</P>
<PRE> Class Room
Room kitchen</PRE><P>
A <TT>set</TT> call on a class creates an instance variable on the
class-object. This variable is unique for all instances, therefore,
it may be referred to as a class variable.
</P>
<H3><A NAME="class_methods"></A>Methods in Classes
</H3>
<P>Methods in classes are called "instprocs". Instprocs are reachable
for the class-object and all other instances of the class. The syntax
for defining an instproc is:
</P>
<PRE STYLE="margin-bottom: 0.2in"><em> ClassName instproc procname args body</em></PRE>
<P>
It is similar to the definition of procs on objects, but uses the
keyword <TT>instproc</TT> to distinguish between the methods defined
on the class-object and those defined on the class. Since all rooms
(in the modeled world) have ceilings, we may want to define a simple
convenience instproc, which is able to set the color:
</P>
<PRE> Room instproc setCeilingColor color {
my set ceilingColor $color
}</PRE><P>
A special instproc, the constructor <TT>init</TT>, was mentioned
already. Now we are able to define such an instproc. Defined on a
class it is responsible for all initialisation tasks, which needed to
be performed, when constructing a new instance object of the class.
The constructor of the <TT>Room</TT> can initialize a variable for
the color, in which the ceiling is painted, to white as default,
since this is the color of ceilings without painting.
</P>
<PRE> Room instproc init args {
my setCeilingColor white
next
}
</PRE>
<P>
After this definition, all instances derived from the <TT>Room</TT>
class have an instance variable <TT>ceilingColor</TT> with the value
<TT>white</TT>. The <TT>args</TT> argument used here is a special
argument in Tcl which allows us to use a list of arguments which may
change its length from call to call.
</P>
<H3><A NAME="class_info"></A>Information on Classes
</H3>
<P STYLE="margin-bottom: 0in">Resembling to objects, information on
classes may be gained through the <TT>info</TT> instance method of the
meta-class <TT>Class</TT>. Note that this instance method does not
only support the class info options, but also the object info options,
since the accessing command refers to the class-object, which itself
is an object and, therefore, offers its informations. The following
table summarizes the additional info options available on classes.
</P>
<P ALIGN=CENTER STYLE="margin-bottom: 0in"><STRONG>The class info
options</STRONG></P>
<CENTER>
<TABLE BORDER=1>
<TBODY>
<TR>
<TD>
<P ALIGN=LEFT><TT>ClassName info heritage ?pattern? </TT>
</P>
</TD>
<TD>
<P ALIGN=LEFT>Returns a list of all classes in the precedence
order of the class hierarchy matching <TT>pattern</TT> or of
all, if <TT>pattern</TT> was not specified.</P>
</TD>
</TR>
</TBODY>
<TBODY>
<TR>
<TD>
<P ALIGN=LEFT><TT>ClassName info instances ?pattern? </TT>
</P>
</TD>
<TD>
<P ALIGN=LEFT>Returns a list of the instances of the class
matching <TT>pattern</TT> or of all, if <TT>pattern</TT> was not
specified.
</P>
</TD>
</TR>
</TBODY>
<TBODY>
<TR>
<TD>
<P ALIGN=LEFT><TT>ClassName info instargs method</TT></P>
</TD>
<TD>
<P ALIGN=LEFT>Returns the arguments of the specified method.</P>
</TD>
</TR>
</TBODY>
<TBODY>
<TR>
<TD>
<P ALIGN=LEFT><TT>ClassName info instbody method</TT></P>
</TD>
<TD>
<P ALIGN=LEFT>Returns the body of the specified method.</P>
</TD>
</TR>
</TBODY>
<TBODY>
<TR>
<TD>
<P ALIGN=LEFT><TT>ClassName info instcommands ?pattern?</TT></P>
</TD>
<TD>
<P ALIGN=LEFT>Returns all commands defined on the class, if
<TT>pattern</TT> was not specified, otherwise it returns all
commands that match the pattern.</P>
</TD>
</TR>
</TBODY>
<TBODY>
<TR>
<TD>
<P ALIGN=LEFT><TT>ClassName info subclass ?classname? </TT>
</P>
</TD>
<TD>
<P ALIGN=LEFT>Returns a list of all subclasses of the class, if
<TT>classname</TT> was not specified, otherwise it returns 1 if
<TT>classname</TT> is a subclass and 0 if not.</P>
</TD>
</TR>
</TBODY>
<TBODY>
<TR>
<TD>
<P ALIGN=LEFT><TT>ClassName info superclass ?classname? </TT>
</P>
</TD>
<TD>
<P ALIGN=LEFT>Returns a list of all super-classes of the class,
if <TT>classname</TT> was not specified, otherwise it returns 1
if <TT>classname</TT> is a superclass and 0 if not.</P>
</TD>
</TR>
</TBODY>
</TABLE>
</CENTER>
<P><BR><BR>
</P>
<H3><A NAME="class_inheritance"></A>Inheritance
</H3>
<P>Besides encapsulation of operations and state in objects, a second
central ability of object-orientation is inheritance. XOTcl supports
single and multiple inheritance with a directed acyclic class
graph. Automatically each new class created by the instance methods
<TT>create</TT> and <TT>alloc</TT> of <TT>Class</TT> inherits from
<TT>Object</TT>. Therefore, it is ensured that all instances of the
new class have access to the common features of objects stored in the
class <TT>Object</TT>.
</P>
<P>To specify further inheritance relationships the instance methods
<TT>superclass</TT> of <TT>Class</TT> is used:
</P>
<PRE STYLE="margin-bottom: 0.2in"><em> ClassName superclass classList</em></PRE><P>
E.g. in the example a kitchen may be seen as a
special room:
</P>
<PRE> Class Room
Class Kitchen -superclass Room</PRE><P>
Now all instances of <TT>Kitchen</TT> are able to access the
operations stored in the <TT>Room</TT> and in the <TT>Kitchen</TT> class.
Note the transition the kitchen was going through: firstly it was a
singular object, then it was an object with a user-defined class,
and now it is a class. This is possible (and not senseless) because
of the per-object specialisation ability and the dual shape of a
class, which is at the same time object and class. Both lead to a
seamless connection of the run-time properties (the object features)
and their descriptive properties (the class features). It is possible
to avoid the strict distinction between them, known from static typed
languages, like C++, Java, etc.
</p>
<p>Moreover, since the syntaxes of constructs expressing the same
concern are nearly identical, we can refactor a solution with very few
changes to the alternative. We will see similar "ease of refactoring"
throughout the XOTcl language. E.g., we can also easily refactor the
class hierarchies or exchange class hierarchies against mixin
solutions with only slight changes in the code.
</P>
<P>Besides single inheritance, as seen, XOTcl provides also multiple
inheritance. This is syntactically solved by giving the <TT>superclass</TT>
instance method a list of classes instead of a single class as
argument.
</P>
<PRE> Class Room
Class 4WallsRoom -superclass Room
Class CookingPlace
Class Kitchen -superclass {4WallsRoom CookingPlace}
</PRE><P>
Now the kitchen class is specialized a bit more. It is a special room
which has four walls <EM>and</EM> it is a cooking place. Multiple
inheritance, as seen here, is as simple to apply as single
inheritance.
</P><P>
Most often when the disadvantages of multiple inheritance are
discussed, the name resolution along the class graph is considered as
the biggest problem. The question is, which method is to be chosen and
which path through class graph is to be taken, if more then one method
of the specified name exist on the class graph.
</P>
<P ALIGN=LEFT STYLE="margin-bottom: 0in">In the example such questions
would arise for an object of the <TT>Kitchen</TT> class, if two
same-named methods are defined on <TT>CookingPlace</TT> and
<TT>4WallsRoom</TT> or if a method of the class <TT>Object</TT> is
called, which is reachable through two paths (along
<TT>CookingPlace</TT> or <TT>Room</TT>).
</P>
<P ALIGN=LEFT STYLE="margin-bottom: 0in">Often - e.g. in the
inheritance model of C++ - the path through the graph is not clearly
determined and/or the rules are too complicated to be understood on
the first glance. The programmer often can only determine by trial
which method is found firstly. Than an explicit naming of the class is
necessary, which means storage of non-local information in
sub-classes. Often different compilers of one language behave
differently. All these issues make code reuse difficult. Moreover
understandability and portability are reduced.
</P>
<P ALIGN=CENTER STYLE="margin-bottom: 0in"><BR>
</P>
<P ALIGN=CENTER STYLE="margin-bottom: 0in"><STRONG>Figure 3:</STRONG>
The example classes and the following next-path</P>
<CENTER>
<TABLE BORDER=0 CELLPADDING=2 CELLSPACING=0>
<COL>
<TR>
<TD>
<IMG SRC="next-path.gif" NAME="Graphic13" ALIGN=LEFT BORDER=0><BR CLEAR=LEFT></TD>
</TR>
</TABLE>
</CENTER>
<P>XOTcl goes an intuitive and unambiguous way to solve this problem.
It resolutes the precedence order along a ``<EM>next-path</EM>''.
Firstly the class of the object is searched, which is <TT>Kitchen</TT>
in example. Then the super-classes are searched in definition order,
which means at first <TT>4WallsRoom</TT>, then <TT>CookingPlace</TT>.
Each branch is searched completely, before changing to the next
branch. That means, <TT>Room</TT> is searched, before the
<TT>CookingPlace</TT> branch is visited. At last the top of the
hierarchy, the class <TT>Object</TT>, is searched.
</P>
<P>The usage of <TT>next</TT> in XOTcl is different to OTcl: In OTcl
it is always necessary to provide the full argument list for every
invocation explicitly. In XOTcl, a call of <TT>next</TT> without
arguments can be used to call the shadowed methods with the same
arguments (which is the most common case). When arguments should be
changed for the shadowed methods, they must be provided explicitly in
XOTcl as well. In the rare case that the shadowed method should
receive no argument, the flag <TT>--noArgs</TT> must be used.
</P>
<H3><A NAME="class_destroy"></A>Destruction of Classes
</H3>
<P>Classes are destroyed by the destruction of the class-object using
the <TT>destroy</TT> method of the <TT>Object</TT> class. The
destruction of super-classes does not destroy the sub-classes. The
super-class is simply removed from the sub-classes' super-class
lists. All classes have the super-class <TT>Object</TT>, if no
super-class is specified. Therefore, if all super-classes are
destroyed or removed, the new super-class is <TT>Object</TT>, not: no
super-class. The destruction of the class of an object does neither
delete the object nor leave it without class. In XOTcl a deleted class
leaves it's instances with the class <TT>Object</TT>.
</P>
<P>So all empty class- and superclass-relationships are automatically
reseted to <TT>Object</TT>. Note, that this are differences to OTcl,
where the destruction of an class destroys all instances and an empty
super-class list remains empty.
</P>
<H3><A NAME="class_method_chaining"></A>Method Chaining
</H3>
<P>A special feature of XOTcl is the method chaining without explicit
naming of the ``mix-in''-method. It allows one to mix the same-named
superclass methods into the current method (modeled after CLOS). The
previously described next-path is the basis for this functionality.
At the point marked by a call to the <TT>next</TT> primitive of XOTcl
the next shadowed method on the next path is searched and, when it is
found, it is mixed into the execution of the current method. When no
method is found, the call of <TT>next</TT> returns an empty string,
otherwise it returns the result of the called method. Note, that the
realization through a primitive command -- similar to the <TT>self</TT>
command -- is a difference to OTcl, where <TT>next</TT> is realized
through an instance method of <TT>Object</TT>. The syntax is:
</P>
<PRE STYLE="margin-bottom: 0.2in"><em> next ?arguments|--noArgs?</em></PRE><P>
Note, that also the usage of <TT>next</TT> in XOTcl is different to
OTcl, since the <TT>next</TT> call without arguments in OTcl means
per default that no arguments are passed. But most often all
arguments are passed through to the shadowed method (since it will
most likely have the same signature like its predecessor). When all
variables should be passed through, in OTcl it is necessary for
correct variable substitution to use:
</P>
<PRE STYLE="margin-bottom: 0.2in"> eval $self next $args</PRE><P>
To avoid such difficulties, we made the passing of all arguments the
default case; a simple
</P>
<PRE STYLE="margin-bottom: 0.2in"> next</PRE><P>
performs the task of passing all arguments to the shadowed methods.
