documents/userman/umsintro.tex

% BEGIN LICENSE BLOCK
% Version: CMPL 1.1
%
% The contents of this file are subject to the Cisco-style Mozilla Public
% License Version 1.1 (the "License"); you may not use this file except
% in compliance with the License.  You may obtain a copy of the License
% at www.eclipse-clp.org/license.
%
% Software distributed under the License is distributed on an "AS IS"
% basis, WITHOUT WARRANTY OF ANY KIND, either express or implied.  See
% the License for the specific language governing rights and limitations
% under the License.
%
% The Original Code is  The ECLiPSe Constraint Logic Programming System.
% The Initial Developer of the Original Code is  Cisco Systems, Inc.
% Portions created by the Initial Developer are
% Copyright (C) 1995 - 2006 Cisco Systems, Inc.  All Rights Reserved.
%
% Contributor(s):
%
% END LICENSE BLOCK
%
% @(#)umsintro.tex	1.11 95/03/17
% $Id: umsintro.tex,v 1.6 2015/01/14 01:31:10 jschimpf Exp $
%
% REL	DATE	AUTHOR		DESCRIPTION
% 2.10	090589	David Miller	Convert to Latex and update for 2.10
%	120489	Micha Meier	Rewritten, added list of {\eclipse} features
%	160590	Joachim Schimpf	Updated for 3.0
%	160590	Joachim Schimpf	Updated for 3.2
%	see SCCS comments

\chapter{Introduction}
\label{chapintro}
%HEVEA\cutdef[1]{section}

\section{What is {\eclipse} ?}
{\eclipse}{} ({\eclipse} Constraint Logic Programming System)
is an open-source software system whose aim is to serve as a platform
for integrating various Logic Programming extensions.
It is used in particular for the cost-effective
development and deployment of constraint programming applications,
e.g. in the areas of planning, scheduling, resource allocation,
timetabling, transport etc.  It is also ideal for teaching most
aspects of combinatorial problem solving, e.g. problem modelling,
constraint programming, mathematical programming, and search
techniques.  It contains several constraint solver libraries, a
high-level modelling and control language, interfaces to third-party
solvers, an integrated development environment and interfaces for
embedding into host environments.

The {\eclipse} programming language has been developed from the Prolog
language (\cite{clocksin81}), more specifically the Edinburgh family
of Prologs and more recently the ISO Prolog standard (\cite{isoprolog95}).
{\eclipse} retains backward compatibility by supporting several language
dialects.

In terms of implementation technology, {\eclipse} is built around an
incremental compiler which compiles {\eclipse} source into WAM-like
code \cite{warren83}, and a runtime system comprising an emulator of
this abstract code, automatic memory management, I/O system and
built-in predicates.


\section{Overview}
\index{ECLiPSe@\eclipse}
\index{SEPIA@\sepia}
\index{MegaLog}
\index{CHIP}
The {\eclipse} logic programming system was originally an integration of
ECRC's \sepia, MegaLog and (parts of the) CHIP systems.
It was then further developed into a Constraint Logic Programming system
with a focus on hybrid problem solving and solver integration.
The documentation is organised as follows:
\begin{description}
\item [The User Manual] describes the functionality
of the {\eclipse} kernel (this document).
\item [The Constraint Library Manual] describes the major {\eclipse} libraries,
in particular the ones implementing constraint solvers.
\item [The Interfacing and Embedding Manual] describes how to interface
{\eclipse} to other programming languages, and in particular how to embed
it into an application as a component.
\item [The Reference Manual] contains detailed descriptions of all the
Built-in predicates and the libraries. This information is also available
from the development system's help/1 command and the tkeclipse library
browser.
\item [The Visualisation Manual] describes the facilities for the
visualisation of constraint propagation and search.
\end{description}
All the documentation can be accessed using an html browser
(refer to the eclipse installation directory under doc/index.html).

%\section{{\eclipse} Features}
%The following sections outline some remarkable features of {\eclipse}:
%
%\subsection{Metaterms}
%{\eclipse} provides the user with the {\bf metaterm} data type which is the key
%\index{metaterm}
%to many extensions to the basic Prolog language.
%It can be seen as a generic data type or as an attributed variable.
%The system calls user-definable event handlers
%when it encounters metaterms in certain contexts, e.g., unification.
%
%\subsection{Incremental Compiler}
%{\eclipse} is based on an {\it incremental interactive compiler},
%it contains no interpreter, all procedures are compiled.
%The compilation is transparent to the user, no loading or linking
%has to be done.
%{\eclipse} programs are both fast (like compiled programs) and flexible (like
%interpreted ones).
%The compiler can compile faster than most other Prologs consult their sources.
%
%\subsection{Source Variable Names}
%\index{variable names}
%{\eclipse} is able to remember the source names
%of variables so that e.g., debugging programs becomes much easier.
%By default, the toplevel loop and the debugger print variables with
%their source names.
%The compiler by default checks if the compiled clauses contain
%singleton variables and emits a warning if it is the case.
%
%\subsection{Flexibility}
%{\eclipse} enables the user to modify most of the system features,
%be it to customise it, build separate applications or include
%new features.
%Most of the built-in predicates can be modified by the user,
%several user-definable events allow the user to
%modify the Prolog top-level loop.
%
%\subsection{User Interface}
%The interface is user-friendly, e.g., in the top-level and in the debugger
%most functionalities can be accessed with a single key stroke.
%The top-level loop does not ask for displaying more alternatives
%when it can determine that there are none.
%All user interface errors are signaled to the user instead of just failing
%so that it is easy to spot trivial errors like misspelling
%a predicate or variable name.
%\index{name completion}
%Automatic name completion is provided:
%when the user types the first characters of
%an atom or functor followed by \verb.^.D, the system will try to
%complete the input from the known atoms and functors in the dictionary.
%For example {\tt gar}\verb.^.D will be expanded into  {\tt garbage_collect}.
%
%\subsection{Full Word Size}
%Unlike other Prolog systems, {\