These arguments are called the <EM>standard arguments</EM>. If the
standard argument feature should not be used, optionally arguments
can be given or the flag <TT>--noArgs</TT> could be set as sole
argument, which means that the shadowed method is called with no
arguments.
</P>
<P>
E.g. the following <tt> next </tt> call ignores the standard arguments
and sends the arguments 1 and 2 instead:
</P>
<PRE STYLE="margin-bottom: 0.2in"> next 1 2</PRE><P>
<P>As an example all classes involved in the previous example should
get a constructor instance method, which simply sets an instance
variable on the object:
</P>
<PRE> Room instproc init args {
my set roomNumber 0
next
}
4WallsRoom instproc init args {
my set doorPosition 0
next
}
CookingPlace instproc init args {
my set stoveType electric
next
}
Kitchen instproc init args {
my set cookName -
next
}</PRE><P>
After creation an object of class <TT>Kitchen</TT> gets automatically
four instance variables <TT>cookName</TT>, <TT>roomNumber</TT>,
<TT>doorPosition</TT> and <TT>stoveType</TT> set up with default
values in this order (since this is the order of the classes in the
next-path). Note, that the order is important, because one missing
next call, in one of the <TT>init</TT> methods, means that succeeding
<TT>init</TT> methods will not be executed. This mechanism functions
equally on all kinds of instprocs, not only on constructors.
</P>
<P>The constructors use the <TT>args</TT> argument, which allows us to
give a list of variable length as arguments. To ensure reusability of
our classes the constructors should use <TT>args</TT> in most cases,
since they may pass through arguments for constructors further up the
class hierarchy.
</P>
<P>If a <TT>proc</TT> with the searched name exists on the object it
shadows all instprocs. A <TT>next</TT> call in a proc leads to
the normal next-paths search, starting with the object's class.
</P>
<p>By the way, an observant reader might notice that the example
above can be rewritten without explicit constructors, just by
using paramters with default values.
</p>
<PRE> Class Room -parameter {{roomNumber 0}}
Class 4WallsRoom -superclass Room -parameter {{doorPosition 0}}
Class CookingPlace -parameter {{stoveType electric}}
Class Kitchen -superclass {4WallsRoom CookingPlace} -parameter {{cookName -}}
</PRE><P>
If an instance of a Kitchen is created it will contain instance
variables for <tt>doorPosition</tt>, <tt>cookName</tt>,
<tt>roomNumber</tt>, and <tt>stoveType</tt>, as the following
statements will show.</p>
<PRE> Kitchen k
puts [k info vars]
</PRE>
<H2><A NAME="class_dynamics"></A>Dynamic Class and Superclass
Relationships
</H2>
<P>Another property of XOTcl that distinguishes it from statically typed
languages are dynamics of class relationships. The realization of the
definition of super-classes as seen above with the <TT>superclass</TT>
method suggests already, that it is not only available at the class
definition time. In the above example its appended to the class
definition with "<TT>-superclass</TT>" as a short syntax
for method invocation at definition time (all other available methods
can also be called with a preceding dash ("-") appended
to definitions).
</P>
<P>At any time the class graph may be changed entirely using the
<tt>superclass</tt> method. Suppose the rooms and kitchens created in
modeling of a house should be displayed to a screen, but it is not
determined, whether the user of the system has the possibilities for
graphical outputs. Two classes <TT>TextOutput</TT> and
<TT>GraphicalOutput</TT> may be defined, which handle the output. Both
have an instproc <TT>paint</TT> which does the painting of the virtual
world on the chosen display type. The common output requirements are
handled by a derived class <TT>VirtualWorldOutput</TT> which calls the
<TT>paint</TT> method of the superclass using <TT>next</TT>. In
statically typed languages it would need more sophisticated constructs
to change the output class at run-time. E.g. a delegation to another
object handling the intrinsic task of the output object would be
introduced solely for the purpose of configuring the output
form. With a dynamic class system we can use the <TT>superclass</TT>
method to do so easily:
</P>
<PRE> Class TextOutput
TextOutput instproc paint args {
# do the painting ...
}
Class GraphicalOutput
GraphicalOutput instproc paint args {
# do the painting ...
}
# initially we use textual output
Class VirtualWorldOutput -superclass TextOutput
VirtualWorldOutput instproc paint args {
# do the common computations for painting ...
next; # and call the actual output
}
# users decides to change to graphical output
VirtualWorldOutput superclass GraphicalOutput
</PRE>
<P>
Sometimes, such a change to new intrinsic properties should not happen
for all instances of a class (or the class hierarchy), but only for
one specific object. Then the usage of a dynamic super-class
relationship is a too coarse-grained means. A second form of such
dynamics is the changing of the relationship between object and
class. This means, objects can also change their class dynamically at
run-time. This feature may be used to model a life-cycle of an object,
without loosing the object's identity, inner state or
per-object-specializations through procs. The <TT>class</TT> instance
method enables this functionality.
</P>
<P>An example would be an agent for the virtual world. Agents may be
placeholders for persons, who interactively travel the world, or
programs, which act automatically. When a person decides at run-time
to give a task it has performed formerly by hand to an automatic
agent, the agents nature changes from interactive agent to automatic
agent, but the identity and the local state (that means the parts of
the task, that are already fulfilled by the person) stay the same.
This is a scenario for changing class relationships, e.g.:
</P>
<PRE> Class Agent
Class AutomaticAgent -superclass Agent
Class InteractiveAgent -superclass Agent
# create a new agent for a person
InteractiveAgent agent1
# the person does something ...
#and decides the change to an automatic agent
agent1 class AutomaticAgent</PRE>
<H2>
<A NAME="meta-classes"></A>Meta-Classes
</H2>
<P>As seen already, a special kind of classes are meta-classes. They
are classes (and like all XOTcl classes also objects) which contain
the features common to all derived classes. We have already shown the
meta-class <TT>Class</TT>, which we was used in the previous sections
to create new classes.
</P>
<P>A new meta-class can be derived from an existing meta-class by
defining <TT>Class</TT> as superclass. Beside that a meta-class
creation is a normal class-definition:
</P>
<PRE STYLE="margin-bottom: 0.2in"><em> Class myMetaClass -superclass Class</em></PRE><P>
This defines a new meta-class <TT>myMetaClass</TT>, which has all the
abilities of meta-classes. That means that the programmer is able to
specify new class features or override old ones. Later she/he may
instantiate these into new classes.
</P>
<P>This is a very powerful language feature, since it allows one to give
some classes further abilities than the others (or to restrict
classes). This way large program structures, like certain design
pattern parts, may be instantiated. Meta-classes hold the common
abstract parts of the structures. They allow one to form libraries of
such structures very easily.
</P>
<P>
As a simple example we can derive a new meta-class
<TT>NoClassInfo</TT> from <tt>Class</tt>. Later we override the
<tt>info</tt> method of <tt>Class</tt>. Thus the classes created with
<TT>NoClassInfo</TT>, have an <TT>info</TT> option that only produces
an error message. All classes created with <TT>NoClassInfo</TT>, like
<TT>Agent</TT> in the example below, are not capable of accessing the class
<tt>info</tt> method anymore:
</P>
<PRE> Class NoClassInfo -superclass Class
# redefine info ability
NoClassInfo instproc info args {
return "No class info avaiable"
}
# derive agent class from meta-class, which
# can not access class info
NoClassInfo Agent</PRE><P>
Now a call like:
</P>
<PRE STYLE="margin-bottom: 0.2in"> Agent info superclass</PRE><P>
returns in the error message.
</P>
<H2>
<A NAME="destroy-logic"></A>Create, Destroy, and Recreate Scheme
</H2>
<P>
<P>
XOTcl allows since version 0.84 for a flexible destroy and recreate scheme.
<tt>create</tt> and <tt>alloc</tt> are both Class instprocs
handling creation for their instances. I.e.:
</P>
<PRE>
className alloc [self]
</PRE>
and
<PRE>
className create [self]
</PRE>
<P>
are used for creating an instance. A similar method <tt>instdestroy</tt>
exists on Class that handles physical destruction of an object. The
method <tt>destroy</tt> on Object which lets an object destroy itself in fact
has the following behavior:
</P>
<PRE> Object instproc destroy args {
[my info class] instdestroy [self]
}
</PRE>
<P>
However, this behavior is not implemented in XOTcl, but in C.
<tt>create</tt> distinguishes between the following situations:
</P>
<ul>
<li> <em>Create a new object:</em>
<tt>create</tt> calls <tt>alloc</tt> and then
<tt>doInitializations</tt>.
<li> <em>Recreate an existing object:</em>
When the specified object exists, it is recreated through the
<tt>recreate</tt> method:
<pre>
<givenClass> recreate [self]
</pre>
<P>
<tt>recreate</tt> can be customized e.g. by overloading or interception.
By default it calls <tt>cleanup</tt> followed by <tt>doInitializations</tt>.
</ul>
<p>
In both cases, the method <tt>doInitializations</tt>
is called automatically from C and has the following default behavior:
</p>
<ul>
<li> Search for parameter default values,
<li> Call parameter initialization methods,
<li> Call the constructor <tt>init</tt>.
</ul>
<P>
Each step has a method call that can be changed, intercepted, etc. Of
course, <tt>cleanup</tt>, <tt>recreate</tt>, <tt>instdestroy</tt>,
etc. can also be overloaded or intercepted.
</P>
<P>
Consider a typical case for overloading <tt>recreate</tt>: a structure
preserving <tt>recreate</tt> that cleans up the class but
preserves the existing class hierarchy (subclass and instance relationships):
</P>
<pre>
Class StructurePreservingRecreate
StructurePreservingRecreate instproc recreate {cl args} {
if {[my isclass $cl]} {
set subclass [$cl info subclass]
set instances [$cl info instances]
}
next
if {[my isclass $cl]} {
foreach sc $subclass {
$sc superclass $cl
}
foreach i $instances {
$i class $cl
}
}
}
Object instmixinappend StructurePreservingRecreate
</pre>
<p>
Now the following code does not change the superclass or instance
relationships of C:
</p>
<pre>
Class A
Class B
Class C -superclass {A B}
Class D
Class E -superclass {C D}
C c1
C c2
# recreate -> is structure preserving
Class C -superclass {A B}
C c2
# test
puts superclass=[C info superclass]
puts subclass=[C info subclass]
puts instances=[C info instances]
puts class=[c1 info class]
puts class=[c2 info class]
</pre>
<!-- PAGE BREAK -->
<TABLE COLS=2 WIDTH=100% BORDER=0 CELLPADDING=2 CELLSPACING=0 BGCOLOR="#000055">
<TR>
<TD WIDTH=75%>
<P><A NAME="interceptors"></A><FONT
COLOR="#ffffff"><FONT FACE="Arial, Helvetica"><FONT SIZE=6>Message
Interception Techniques
</FONT></FONT></FONT>
</P>
</TD>
<TD>
<IMG SRC="logo-100.jpg" NAME="Graphic7" ALIGN=RIGHT WIDTH=102 HEIGHT=42 BORDER=0></TD>
</TR>
</TABLE>
<P>Even though object-orientation orders program structures around
data, objects are characterized primarily by their behavior.
Object-oriented programming style encourages the access of
encapsulated data only through the methods of an object, since this
enables data abstractions. A method invocation can be interpreted as a
message exchange between the calling and the called object.
Therefore, objects are at runtime only traceable through their message
exchanges. At this point the message interceptors can be applied to
catch and manipulate all incoming and outgoing messages of an
object.
<P>
</P>Generally interceptors can be applied to attach additional or
extrinsic concerns to an object or a class or a class hierarchy. For
instance roles or aspects can be implemented this way on various
levels of scale.
</P>
<P>We have already discussed some interception techniques
implicitly. E.g., the <tt>unknown</tt> mechanism intercepts messages
that have not be found on the object. It can be used as a very useful
programming technique, e.g., the define a default behavior for an
object. The interceptors presented in this section have a different
character: They are applied before/after the original method <em>even
if the method is defined for the target object</em>. Thus these
interception techniques may be applied
</P>
<P>We will discuss the message interceptors in this section in
detail. The table below gives an impression, when which interceptor
may be applied.
<P ALIGN=CENTER STYLE="margin-bottom: 0in"><STRONG>Message
Interceptors Overview</STRONG></P>
<CENTER>
<TABLE BORDER=1 CELLPADDING=2>
<TR>
<TD></TD>
<TD>
<P ALIGN=LEFT><strong>Applied When</strong</P>
</TD>
<TD>
<P ALIGN=LEFT><strong>Primary Target Structure</strong</P>
</TD>
<TD>
<P ALIGN=LEFT><strong>Coverage</strong</P>
</TD>
</TR>
<TR>
<TD>
<P ALIGN=LEFT><em> Per-Object Filter</em>
</P>
</TD>
<TD>
<P ALIGN=LEFT>before/after a call
</P>
</TD>
<TD>
<P ALIGN=LEFT> object
hierarchies
</P>
</TD>
<TD>
<P ALIGN=LEFT>all methods
</P>
</TD>
</TR>
<TR>
<TD>
<P ALIGN=LEFT><em> Per-Class Filter</em>
</P>
</TD>
<TD>
<P ALIGN=LEFT>before/after a call
</P>
</TD>
<TD>
<P ALIGN=LEFT> class and class
hierarchies
</P>
</TD>
<TD>
<P ALIGN=LEFT>all methods
</P>
</TD>
</TR>
</TBODY>
<TBODY>
<TR>
<TD>
<P ALIGN=LEFT><em> Per-Object Mixin</em>
</P>
</TD>
<TD>
<P ALIGN=LEFT> before/after a call
</P>
</TD>
<TD>
<P ALIGN=LEFT> object
</P>
</TD>
<TD>
<P ALIGN=LEFT> specific methods
</P>
</TD>
</TR>
<TR>
<TD>
<P ALIGN=LEFT><em> Per-Class Mixin</em>
</P>
</TD>
<TD>
<P ALIGN=LEFT> before/after a call
</P>
</TD>
<TD>
<P ALIGN=LEFT> class and class
hierarchies
</P>
</TD>
<TD>
<P ALIGN=LEFT> specific methods
</P>
</TD>
</TR>
<TR>
<TD>
<P ALIGN=LEFT><em> Unknown Mechanism</em>
</P>
</TD>
<TD>
<P ALIGN=LEFT> after method
was not found
</P>
</TD>
<TD>
<P ALIGN=LEFT> object
</P>
</TD>
<TD>
<P ALIGN=LEFT> all unknown calls
</P>
</TD>
</TR>
</TABLE>
</CENTER>
<br>
<H2><A NAME="filter"></A>Filter
</H2>
<P>The filter (see <a href="#xotcl-filter">[Neumann and Zdun
1999a]</a> for more details) is a language construct to implement
broader extensional concerns either for a single object or for several
classes or class hierarchies. This way large program structures at the
scale of several classes or class hierarchies can be managed. It is a
very general interception mechanism which can be used in various
application areas. E.g. a very powerful programming language support
for certain design patterns is easily achievable, but there are also
several other domains which are covered, like tracing of program
structures, self-documentation at run-time, re-interpretation of the
running program, etc.
</P>
<P>A <I>per-class filter</I> is a special instance method that is
registered for a class <I>C</I>. A <I>per-object filter</I> is a
special instance method that is registered for a object
<I>o</I>. Every time an object of class, <I>C</I> or the object
<I>o</I> respectively, receives a message,
the <I>filter</I> method is invoked automatically.
</P>
<H3><A NAME="filter_usage"></A>Usage of Filters
</H3>
<P>All messages to a filtered object must go through the filter before
they reach their destination object. A simple example would be a sole
filter on the class of the object. To define such a filter two steps
are necessary. Firstly an filter method has to be defined, then the
filter has to be registered. The filter method consists of three parts
which are all optional. A filter method has the following form:
</P>
<PRE><EM> ClassName instproc FilterName args {
pre-part
next
post-part
}
</em></PRE>
<P>
When a filter comes to execution at first the actions in the <EM>pre-part</EM>
are processed. The filter is free in what it does with the
message. Especially it can (a) pass the message, which was perhaps
modified in the <EM>pre-part</EM>, to other filters and finally to
the object. It can (b) redirect it to another destination. Or it can
(c) decide to handle the message on its own. The forward passing of
messages is implemented through the <TT>next</TT> primitive of XOTcl.
After the filter has passed its pre-part, the actual called method is
invoked through <TT>next</TT>.
</P>
<P>After the call of <TT>next</TT> is processed, the execution returns
to the point in the filter, where the <TT>next</TT> call is located
and resumes execution with the actions of the <EM>post-part</EM>.
These may contain arbitrary statements, but especially may take the
result of the actual called method (which is returned by the
next-call) and modify it. The caller then receives the result of the
filter, instead of the result of the actual called method.
</P>
<P>The pre- and post-part may be filled with any ordinary
XOTcl-statements. The distinction between the three parts is just a
naming convention for explanation purposes.
</P>
<P>The filter uses the <TT>args</TT> argument which lets us use a list of
variable length as arguments, since it must filter a lot of different
calls, which may have different argument lists. Furthermore, it may
pass through arguments to other filters and the preceding filters may
change the argument list.
</P>
<P>Since any proc/instproc may be a filter, a registration of the
filter is necessary, in order to tell XOTcl, which instprocs are
filters on which classes. The <TT>filter</TT> and <TT>instfilter</TT>
instance methods are able to handle this task for per-object filters
and per-class filters respectively. Similar to the XOTcl language
introduced so far, the filter registration is dynamic at run-time. By
supplying a new list of filters to
<TT>filter</TT>/<TT>instfilter</TT>, the programmer can change the
filters registered on a class at arbitrary times. The filter instance
method has the syntax:
</P>
<PRE STYLE="margin-bottom: 0.2in"><em> ClassName instfilter filterList</em></PRE>
for per-class filters and:
<PRE STYLE="margin-bottom: 0.2in"><em> objName filter filterList</em></PRE>
for per-object filters.
<P>
Now a simple example should show the filter's usage. In the preceding
examples we have defined several rooms. Every time a room action
occurs it is likely that the graphical sub-system has to change
something on the output of that particular room. Therefore, at first
we need a facility to be informed every time an action on a room
happens. This is quite easily done using filters:
</P>
<PRE> Class Room
Room r1; Room r2; # just two test objects
Room instproc roomObservationFilter args {
puts "now a room action begins"
set result [next]
puts "now a room action ends - Result: $result"
return $result
}
Room instfilter roomObservationFilter</PRE><P>
Now every action performed on room objects is notified with a pre-
and a post-message to the standard output stream. We return the
result of the actual called method, since we don't want to change the
program behavior at all. E.g. we can set an instance variable on both
of the two room objects:
</P>
<PRE> r1 set name "room 1"
r2 set name "room 2"</PRE><P>
The output would be:
</P>
<PRE> now a room action begins
now a room action ends - Result: room 1
now a room action begins
now a room action ends - Result: room 2</PRE><P STYLE="margin-bottom: 0in">
<BR>
</P>
<P ALIGN=CENTER STYLE="margin-bottom: 0in"><A NAME="oneFilter"></A><A NAME="718"></A>
  
</P>
<P ALIGN=CENTER STYLE="margin-bottom: 0in"><STRONG>Figure 4:</STRONG>
Cascaded Message Filtering</P>
<CENTER>
<TABLE WIDTH=480 BORDER=0 CELLPADDING=2 CELLSPACING=0>
<COL WIDTH=476>
<TR>
<TD WIDTH=476>
<P><FONT SIZE=1 STYLE="font-size: 2pt"><IMG SRC="cascaded-message-filter.gif" NAME="Graphic14" ALIGN=BOTTOM WIDTH=474 HEIGHT=281 BORDER=0></FONT></P>
</TD>
</TR>
</TABLE>
</CENTER>
<P><BR><BR>
</P>
<P>All classes may have more than one filter. In fact they may have a
whole filter chain, where the filters are cascaded through <TT>next</TT>.
The <TT>next</TT> method is responsible for the forwarding of
messages to the remaining filters in the chain one by one till all
pre-parts are executed. Then the actual method is executed and
then the post-parts come to turn. If one next-call is omitted the
chain ends in this filter method. As an example for an additional
filter we may register a filter that just counts the calls to
rooms.
</P>
<PRE> Room set callCounter 0 ;# set class variable
Room instproc counterFilter args {
[self class] instvar callCounter
incr callCounter
puts "the call number callCounter to a room object"
next
}
Room instfilter {roomObservationFilter counterFilter}</PRE><P>
Filters are invoked in registration order. The order may be changed
by removing them and adding them in new order. Filters are inherited
by sub-classes. E.g. in the preceding example for the next path, an
<TT>OvalOffice</TT> was derived from the <TT>Room</TT> class. Without
a change to the program each <TT>OvalOffice</TT> object automatically
produces the same filter output as rooms.
</P>
<P STYLE="margin-bottom: 0in"><BR>
</P>
<P ALIGN=CENTER STYLE="margin-bottom: 0in"><A NAME="filterInheritance"></A><A NAME="734"></A>
  
</P>
<P ALIGN=CENTER STYLE="margin-bottom: 0in"><STRONG>Figure 5:</STRONG>
Filter Inheritance</P>
<CENTER>
<TABLE WIDTH=2 BORDER=0 CELLPADDING=2 CELLSPACING=0>
<COL WIDTH=0>
<TR>
<TD>
<P><FONT SIZE=1 STYLE="font-size: 2pt"><IMG SRC="filter-inheritance.gif" NAME="Graphic15" ALIGN=BOTTOM WIDTH=508 HEIGHT=350 BORDER=0></FONT></P>
</TD>
</TR>
</TABLE>
</CENTER>
<P><BR>Filter chains can also be combined through (multiple)
inheritance using the <TT>next</TT> method. When the filter chain of
the object's class is passed, the filter chains of the superclasses
are invoked using the same precedence order as for inheritance. Since
on the subclass there may also be a another filter chain, without
sophisticated computing in the pre- and post-parts one can produce
easily a powerful tracing facility. E.g. if we want to distinguish an
<TT>OvalOffice</TT> from other rooms we may want to add a filter
solely for rooms of the type <TT>OvalOffice</TT>:
</P>
<PRE> Class OvalOffice -superclass Room
OvalOffice o1; # test object
OvalOffice instproc ovalOfficeObservationFilter args {
puts "actions in an oval office"
next
}
OvalOffice instfilter ovalOfficeObservationFilter</PRE><P>
A simple call to the <TT>o1</TT> object, like:
</P>
<PRE STYLE="margin-bottom: 0.2in"> o1 set location "Washington"</PRE><P>
produces the following output:
</P>
<PRE> actions in an oval office
now a room action begins
the call number 3 to a room object
now a room action ends - Result: Washington</PRE><P>
As seen already, filter registrations can be added dynamically at
runtime. But they may also be removed. Perhaps the counting on rooms
should stop after a while, then a simple call of the <TT>instfilter</TT>
method is sufficient:
</P>
<PRE STYLE="margin-bottom: 0.2in"> Room instfilter roomObservationFilter</PRE>
<P>Filters can be removed completely by giving an empty list to the
registration method:</P>
<PRE STYLE="margin-bottom: 0.2in"> Room instfilter {}</PRE>
<P> Per-object filters operate on a single object. E.g. if we only
want to observe a single Room object room1, we can use the filter
method to register the roomObservationFilter only for this particular
instance:</P>
<PRE STYLE="margin-bottom: 0.2in"> room1 filter roomObservationFilter</PRE>
<P> As a filter we can register any method in the precedence order of
the class or object. Thus we can also register procs as per-object
filters. Additionally, meta-class methods may be registered as
per-class filters. Filters are linearized so that each filter is only
executed once, even if it is registered multiple times.
</P>
<P><BR><BR>
</P>
<H3><A NAME="filter_info"></A>Introspection on Filters
</H3>
In order to gain information about the currently registered filters on
a certain object/class, the object info option <TT>filters </TT> and
the class info option <TT>instfilters </TT> may be
queried. It returns a list of the currently registered filters:
</P>
<PRE STYLE="margin-bottom: 0.2in"><em> ClassName info
instfilter</em></PRE>
<PRE STYLE="margin-bottom: 0.2in"><em> objName info filter</em></PRE>
<p>
A special call-stack info option for filters is <TT>self
filterreg</TT>. It returns the name of the object or class on which
the filter is registered. Since the filter may be registered on other
objects/classes than the one on which it is defined, this may vary from
<TT>self class</TT> in the filter.
The command returns a list of the form:
<PRE STYLE="margin-bottom: 0.2in"><em> objName filter filterName</em></PRE>
or:
<PRE STYLE="margin-bottom: 0.2in"><em> className instfilter filterName</em></PRE>
respectively.
</P>
<P><BR><BR>
</P>
<H3><A NAME="filter_trace"></A>A simple Trace Program Examples
</H3>
<P>The trace example primarily demonstrates the inheritance of filter
chains. Since all classes inherit from <TT>Object</TT>, a filter on
this class is applied on all messages to objects. The <TT>Trace</TT>
object encapsulates methods for managing the tracing:
</P>
<PRE> Object Trace
Trace set traceStream stdout
Trace proc openTraceFile name {
my set traceStream [open $name w]
}
Trace proc closeTraceFile {} {
close $Trace::traceStream
my set traceStream stdout
}
Trace proc puts line {
::puts $Trace::traceStream $line
}
Trace proc add classname {
$classname instfilter [concat [$classname info filter] traceFilter]
}</PRE><P>
First we define the object and set a variable for the stream to which
we send the trace outputs (here: stdout). With a method for opening
and a method for closing a file we can redirect the trace stream to a
file. <TT>puts</TT> is helper method for the filter to print an
output to the selected output stream. In <TT>add</TT> the <TT>traceFilter</TT>
is appended to the existing filters of a specified class. The actual
filter method (see below) displays the calls and exits of methods
with an according message. The calls are supplied with the arguments,
the exit traces contain the result values. We have to avoid the
tracing of the trace methods explicitly.
</P>
<PRE> Object instproc traceFilter args {
# don't trace the Trace object
if {[string equal [self] ::Trace]} {return [next]}
::set context "[self class]->[self callingproc]"
::set method [self calledproc]
switch -- $method {
proc -
instproc {::set dargs [list [lindex $args 0] [lindex $args 1] ...] }
default {::set dargs $args }
}
Trace::puts "CALL $context> [self]->$method $dargs"
::set result [next]
Trace::puts "EXIT $context> [self]->$method ($result)"
return $result
}</PRE><P>
As trace message we write the callee´s context (class and
proc), the invoked method (using <TT>calledproc</TT>), and the given
arguments. In the switch statement we avoid to print whole method
bodies.
</P>
<P>With
</P>
<PRE STYLE="margin-bottom: 0.2in"> Trace add Room</PRE><P>
messages to all rooms, including all instances of <TT>Room</TT>´s
sub-classes, are surrounded with a CALL and an EXIT output. With
</P>
<PRE STYLE="margin-bottom: 0.2in"> Trace add Object</PRE><P>
messages to all objects in an XOTcl environment are surrounded with a
CALL and an EXIT output. In general, it is possible to restrict the
trace to instances of certain classes, or to produce trace output for
only certain methods. This requires registration methods and a more
sophisticated implementation of the filter method.
</P>
<P><BR><BR>
</P>
<H3><A NAME="filter_guards"></A> Filter Guards</H3>
<P>
A filter guard is a set of conditions that determine whether a filter
is to be executed upon a certain call or not. Syntactically we can
append a filter guard to the filter registration or it can be
registered using the methods filterguard for filters and
instfilterguard for instfilters.
</P><P>
Each filter guard is an ordinary condition. A filter guard is executed
in the call frame of the filter to be executed, if the filter guard
returns 1. Thus, the callstack information are already set to the
values of the targeted filter.
</P><P>
Let us consider a simple filter guard as an example.
</P>
<PRE>
Room instfilter {
{loggingFilter {[self calledproc] == "open" || [self calledproc] == "close"}}
}
</PRE><P>
Here we limit the filter application of the logging filter on rooms
to calls to open and close. All other calls to requests are
not filtered at all. Actually, the above syntax is a short form of:
</P>
<PRE>
Room instfilter loggingFilter
Room instfilterguard {
[self calledproc] == "open" || [self calledproc] == "close"}
}
</PRE><P>
The filter guard language construct is registration centric. It only
applies for the class or object on which a filter is registered, not
for all applications of the filter method. E.g. if we use
loggingFilter on another class we may give no or completely
different filter guards.
</P><P>
If no filter guard is given for a filter, we assume that it is to be
applied on all methods (equivalent to the filter guard '1' which is
always true). In principal, a filter guard may be expressed in an
ordinary filter as well by starting the filter method with a condition:
</P>
<PRE>
Room instproc loggingFilter args {
if {!([self calledproc] == "open" ||
[self calledproc] == "close")} {
next
### filter code
}
</PRE><P>
There are several advantages of using filter guards instead of
ordinary if-statements in filters. First, filter guards are a more
handy syntax. Secondly, as we will see in this paper, filter guards
can be used to define conditions in a reusable way. In this paper, we
will use them to define reusable pointcuts. The filters themselves
become more reusable as well, since we can define them independently
from the conditions bundled with their registration. Thirdly, filter
guards offer a better performance because we do not have to call and
evaluate the filter method, if the filter guard returns
FALSE. Moreover, filter guards may call methods. To avoid recursive
filtering during the application of filter guards (which would limit
performance even more), filtering is disabled during execution of the
guard.
</P><P>
If we call a method during a filter, as for instance callsMethod:
</P>
<PRE>
Room instfilterguard {[my callsMethod openURL]}
</PRE><P>
we have to find out the setting of the call-stack
information in the method callsMethod from the
filter (here: calledproc is interesting).
This can be done using the getGuardedScope method which returns
the level of the filter scope that is guarded. With Tcl's uplevel
we can switch into this scope and get the relevant filter information
from there, e.g.:
</P>
<PRE>
Room instproc callsMethod {method} {
set level [my getGuardedScope]
set calledproc [uplevel $level self calledproc]
return [string match $calledproc $method]
}
</PRE><P>
Here, we first determine the guarded scope's level. Then we get the
called proc information from this scope. Finally, we check whether it
matches the given method name or not.
<H2><A NAME="mixins"></A>Mixin Classes
</H2>
<P>Per-object and per-class mixins (see <a
href="#xotcl-mixin">[Neumann and Zdun 1999c]</a> for more details) are
another interception technique of XOTcl to handle complex
data-structures dynamically. Here, we use mixin as a short form for
mixin class. All methods which are mixed into the execution of the
current method, by method chaining or through a mixin class, are
called <I>mixin methods</I>. Mixin classes resembles the filter
presented in the preceding section. While the filters work on all
calls to all methods of an object/class hierarchy, the mixin classes
are applied on specific methods. The filter is defined in a single
method, while the mixin is composes several method in a class.
</P>
<H3><A NAME="mixin_supplemental"></A>Supplemental Classes
</H3>
<P>Mixin classes cover a problem which is not solvable elegantly just
by the method chaining, introduced so far. To bring in an addition to
a class, the normal XOTcl way is to define a mixin method and chain
the methods through <TT>next</TT>, e.g.:
</P>
<PRE> Class Basic
Basic instproc someProc {
# do the basic computations
}
Class Addition
Addition instproc someProc {
# do the additional computations
next
}</PRE><P>
In order to mix-in the additional functionality of the <EM>supplemental</EM>
class <TT>Addition</TT> a new helper class (sometimes called
intersection class) has to be defined, like:
</P>
<PRE STYLE="margin-bottom: 0.2in">Basic+Addition -superclass Addition Basic</PRE><P>
This is even applicable in a dynamical manner, every object of the
class <TT>Basic</TT> may be changed to class <TT>Basic+Addition</TT>
at arbitrary times, e.g.:
</P>
<PRE> Basic+Addition basicObj
...
basicObj class Basic+Addition</PRE><P>
Now consider a situation with two addition classes. Then following
set of classes has to be defined to cover all possible combinations:
</P>
<PRE> Class Basic
Class Addition1
Class Addition2
Class Basic+Addition1 -superclass Addition1 Basic
Class Basic+Addition2 -superclass Addition2 Basic
Class Basic+Addition1+Addition2 -superclass Addition2 Addition1 Basic</PRE><P>
The number of necessary helper classes rises exponential. For <I>n</I>
additions, 2<I><SUP>n</SUP></I>-1(or their permutations if order
matters) artificially constructed helper-classes are needed to
provide all combinations of additional mix-in functionality.
Furthermore it is possible that the number of additions is unlimited,
since the additions may produce other additions as side-effects. This
demonstrates clearly that the sub-class mechanism provides only a
poor mechanism for mix-in of orthogonal functionality. Therefore we
provide an extension in the form of object mixin classes, which are
added in front of the search precedence of classes.
</P>
<H3><A NAME="mixin-usage"></A>Per-Object Mixins
</H3>
<P>The mix-ins methods extend the next-path of shadowed methods.
Therefore, per-object mix-in methods use the <TT>next</TT> primitive
to access the next shadowed method. Consider the following example:
</P>
<PRE> Class Agent
Agent instproc move {x y} {
# do the movement
}
Class InteractiveAgent -superclass Agent
# Addition-Classes
Class MovementLog
MovementLog instproc move {x y} {
# movement logging
next
}
Class MovementTest
MovementTest instproc move {x y} {
# movement testing
next
}</PRE><P>
An agent class is defined, which allows agents to move around. Some
of the agents may need logging of the movements, some need a testing
of the movements, and some both (perhaps only for a while). These
functionalities are achieved through the additional classes, which we
will apply through per-object mixins.
</P>
<P>Before we can use the per-object mix-ins on a particular object,
we must register the mixins on it with the <TT>mixin</TT> instance
method. It has the syntax:
</P>
<PRE STYLE="margin-bottom: 0.2in"><em> objName mixin mixinList</em></PRE><P>
For example we may create two interactive agents, where one is logged
and one is tested:
</P>
<PRE> InteractiveAgent i1; InteractiveAgent i2
i1 mixin MovementLog
i2 mixin MovementTest</PRE><P>
At arbitrary times the mixins can be changed dynamically. For example
<TT>i2</TT>'s movements can also be logged:
</P>
<PRE STYLE="margin-bottom: 0.2in"> i2 mixin MovementTest MovementLog</PRE><P STYLE="margin-bottom: 0in">
<BR>
</P>
<P ALIGN=CENTER STYLE="margin-bottom: 0in"><A NAME="per-obj-mixin"></A><A NAME="662"></A>
  
</P>
<P ALIGN=CENTER STYLE="margin-bottom: 0in"><STRONG>Figure 6:</STRONG>
Per-Object Mix-ins: Next-Path for the Example</P>
<CENTER>
<TABLE WIDTH=315 BORDER=0 CELLPADDING=2 CELLSPACING=0>
<COL WIDTH=311>
<TR>
<TD WIDTH=311>
<P><FONT SIZE=1 STYLE="font-size: 2pt"><IMG SRC="next-path-mixin-movement.gif" NAME="Graphic16" ALIGN=BOTTOM WIDTH=307 HEIGHT=187 BORDER=0></FONT></P>
</TD>
</TR>
</TABLE>
</CENTER>
<P><BR><BR>
</P>
<P>The <TT>mixin</TT> option of the <TT>info</TT> instance method
allows us to introspect the per-object mix-ins. It has the syntax:
</P>
<PRE STYLE="margin-bottom: 0.2in"><em> objName info mixins ?class?</em></PRE><P STYLE="margin-bottom: 0in">
It returns the list of all mix-ins of the object, if <TT>class</TT>
is not specified, otherwise it returns <TT>1</TT>, if <TT>class</TT>
is a mixin of the object, or <TT>0</TT> if not.
</P>
<P> Note, that the constructors (init methods) of per-object mixins (and per-class mixins)
are only called, if the mixin is registered already during object
initialization (when init is called). For per-object mixins, one can
achieve the initialization of a mixin via an idiom like <PRE>
<TT>Object o -mixin M -init</TT>
</PRE> that
registers the mixin before <tt>init</tt> is called. When a mixin is registered
after object creation and it needs initializations, it is necessary to
define special methods for this. Note, that the behavior described
here is introduced in version 0.84 to ensure consistent behavior of
intrinsic classes, per-object and per-class mixins, and to achieve
predictable behavior for dynamic registration for all kind of mixins,
and as well during recreations of objects having mixins
registered. Older versions used heuristics for the initialization of
per-object mixins.
</P>
<H3><A NAME="per-class-mixins"></A>Per-Class Mixins
</H3>
<P>Per-class mixins are exactly identical in their behavior to
per-object mixins, but they operate on classes. Thus they are the
class-specific variant of the per-object mixins, like instprocs are a
class-specific variant of procs. Therefore, in the language the
per-class mixins are called instmixins.
</p>
<P>
In general a per-class mixin is a class which is mixed into the
precedence order of all instances of the class and all its subclasses
it is registered for. It is also searched before the object's class
itself is searched, but after per-object mixins.
</p>
<P>
Per-class mixins are <em>linearized</em> according to the
<a href='#precedence order'>precedence order</a>
like classes on the superclass hierarchy. I.e. from the full
list of per-object mixins, per-class mixins, and intrinsic classes
(and all the superclasses of all these classes) always the last
occurrence is used.
</p>
<P>
From the point of view of language expressibility instmixins are not
required, because they cannot express anything that per-object mixins
cannot express already (like procs can express any instproc
feature). As alternative to instmixins, we could simply register the
per-object mixins in the constructor of the class.
</p>
<P>
But there at least the following reasons for instmixins as an
additional language construct:
<OL>
<LI> we can at runtime determine with <tt>info mixin</tt>
and <tt>info instmixin</tt> whether it is a class- or object-specific
mixin. Thus we get a better structuring at runtime.
<LI> We have not to 'pollute' the constructors with per-class mixin
registrations. Therefore, the constructors get more understandable.
<LI>If it is required to add (and remove) dynamically interceptors
to a set of objects, which are instances of a certain type, per-class
mixins are much easier to handle (e.g. add an instmixin to Object
to intercept e.g. all calls to certain predefined methods).
<LI>The language is more 'symmetrical', since any object-specific
feature in XOTcl has a class-specific variant.
</OL>
<P>
<P>The mix-ins methods of per-class mixins extend the next-path of
shadowed methods in the same way as per-object mixin methods. Before
we can use a per-class mix-in on a particular class, we must
register the mixin on it with the <TT>instmixin</TT> instance method. It
has the syntax:
</P>
<PRE STYLE="margin-bottom: 0.2in"><em>ClassName instmixin mixinList</em></PRE><P>
</P>
<P>
Now consider that in the given per-object mixin example all
interactive agents should be tested. We could either build a subclass
<tt>TestedInteractiveAgent</tt> or register the per-object mixin in
the constructor of the interactive agent class. The subclass solution
leads to the same combinatorial explosion of intersection classes as
discussed in the previous section, if more supplemental classes are
added. The per-object mixin solution pollutes the constructor and does
not prevail the structural semantics that the 'tested' property
belongs to the interactive agent class at runtime
</P>
<P>
Here, we can use a per-class mixin:
</P>
<PRE> Class Agent
Agent instproc move {x y} {# do the movement}
Class InteractiveAgent -superclass Agent
Class MovementTest
MovementTest instproc move {x y} {
# movement testing
next
}
# now register the instmixin
InteractiveAgent instmixin MovementTest
</PRE>
<P> The per-class mixin now operates on all interactive agent
including the instances of subclasses. E.g. for interactive agents
<TT>i1</TT> and <TT>i2</TT> we automatically have movement
testing. <TT>i2 </TT> is also logged, since it has the logging class
as object-specific mixin:
</P>
<PRE> InteractiveAgent i1
InteractiveAgent i2 -mixin MovementLog
i1 move 3 4
i2 move 1 2
</PRE>
<P>
At arbitrary times the instmixins can be changed dynamically.
</P>
<P>The <TT>instmixin</TT> option of the class <TT>info</TT> instance
method allows us to introspect the per-class mixins. It has the
syntax:
</P>
<PRE STYLE="margin-bottom: 0.2in"><em> ClassName info instmixins ?class?</em>
</PRE>
<P STYLE="margin-bottom: 0in">
It returns the list of all instmixins of the the class, if <TT>class</TT>
is not specified, otherwise it returns <TT>1</TT>, if <TT>class</TT>
is a mixin of the object, or <TT>0</TT> if not.
</P>
<p>Per-class mixins are applied transitively. That means the per-class
mixin A of a per-class mixin B is also applied for an object in in B's
scope. This is exactly the same as how superclasses are applied for
instances. Consider the following example</p>
<pre>
Class X11 \
-instproc test args {
puts [self class]
next
}
Class X12 \
-instproc test args {
puts [self class]
next
}
Class X \
-instmixin {X11 X12} \
-instproc test args {
puts [self class]
next
}
Class Y \
-instmixin X
Y create y -test
X create x -test
</pre>
<p> Here the application as a superclass (for x) yields the same
result as the application as an instmixin (for y):
<pre>
::X11
::X12
::X
</pre>
<H2>
<A NAME="callstack_info"></A>Callstack Information
</H2>
<P STYLE="margin-bottom: 0in">Since the presented interceptors are
normal XOTcl instprocs they can access all XOTcl introspection
abilities introduced so far. In instprocs all recent information is
accessible within their scope. But the interceptors are mechanisms,
which cover more then their sole scope. The meaningful usage of the
meta-programming abilities often requires to go further and to get
information from the caller's and the callee's scope (e.g for
delegation decisions). Therefore, we introduced rich call-stack
informations for the interceptors. Note, that these are also available
for ordinary methods, but the "called..." info options return empty
strings.
</P>
<P> All call-stack information are packed compactly into the
<tt>self</tt> primitive as additional options. Note, before XOTcl
version 0.84 these were implemented as a part of the <tt>info</tt>
method. They are part of the <tt>self</tt> command for conceptual
integrity: introspection options in <tt>info</tt> can be expected to
produce the same result, when they are not explicitly changed. In
contrast, all information provided by <tt>self</tt> are
call-stack dependent.
</P>
<P ALIGN=CENTER STYLE="margin-bottom: 0in"><STRONG>The additional
Call-stack Information Options in <tt>self</tt> </STRONG></P>
<CENTER>
<TABLE BORDER=1>
<TR>
<TD>
<P ALIGN=LEFT><TT>self calledproc</TT></P>
</TD>
<TD VALIGN=TOP>
<P ALIGN=LEFT>Returns the name of the proc which was invoked in
the original call.</P>
</TD>
</TR>
<TR>
<TD >
<P ALIGN=LEFT><TT>self calledclass</TT></P>
</TD>
<TD VALIGN=TOP>
<P ALIGN=LEFT>Returns the name
of the class which presumably (if no dynamic class change occurs
afterwards) is invoked in
the original call.</P>
</TD>
</TR>
<TR>
<TD >
<P ALIGN=LEFT><TT>self callingclass</TT></P>
</TD>
<TD VALIGN=TOP>
<P ALIGN=LEFT>Returns the name of the class from which the
call was invoked (if one exists, otherwise an empty
string).</P>
</TD>
</TR>
<TR>
<TD >
<P ALIGN=LEFT><TT>self callingproc</TT></P>
</TD>
<TD VALIGN=TOP>
<P ALIGN=LEFT>Returns the name of the proc from which the
call was invoked (if one exists, otherwise an empty
string).</P>
</TD>
</TR>
<TR>
<TD >
<P ALIGN=LEFT><TT>self callingobject</TT></P>
</TD>
<TD VALIGN=TOP>
<P ALIGN=LEFT>Returns the name of the object from which the
call was invoked (if one exists, otherwise an empty
string).</P>
</TD>
</TR>
<TR>
<TD >
<P ALIGN=LEFT><TT>self filterreg</TT></P>
</TD>
<TD VALIGN=TOP>
<P ALIGN=LEFT>In a filter: returns the name
of the object/class on which the filter is registered. Returns either
'objName filter filterName' or 'className instfilter filterName'.</P>
</TD>
</TR>
<TR>
<TD >
<P ALIGN=LEFT><TT>self next</TT></P>
</TD>
<TD VALIGN=TOP>
<P ALIGN=LEFT>Return the
"next" method on the path as a string, i.e. the method which will be
called by [next].</P>
</TD>
</TR>
</TABLE>
</CENTER>
</P>
<P>Note, that three options with the prefix <TT>calling</TT>
represent the values of <TT>self</TT>, <TT>self proc</TT>, and <TT>self
class</TT> in the scope where the original call was invoked. In the
following section we will show a simple program in which all of the
<TT>info</TT> options have different values.
<H3><A NAME="filter_info_example"></A><BR>Filter Call-stack Information Example
</H3>
<P>Now we discuss a simple example that shows that all filter
introspection options may have different values:
</P>
<PRE> Class InfoTrace
InfoTrace instproc infoTraceFilter args {
puts "SELF: [self]"
puts "SELF PROC: [self proc]"
puts "SELF CLASS: [self class]"
puts "INFO CLASS: [my info class]"
puts "CALLED PROC: [self calledproc]"
puts "CALLING PROC: [self callingproc]"
puts "CALLING OBJECT: [self callingobject]"
puts "CALLING CLASS: [self callingclass]"
puts "REGISTRATION CLASS: [self filterreg]"
next
}
Class CallingObjectsClass
CallingObjectsClass callingObject
Class FilterRegClass -superclass InfoTrace
Class FilteredObjectsClass -superclass FilterRegClass
FilteredObjectsClass filteredObject
CallingObjectsClass instproc callingProc args {
filteredObject set someVar 0
}
FilterRegClass instfilter infoTraceFilter</PRE><P>
The invocation of <TT>callingObject callingProc</TT> produces the
following output:
</P>
<PRE> SELF: ::filteredObject
SELF PROC: infoTraceFilter
SELF CLASS: ::InfoTrace
INFO CLASS: ::FilteredObjectsClass
CALLED PROC: set
CALLING PROC: callingProc
CALLING OBJECT: ::callingObject
CALLING CLASS: ::CallingObjectsClass
REGISTRATION CLASS: ::FilterRegClass instfilter infoTraceFilter</PRE><P>
The filter reports for <TT>self</TT> the value <TT>filteredObject</TT>,
since this is the object on which the <TT>set</TT> call is invoked;
<TT>infoTraceFilter</TT> is the method of the filter, and therefore,
the actual proc, while the actual class is <TT>InfoTrace</TT>, the
filter's class. The class of the actual object is
<TT>FilteredObjectsClass</TT>.
</P>
<P>The called procedure is <TT>set</TT>. While the program stays in a
XOTcl-instproc all calling-info-options are set, the calling
procedure is <TT>callingProc</TT>, the calling class is the class,
where the method is defined (namely <TT>CallingObjectsClass</TT>),
and the object from which the call invoked is <TT>callingObject</TT>.
</P>
<P>Since the filter's registration class differs from the class,
where it is defined, the corresponding information is still missing
(in this example <TT>FilterRegClass</TT>).
</P>
<H2><A NAME="precedence order"></A>Precedence Order
</H2>
<P>The precedence order is composed by the precedence order of the
superclass hierarchy (as explained earlier) and the message
interceptors. In general, filters precede mixins and the superclass
hierarchy. They are applied in the order of the next path of the
object. Thus per-object filters are ordered before per-class
filters.</p>
<p>Mixins are processed after the filters. Again, they are applied in
the order of the next path of the object. Thus per-object mixins are
ordered before per-class mixins.</p>
<p>Finally, the object's own heritage order comes in the order: object,
class, superclasses.</p>
<p>The three precedence order lists (filters, mixins, and classes) are
pre-calculated and cached.</p>
<p>Filters as well as classes (mixins and ordinary classes) are
linearized. That means, each filter and each class can be only once on
a precedence order list. If a filter or class can be reached more than
once, than the last occurrence is used.</p>
<p>For instance, consider a class A is superclass, per-class mixin,
and per-object mixin. On the precedence order lists only the last
occurrence as a superclass is used after linearization.</P>
<!-- PAGE BREAK -->
<TABLE COLS=2 WIDTH=100% BORDER=0 CELLPADDING=2 CELLSPACING=0 BGCOLOR="#000055">
<TR>
<TD WIDTH=75%>
<P><A NAME="nesting"></A><FONT COLOR="#ffffff"><FONT FACE="Arial, Helvetica"><FONT SIZE=6>Class
Nesting and Dynamic Object Aggregations </FONT></FONT></FONT>
</P>
</TD>
<TD>
<IMG SRC="logo-100.jpg" NAME="Graphic9" ALIGN=RIGHT WIDTH=102 HEIGHT=42 BORDER=0></TD>
</TR>
</TABLE>
<P>Most object-oriented analysis and design methods are based on the
concepts of generalization and aggregation. Generalization is
achieved through class hierarchies and inheritance, while static
aggregation is provided through embedding. Since version 8.0 Tcl
offers a namespace concept which can be used as a mechanism to
provide dynamic aggregations.
</P>
<P>A <EM>namespace</EM> provides an encapsulation of variable and
procedure names in order to prevent unwanted name collisions with
other system components. Each namespace has a unique identifier which
becomes part of the fully qualified variable and procedure names.
Namespaces are therefore already object-based in the terminology of
Wegner. OTcl is object-oriented since it offers classes and class
inheritance. Its objects are also namespaces, but an object is more
than only a namespace. Therefore, two incompatible namespace concepts
have existed in OTcl in parallel.
</P>
<P>Extended OTcl combines the namespace concept of Tcl with the
object concept of OTcl. Every object and every class in XOTcl is
implemented as a separate Tcl namespace. The biggest benefit of this
design decision aside from performance advantages is the ability to
aggregate objects and nest classes. Contrary in OTcl every object has
a global identifier. Through the introspection abilities of
namespaces nested classes are also traceable at runtime and can be
changed dynamically. In XOTcl objects are allowed to contain nested
objects, which are dynamically changeable aggregates of the
containing object.
</P>
<H2><A NAME="nested_classes"></A>Nested Classes
</H2>
<P>The notation for nested classes follows the syntax of Tcl
namespaces by using ``::'' as a delimiter. For example the
description of a oval carpet and a desk can nest inside of the
<TT>OvalOffice</TT> class:
</P>
<PRE> Class OvalOffice
# general carpet
Class Carpet
Class OvalOffice::Desk
# special oval carpet - no name collision
Class OvalOffice::Carpet -superclass ::Carpet</PRE><P>
Nested classes can be used exactly like ordinary classes, a user can
sub-class it, derive instances, etc. The information about the
nesting structure of classes is available through the <TT>info</TT>
instance method:
</P>
<PRE><EM> ClassName info classchildren
ClassName info classparent
</em></PRE><P>
The <TT>classchildren</TT> option returns a list of children, if one
or more exist, otherwise it returns an empty string. <TT>classparent</TT>
results in the name of the parent class, if the class is nested.
Since nested classes are realized through namespaces, all
functionality offered by Tcl's <TT>namespace</TT> command is usable
from XOTcl as well.
</P>
<H2><A NAME="obj-agg"></A>Dynamic Object Aggregations
</H2>
<P>The nested classes only provide an aggregation of the descriptive
not of the runtime properties of an object. We have pointed out the
difference of object and class in XOTcl. Because of the splitting of a
class into class and class-object it is possible to give each object
its own namespace. The internal implementation of objects enable them
to contain nested objects, which are aggregates of the containing
object. In XOTcl these can be changed dynamically and introspected
through the language support of dynamic object aggregations <a
href="#xotcl-aggregation">[Neumann and Zdun 2000b]</a>. Suppose an
object of the class <TT>Agent</TT> should aggregate some property
objects of an agent, such as head and body:
</P>
<PRE> ClassAgent
Agent myAgent
Class Agent::Head
Class Agent::Body
Agent::Head ::myAgent::myHead
Agent::Body ::myAgent::myBody</PRE><P>
Now the objects <TT>myHead</TT> and <TT>myBody</TT> are part of the
<TT>myAgent</TT> object and they are accessible through a
qualification using ``::'' (or through Tcl's namespace command). But
in the common case they will be accessed, as introduced so far: the
explicit full qualification is not necessary when such variables are
being accessed from within XOTcl methods, since the object changes to
its namespace.
</P>
<P>The information about the part-of relationship of objects can be
obtained exactly the same way as for classes through the
<TT>info</TT> interface:
</P>
<PRE><em> objName info children
objName info parent
</em></PRE>
<H2><A NAME="nest_agg"></A>
Relationship between Class Nesting and Object Aggregation
</H2>
<P>The classes <TT>Head</TT> and <TT>Body</TT> are children of the
<TT>Agent</TT> class. It is likely that all agents, interactive or
not, have properties for head and body. This implies a static or
predetermined relationship between class nesting and object
aggregation. Such predetermination do not exist in XOTcl, but are
simply build, when specifying the relationship in the constructor,
e.g.:
</P>
<PRE> Agent instproc init args {
::Agent::Head [self]::myHead
::Agent::Body [self]::myBody
}</PRE><P>
Now all agents derived from the class have the two property objects
aggregated after creation. But still they are changeable in a
dynamical manner, e.g. with:
</P>
<PRE> Agent myAgent
myAgent::myHead destroy</PRE><P>
The agent turns into a headless agent. In companion of the
introspection mechanisms such constructions could be very useful.
Suppose, that in the virtual world the agents heads may be slashed
from their bodies. The graphical system simply needs to ask with <TT>info
children</TT> on the agent's object, whether it has a head or not and
can choose the appropriate graphical representation.
</P>
<P STYLE="margin-bottom: 0in">Note, that the not existing
relationship means a great deal of freedom and dynamics, which goes
together with the ideas behind OTcl, e.g. like the renunciation of
protection mechanisms. This policy in programming language design
means, on the one hand, ease of programming and more expressiveness,
but, on the other hand, it contains no protection against bad
software architectures or programming style. We believe that no such
mechanisms could hinder the programmer to do silly things, so our
policy was, to give the programmer rather more powerful constructs
then to make decisions in his place.
</P>
<H2><A NAME="copy_move"></A>Copy/Move
</H2>
Often an object has to be copied/moved. This is a very useful
functionality when XOTcl should be used as a prototyping language.
The XOTcl method <TT>move</TT> provides this functionality. Another
common behavior is implemented by the <tt>copy</tt> method which
clones the actual object to a destination object. The two methods have
the syntax:
<PRE STYLE="margin-bottom: 0.2in"><em> objName move destination </em></PRE><P>
<PRE STYLE="margin-bottom: 0.2in"><em> objName copy destination</em></PRE><P>
Copy and move operations work with all object/class information, i.e.,
information on filters, mixins, parameters, etc. are
automatically copied. Copy and move are integrated with class nesting
and object aggregations. All copy/move operations are deep copy
operations: all nested objects/classes are automatically copied/moved,
too.
E.g. if we want to reuse an imperial march object of star wars for
star wars 2, we can just copy the object:
<PRE>
starWars::imperialMarch copy starWars2::imperialMarch
</PRE>
<TABLE COLS=2 WIDTH=100% BORDER=0 CELLPADDING=2 CELLSPACING=0 BGCOLOR="#000055">
<TR>
<TD WIDTH=75%>
<P><A NAME="assertions"></A><FONT COLOR="#ffffff"><FONT FACE="Arial, Helvetica"><FONT SIZE=6>Assertions
</FONT></FONT></FONT>
</P>
</TD>
<TD>
<IMG SRC="logo-100.jpg" NAME="Graphic10" ALIGN=RIGHT WIDTH=102 HEIGHT=42 BORDER=0></TD>
</TR>
</TABLE>
<P>In order to improve reliability and self documentation we added
assertions to XOTcl. The implemented assertions are modeled after the
``design by contract'' concept of Bertrand Meyer. In XOTcl assertions
can be specified in form of formal and informal pre- and
post-conditions for each method. The conditions are defined as a list
of and-combined constraints. The formal conditions have the form of
normal Tcl conditions, while the informal conditions are defined as
comments (specified with a starting ``<TT>#</TT>''). The lists
containing the pre- and post-conditions are appended to the method
definition (see example below).
</P>
<P>Since XOTcl offers per-object specialization it is desirable to
specify conditions within objects as well (this is different to the
concept of Meyer). Furthermore there may be conditions which must be
valid for the whole class or object at any visible state (that means
in every pre- and post-condition). These are called invariants and
may be defined with following syntax for class invariants:
</P>
<PRE STYLE="margin-bottom: 0.2in"><em> ClassName instinvar invariantList</em></PRE><P>
or for objects invariants:
</P>
<PRE STYLE="margin-bottom: 0.2in"><em> objName invar invariantList</em></PRE><P>
Logically all invariants are appended to the pre- and post-conditions
with a logical ``and''. All assertions can be introspected.
</P>
<P>Since assertions are contracts they need not to be tested if one
can be sure that the contracts are fulfilled by the partners. But for
example when a component has changed or a new one is developed the
assertions could be checked on demand. For this purpose the <TT>check</TT>
method can be used either to test the pre- or the post-conditions.
The syntax is:
</P>
<PRE STYLE="margin-bottom: 0.2in"><em> objName check ?all? ?instinvar? ?invar? ?pre? ?post?</em></PRE><P>
Per default all options are turned off. <TT>check all</TT> turns all
assertion options for an object on, an arbitrary list (maybe empty)
can be used for the selection of certain options. Assertion options
are introspected by the <TT>info check</TT> option. The following
class is equipped with assertions:
</P>
<PRE> Class Sensor -parameter {{value 1}}
Sensor instinvar {
{[regexp {^[0-9]$} [my value]] == 1}
}
Sensor instproc incrValue {} {
my incr value
} {
{# pre-condition:}
{[my value] > 0}
} {
{# post-condition:}
{[my value] > 1}
}</PRE><P>
The <TT>parameter</TT> instance method defines an instance variable
<TT>value</TT> with value <TT>1</TT>. The invariant expresses the
condition (using the Tcl command <TT>regexp</TT>), that the value
must be a single decimal digit. The method definition expresses the
formal contract between the class and its clients that the method
<TT>incrValue</TT> only gets input-states in which the value of the
variable <TT>value</TT> is positive. If this contract is fulfilled by
the client, the class commits itself to supply a post-condition where
the variable's value is larger than 1. The formal conditions are
ordinary Tcl conditions. If checking is turned on for sensor <TT>s</TT>:
</P>
<PRE STYLE="margin-bottom: 0.2in">s check all</PRE><P>
the pre-conditions and invariants are tested at the beginning and the
post-condition and invariants are tested at the end of the method
execution automatically. A broken assertion, like calling <TT>incrValue</TT>
9 times (would break the invariant of being a single digit) results
in an error message.
</P>
<p>
In assertions we do not check methods that modify or introspect
assertions. These are
<tt>check</tt>,<tt>info</tt>,<tt>proc</tt>,<tt>instproc</tt>,<tt>invar</tt>,
and <tt>instinvar</tt>. The reason for this is that we want to be able
to recover a malicious action in a <tt>catch</tt> error handler, like:
</P>
<pre>
...
if {[catch {my assertionBreakingAction} errMsg]} {
puts "CATCHED ERROR: $errMsg"
# remember checking options, for turning them on later again
set check [my info check]
my check {}
# recover from broken assertion
...
# turning checking on again
$fb check $check
}
</pre>
<!-- PAGE BREAK -->
<TABLE COLS=2 WIDTH=100% BORDER=0 CELLPADDING=2 CELLSPACING=0 BGCOLOR="#000055">
<TR>
<TD WIDTH=75%>
<P><A NAME="meta-data"></A><FONT
COLOR="#ffffff"><FONT FACE="Arial, Helvetica"><FONT SIZE=6>Meta-Data
and Automatic Documentation
</FONT></FONT></FONT>
</P>
</TD>
<TD>
<IMG SRC="logo-100.jpg" NAME="Graphic11" ALIGN=RIGHT WIDTH=102 HEIGHT=42 BORDER=0></TD>
</TR>
</TABLE>
<P>To enhance the understandability and the consistency between
documentation and program it is useful to have a facility to make the
documentation a part of the program. There are several kinds of
meta-data which are interesting for a class, e.g. the author, a
description, the version, etc.
</P>
<P>
Older versions of XOTcl have contained a special meta-data command
<tt>metadata</tt>. This command is now (from version 0.83) deprecated
and replaced by an integrated solution with XOTcl's API documentation
functionality. The object <tt>@</tt> is used for documentation and
metadata issues. Per default it is not evaluated at all. Everything
that is send to <tt>@</tt> is simply ignored. That way we do not waste
memory/performance at runtime, if we do not require to parse the
metadata/documentation.
</P>
<P>
If we have to know the meta-data/documentation, as for instance in the
<tt>xoDoc</tt> component and the <tt>makeDoc</tt> tool, that handle
XOTcl's internal documentation, we have to re-define the documentation
object. Alternatively, we can partially parse the source code for
<tt>@</tt> commands.
</P>
<P>
With <tt>@</tt> the meta-data/documentation is handled by first class
XOTcl objects. By defining alternate @ implementations - as in
<tt>xoDoc</tt>/<tt>makeDoc</tt> - we can evaluate the
meta-data/documentation arbitrarily. <tt>xoDoc</tt>/<tt>makeDoc</tt>
are only an HTML back-end, but the basic idea is to provide support for
several other usages as well (e.g. XML, RDF, on-line help,
documentation of dynamic structures, etc).
</P>
<P>
The object<tt>@</tt> handles comments via its <tt>unknown</tt>
method. <tt>xoDoc</tt> adds the appropriate instprocs to t<tt>@</tt> to produce HTML
output. The appropriate command is:
</P>
<PRE STYLE="margin-bottom: 0.2in">
tclsh src/lib/makeDoc.xotcl <DOCDIR> <DOCFILES>
</PRE><P>
The source of a documentation is structurally very similar to the
XOTcl constructs being commented. E.g. one can copy an instproc and
add comments at the right places, like:
</P>
<PRE STYLE="margin-bottom: 0.2in">
Class C
C instproc m {a1 a2} {
return [expr {$a1+$a2}]
}
</PRE><P>
can be commented as follows
</P>
<PRE STYLE="margin-bottom: 0.2in">
@ Class C { description { "my sample class"} }
@ C instproc m {a1 "first number" a2 "second number"} {
description "add two numbers"
return "sum of a1 and a2"
}
</PRE></P>
<P>
One can do essentially a copy+paste of the source and add the
comments via attribute value pairs. Every basic language construct
can have a "description". If you want to include other properties to
the description, you can add them like:
</P>
<PRE STYLE="margin-bottom: 0.2in">
@ C instproc m {a1 "first number" a2 "second number"} {
author "GN+UZ"
date "Feb 31"
description "add two numbers"
return "sum of a1 and a2"
}
</PRE><P>
This way, author and date are added automatically to the generated
HTML file.
In addition, there is a <tt>@File</tt> hook for a per file
description, like:
</P>
<PRE STYLE="margin-bottom: 0.2in">
@ @File {
description {
This is a file which provides a regression test
for the features of the XOTcl - Language.
}
}
</PRE><P>
<TABLE COLS=2 WIDTH=100% BORDER=0 CELLPADDING=2 CELLSPACING=0 BGCOLOR="#000055">
<TR>
<TD WIDTH=75%>
<P><A NAME="additional-functionalities"></A><FONT COLOR="#ffffff"><FONT FACE="Arial, Helvetica"><FONT SIZE=6>Additional
Functionalities </FONT></FONT></FONT>
</P>
</TD>
<TD>
<IMG SRC="logo-100.jpg" NAME="Graphic12" ALIGN=RIGHT WIDTH=102 HEIGHT=42 BORDER=0></TD>
</TR>
</TABLE>
<H2><A NAME="abstract-classes"></A>Abstract Classes
</H2>
<P>In XOTcl a class is defined abstract if at least one method of
this class is abstract. The instance method <TT>abstract</TT> defines
an abstract method and specifies its interface. Direct calls to
abstract methods produce an error message. E.g. a <TT>Storage</TT> class
provides an abstract interface for access to different storage forms:
</P>
<PRE> Class Storage
Storage abstract instproc open {name}
Storage abstract instproc store {key value}
Storage abstract instproc list {}
Storage abstract instproc fetch key
Storage abstract instproc close {}
Storage abstract instproc delete {k} </PRE><P>
All kinds of storage have to implement every method from the
interface. E.g. a GNU Database Access, a relational database access,
and several other storage forms may be derived by sub-classing
(therefore, all conform to the same storage access interface).
</P>
<H2><A NAME="parameter"></A>Parameter</H2>
<P>Classes may be equipped with <TT>parameter</TT> definitions which
are automatically created for the convenient setting and querying of
instance variables. Parameters may have a default value, e.g.:
</P>
<PRE> Class Car -parameter {
owner
{doors 4}
}</PRE><P>
Each instance of class <TT>Car</TT> gets two instance variables
defined. <TT>owner</TT> has no default value, and <TT>doors</TT>
defaults to 4. E.g. the following defines a new person object with
the two parameters set:
</P>
<PRE STYLE="margin-bottom: 0.2in"> Car mercedes</PRE><P>
Additionally the <TT>parameter</TT> method automatically creates a
new getter/setter instance method for each parameter -- same named to
the parameter, which queries the parameter if it no argument is given
or sets the parameter if with an given argument. E.g. a car with only
two doors can be created by:
</P>
<PRE STYLE="margin-bottom: 0.2in"> Car porsche -doors 2</PRE><P>
The owner of the first car is set by:
</P>
<PRE STYLE="margin-bottom: 0.2in"> mercedes owner Marion</PRE><P>
and the doors of the first car can be queried (and printed to the
screen) by:
</P>
<PRE STYLE="margin-bottom: 0.2in"> puts "The mercedes got [mercedes doors] doors and is owned by [mercedes owner]"</PRE><P>
<TT>parameter</TT> are inherited by subclasses. The parameters
specified in the class hierarchy are combined, default values
can be redefined. Example:
<PRE STYLE="margin-bottom: 0.2in">
Class Car -parameter {{doors 4} owner}
Class SportsCar -superclass Car -parameter {{doors 2}}
Class Limo -superclass Car
Class Porsche -superclass SportsCar
Class Mercedes -superclass Limo
Porsche p1 -owner peter
Mercedes m1 -owner marion
puts "[p1 owner]'s [p1 info class] has [p1 doors] doors"
puts "[m1 owner]'s [m1 info class] has [m1 doors] doors"
</PRE>
<p>
An entry of the list specified for <tt>parameter</tt> is either
<ul>
<li>
a single element denoting the name of the parameter (i.e. a method
to be defined to set (or get) an instance variable with the same
name, or it might be</li>
<li>
a list of two elements denoting the name and a default value,
or it might be </li>
<li>
a list of more than two elements (specification list), where it is
possible to call methods and to specify additional information about
the parameter to be set. For example, the following definition
<PRE STYLE="margin-bottom: 0.2in">
Class Car -parameter {{doors 4} owner}
</PRE>
is actually a short form for
<PRE STYLE="margin-bottom: 0.2in">
Class Car -parameter {{doors -default 4} owner}
</PRE>
</li>
</ul>
<H4>Objects as Parameter</H4>
<p>In the specification list of a parameter it is possible to specify
a Class; Without the <tt>-Class</tt> given, a parameter denotes solely
to an instance variable; when <tt>-Class</tt> is given an child object
of the specified class is created, and the corresponding instance
variable keeps the name of the object. The classes to be used for
parameters must be instances of the meta class
<tt>::xotcl::Class::Parameter</tt>.
</p>
<p>The following example shows how to define and use parameters of
Class <tt>Point</tt> and <tt>Rectangle</tt>
<PRE STYLE="margin-bottom: 0.2in">
::xotcl::Class::Parameter Point -parameter {{x 0} {y 0} {z 0}}
::xotcl::Class::Parameter Rectangle -parameter {
color
{leftCorner -Class "Point -x 0 -y 0" -default 1}
{rightCorner -Class "Point -x 10 -y 10" -default 1}
}
Class Container -parameter {
name
{startPoint -Class Point}
{endPoint -Class Point}
{r -Class "Rectangle -color green" -default 1}
}
Container create c1 \
[list -startPoint -x 1 -y 2] [list -endPoint -z 3]
</PRE>
Similar to parameters without <tt>-Class</tt> they are created either
when a default value is given or when they are specified during
creation of the object. The default value is passed to the constructor
of the parameter object. Therefore, upon creation of the Container
in the example above, the following objects are created:
<ul>
<li>the Container <tt>c1</tt></li>
<li>the Point <tt>startPoint</tt> as a child object of c1
(since it was specified in the creation of c1)</li>
<li>the Point <tt>endPoint</tt> as a child object of c1
(since it was specified in the creation of c1)</li>
<li>the Rectangle <tt>r</tt> as a child object of c1
(since it has a default value)</li>
<li>the Points <tt>leftCorner</tt> and <tt>rightCorner</tt>
as a child object of the rectangle
(since they have default values).</li>
</ul>
As all other parameters, parameters with classes can be
reconfigured as well; the following commands sets the
z coordinate of the <tt>endPoint</tt> of the container <tt>c1</tt>
to 1.
<PRE STYLE="margin-bottom: 0.2in">
c1 endPoint -z 1
</PRE>
<H4>Setter and Getter Methods for Parameter</H4>
By default, the methods defined to access the parameter values are
instcommands implemented in C. However, it is possible to specify
custom setter and getters that might perform additional tasks. There
are two ways to specify custom setter/getter methods for parameters:
<ul>
<li>(a) the custom setter/getter methods can be defined as methods
within the class hierarchy of the object, or </li>
<li>(b) the custom getter/setter can
be specified on a different object. The set and get calls
are delegated to that object, which might be e.g. a
database instance. </li>
</ul>
In both cases when the custom getters and/or setter are defined they
will be called automatically from the standard setter/getter methods.
In order to use approach (a) the parameter methods -getter
and -setter can be used to specify the custom getter and
and setter methods:
<PRE STYLE="margin-bottom: 0.2in">
Class C -parameter {{a -setter myset -getter myget}}
</PRE>
The methods myset and myget are called like set with
one or two arguments. They are responsible for setting and
retrieving the appropriate values. It is possible to
specify any one of these parameter methods. In the following
example "<TT>c1 myset a 100</TT>" will be called by the
first line to set the value of a, "<TT>c1 myget a</TT>"
will be called by the second line to obtain the value of a.
<PRE STYLE="margin-bottom: 0.2in">
C c1 -a 100
c1 a
</PRE>
In order to use approach (b) a parameter method <tt>-access</tt> is
used to specify an object responsible for setting/getting the
parameter's values. This has the advantage that the custom getter and
setter methods can be inherited from a separate class hierarchy, such
they can used for any object without cluttering its interface.
<p>
In order to keep the parameter specification short the access
object can contain instance variables setter or getter, naming its
the setter/getter methods. If these instance variables are not
in the access object, "set" is used per default for getter and
setter. These default values can be still overridden by the
parameter methods -setter or -getter. Here is a simple example
showing this mechanism.
<PRE STYLE="margin-bottom: 0.2in">
Object db
db set setter myset
db set getter myget
db proc myset {o var value} { my set $var $value }
db proc myget {o var} { my set $var }
Class D -parameter {{x -access db}}
D d1
d1 x 100
puts x=[d1 x],vars=[db info vars]
</PRE>
Note that myset and myget obtain the name of the object as well,
such they can set theses instance variables in the object if
desired.
<H4>Alternative Parameter Classes</H4>
<p>For further customization, the class <TT>Class::Parameter</TT>
containing the described behavior can be as well extended (and
sub-classed).
The basic idea is to make the parameter mechanism extensible in a
similar way as the extension mechanisms work for normal object-oriented
methods. One can extend the predefined <tt>Class::Parameter</TT>
class with <tt>someInstproc</TT> and use later
</p>
<PRE STYLE="margin-bottom: 0.2in">
C c1 {{a -default 1 -someInstproc x} ...}
</pre>
<p>
or subclass it like:
</p>
<PRE>
Class MyParameter -superclass Class::Parameter
Class X -parameterclass MyParameter -parameter ...
</pre>
<p>
Upon object initialization, the parameters are firstly evaluated for
all mixins and then for the class hierarchy. E.g. in the following
example:
</p>
<PRE STYLE="margin-bottom: 0.2in">
Class A -parameter {
{pcm 1}
}
Class B -instmixin A -parameter {
{cl 4}
}
B b
</pre>
<p>
at first the mixin parameter 'pcm' is set, then the class parameter
'cl'. However, since parameters are applied before the '-' methods,
the per-object mixin parameter defaults in the following example are
not set by the standard initialization routine:
</p>
<PRE STYLE="margin-bottom: 0.2in">
Class C -parameter {
{pom 1}
}
B b -mixin C
</pre>
<p>
If this parameter or any other parameter default, which is introduced
later than the standard initialization routine, is required, then we
can evaluate the parameter defaults manually, like:
</p>
<PRE STYLE="margin-bottom: 0.2in">
[B info parameterclass] searchDefaults b
</pre>
<p>
This searches all default values for the object b which are defined on
mixins or on the class hierarchy.
</P>
<H2><A NAME="cmdCheck"></A>Checking Commands for being Objects,
Classes, or Meta-Classes
</H2>
<P>Since XOTcl is a hybrid language containing several Tcl commands,
sometimes its necessary for applications to distinguish between Tcl
commands and object commands for XOTcl. </TT>method of the
<TT>Object</TT> class looks up an <TT>objName</TT> and returns 1 if it
is an object and 0 if not:
<PRE STYLE="margin-bottom: 0.2in"><em> anyXOTclObject isobject objName</em></PRE><P>
If one can be sure that a command represents an
object, it might be unsure if the command is only an object or also
class or even meta-class. The two instance methods <TT>isclass</TT>
and <TT>ismetaclass</TT> check in the same manner, whether a class or
meta-class is given (since ever XOTcl class is an object, they also
return 0, when objName is not an XOTcl object).
<PRE><em>
anyXOTclObject isclass objName
anyXOTclObject ismetaclass objName
</em></PRE>
<H2>
<A NAME="Exit Handler"></A>Exit Handler
</H2>
<P>A task for a programming language, sometimes of similar importance
as object creation, is the object destruction. XOTcl ensures that all
objects are destroyed and their destructors are invoked when XOTcl
applications terminate. For that reason objects and classes are
destroyed in the order objects, classes, meta-classes. Sometimes
further destruction order is of importance. For these cases, the XOTcl
language provides an exit handler, which is a user-defined proc, which
invokes user-defined exit handling just before the destruction of
objects, classes, meta-classes is invoked. For instance, the exit
handler lets the user specify objects which have to be destroyed
before all other objects.
</P>
<P> The exit handler is defined as a proc of <TT>Object</TT>, which is per default empty:
<PRE>
::xotcl::Object proc __exitHandler {} {
# clients should append exit handlers to this proc body
;
}
</PRE>
<P> There are some procs of the <TT>Object</TT> class pre-defined,
which let us specify an exit handler conveniently:
</P>
<PRE> Object setExitHandler body
Object getExitHandler
Object unsetExitHandler
</PRE><P STYLE="margin-bottom: 0in">
<FONT FACE="courier, monospace">setExitHandler </FONT>lets us specify
a proc body that actually contains the user-defined exit handling.
E.g.
</P>
<P STYLE="margin-bottom: 0in"><BR>
</P>
<PRE> Object setExitHandler {
aObj destroy
puts "existing"
}
</PRE><P STYLE="margin-bottom: 0in">
destroys the object <FONT FACE="courier, monospace">aObj</FONT> before
all other objects and prints the message existing to the screen. With
<FONT FACE="courier, monospace">getExitHandler</FONT> the exit
handler can be introspected. E.g. if we just want to append the
destruction of object <FONT FACE="courier, monospace">bObj</FONT> to
an existing exit handler, we use <FONT FACE="courier, monospace">getExitHandler</FONT>:
</P>
<P STYLE="margin-bottom: 0in"><BR>
</P>
<PRE> Object setExitHandler "[Object getExitHandler]; bObj destroy"
</PRE><P STYLE="margin-bottom: 0in">
<FONT FACE="courier, monospace">unsetExitHandler</FONT> deletes the
exit handler.
</P>
<PRE STYLE="margin-top: 0.17in; margin-bottom: 0.2in; page-break-after: avoid">
</pre>
<H2><A NAME="autonames">Automatic Name Creation</A>
</H2>
The XOTcl <FONT SIZE=2>autoname</FONT>
instance method provides an simple way to take the task of
automatically creating names out of the responsibility of the
programmer. The example below shows how to create on each invocation
of method <FONT SIZE=2>new</FONT> an agent with a fresh name
(prefixed with <FONT SIZE=2>agent</FONT>):
</P>
<PRE>
Agent proc new args {
eval my [my autoname agent] $args
}
</PRE>
<p>
Autonames may have format strings as in the Tcl 'format' command.
E.g.:
</P>
<pre>my autoname a%06d</pre>
<p>
produces a000000, a000001, a000002, ...
</P>
<!-- PAGE BREAK -->
<TABLE COLS=2 WIDTH=100% BORDER=0 CELLPADDING=2 CELLSPACING=0 BGCOLOR="#000055">
<TR>
<TD WIDTH=75%>
<P><A NAME="cext"></A><FONT COLOR="#ffffff"><FONT FACE="Arial, Helvetica"><FONT SIZE=6>Integrating XOTcl Programs with C Extensions (such as TK)
</FONT></FONT></FONT>
</P>
</TD>
<TD>
<IMG SRC="logo-100.jpg" NAME="Graphic2" ALIGN=RIGHT WIDTH=102 HEIGHT=42 BORDER=0></TD>
</TR>
</TABLE>
<p>Because all XOTcl commands are in the ::xotcl namespace, it is
usually no problem to integrate XOTcl with other Tcl extensions. Most
often it works to import the XOTcl commands (like Object, Class) into
the current namespace because there are no name-clashes with the
commands defined by other extensions.</p>
<p>Consider you want to perform a deeper integration of an other
extension and XOTcl because you want to benefit from XOTcl's object
system. For instance, you might want to introduce composite TK widgets
(sometimes called mega-widgets) as classes and inherit from these
classes. Here, you have two options: you can change or extend the C
code of that other extension to provide XOTcl classes or objects, or
you can write an XOTcl wrapper in Tcl. For the first alternative,
there are some examples provided in the XOTcl distribution. XOTclGdbm
provides an OO Tcl interface to the GDBM database, for
instance. XOTclSdbm does the same for SDBM, and the TclExpat wrapper
provides a class-based interface to the TclExpat XML parser.</p>
<p>Consider you do not want to change the C code of a Tcl
extension. Then you can write an OO wrapper in XOTcl for the commands
of the other extension. For stateless commands, you can simply write
forwarder methods. If the extension maintains some state, you
typically associate the state handle with an XOTcl parameter, acquire
the state in the XOTcl constructor, and align the XOTcl destructor
with the stateful instance.</p>
<p>Consider you want to wrap the Tk button widget. You can acquire the
widget in the constructor, and maintain the widget ID in a
parameter. You now can forward invocations to this widget ID
(e.g. when using "pack"), or register command callbacks (like
buttonPressed). Note that we let the "self" command be replaced in the
scope of the current method so that TK receives the correct object ID
for the callback. In the destructor we destroy the widget as well (we
use "catch" because sometimes widgets can destroyed by other means as
well (e.g. by their parent widget, when a widget/object hierarchy is
destroyed at once).</p>
<pre>
Class MyButton -parameter {button}
MyButton instproc buttonPressed args {
puts "pressed [my set button]"
}
MyButton instproc init args {
set ID [namespace tail [self]]
my instvar button
set button [button .$ID \
-text "My Button $ID" \
-command [list [self] buttonPressed]]
pack $button
next
}
MyButton instproc destroy args {
catch {destroy [my set button]}
next
}
# a test -> 3 buttons, destroy one of them
foreach b {a b c} {
MyButton $b
}
b destroy
</pre>
<p> The "trick" to substitute "self" within the current method scope
works for all kinds of command callbacks. Extensions such as TK,
however, often work with bindings to (global) variables as well. Using
global variables is frowned upon in the OO community. Instead you
should use instance variables of objects. As Tcl can only bind to
existing namespace variables (and XOTcl acquires the namespace of an
object on demand), you have to make sure that the namespace of an
object exists before binding a variable. That can be done with
"requireNamespace":</p>
<pre>
GUIClass instproc buildEntry win {
my requireNamespace
entry $win -textvariable [self]::entryValue
my set entryValue {Init Value}
}
</pre>
<p>Note that in the above example we have used to tail of the object ID
as ID for the widget. Usually, it is a good idea to the object name,
if possible, for TK (and other extensions) IDs as well. Another option
is to use a autoname to get a unique name for the ID.</p>
<p>Sometimes you want to simply send all invocations, not implemented by
XOTcl, to the wrapped command. Here, it is tedious to write a wrapper
for each of these methods. Instead you can use "unknown" to handle
automatic forwarding. Consider you want to wrap TK commands like pack
and replace XOTcl object names with their TK widget ID, so that you can
use both IDs synonymously. You can rename the respective TK commands in
the following way:
<pre>
foreach tkCommand [list \
bell bind bindtags clipboard event focus font grid \
image lower option pack place raise selection send \
tk tkwait winfo wm] {
rename ::$tkCommand __tk_$tkCommand
TkCommand ::$tkCommand
::$tkCommand set wrapped __tk_$tkCommand
</pre>
<p>The XOTcl class handling the ID substitution for the TK command
might look as follows:</p>
<pre>
Class TkCommand -parameter wrapped
TkCommand instproc unknown args {
[self] instvar wrapped
set args [Widget replaceWithWidgetIDs $args]
# now call the command
eval $wrapped $args
}
</pre>
<p></p>
<!-- PAGE BREAK -->
<TABLE COLS=2 WIDTH=100% BORDER=0 CELLPADDING=2 CELLSPACING=0 BGCOLOR="#000055">
<TR>
<TD WIDTH=75%>
<P><A NAME="references"></A><FONT COLOR="#ffffff"><FONT FACE="Arial, Helvetica"><FONT SIZE=6>References
</FONT></FONT></FONT>
</P>
</TD>
<TD>
<IMG SRC="logo-100.jpg" NAME="Graphic2" ALIGN=RIGHT WIDTH=102 HEIGHT=42 BORDER=0></TD>
</TR>
</TABLE>
<P></P><A NAME="xotcl-filter"><STRONG>[Neumann and Zdun 1999a]</STRONG></A>
G. Neumann and U. Zdun.
Filters as a language support for design patterns in object-oriented
scripting languages.
In <EM>Proceedings of COOTS'99, 5th Conference on Object-Oriented
Technologies and Systems</EM>, San Diego, May 1999.
<P></P><A NAME="xotcl-objpattern"><STRONG>[Neumann and Zdun 1999b]</STRONG></A>
G. Neumann and U. Zdun.
Implementing object-specific design patterns using per-object mixins.
In <EM>Proc. of NOSA`99, Second Nordic Workshop on Software
Architecture</EM>, Ronneby, Sweden, August 1999.
<P></P><A NAME="xotcl-mixin"><STRONG>[Neumann and Zdun 1999c]</STRONG></A>
G. Neumann and U. Zdun.
Enhancing object-based system composition through per-object mixins.
In <EM>Proceedings of Asia-Pacific Software Engineering Conference
(APSEC)</EM>, Takamatsu, Japan, December 1999.
<P></P><A NAME="xotcl"><STRONG>[Neumann and Zdun 2000a]</STRONG></A>
G. Neumann and U. Zdun.
XOT<SMALL>CL</SMALL>, an object-oriented scripting language.
In <EM>Proceedings of Tcl2k: The 7th USENIX Tcl/Tk Conference</EM>,
Austin, Texas, February 2000.
<P></P><A NAME="xotcl-aggregation"><STRONG>[Neumann and Zdun 2000b]</STRONG></A>
G. Neumann and U. Zdun. Towards the Usage of Dynamic Object
Aggregations as a Form of Composition
In: <EM>Proceedings of Symposium of Applied Computing (SAC'00)</EM>, Como, Italy, Mar 19-21, 2000.
<P></P><A NAME="tcl"><STRONG>[Ousterhout 1990]</STRONG></A>
J. K. Ousterhout.
Tcl: An embeddable command language.
In <EM>Proc. of the 1990 Winter USENIX Conference</EM>, January 1990.
<P></P><A NAME="ousterhout"><STRONG>[Ousterhout 1998]</STRONG></A>
J. K. Ousterhout.
Scripting: Higher Level Programming for the 21st Century, IEEE Computer 31(3), March 1998.
<P></P><A NAME="otcl"><STRONG>[Wetherall and Lindblad 1995]</STRONG></A>
D. Wetherall and C. J. Lindblad. Extending Tcl for Dynamic
Object-Oriented Programming. Proc. of the Tcl/Tk Workshop '95, July 1995.
</BODY>
</HTML>
<!-- LocalWords: mixins mixin instproc instmixins superclasses XOTcl Zdun Tcl
-->
<!-- LocalWords: Wetherall Lindblad OTcl glueing namespaces Beckenbauer procs
-->
<!-- LocalWords: Bayern instprocs CLOS args incr namespace instfilter RDF
-->
<!-- LocalWords: instmixin calledclass calledproc callingclass callingproc
-->
<!-- LocalWords: callingobject filterreg instcommands isclass isobject eval
-->
<!-- LocalWords: ismetaclass destructors autoname Ousterhout
-->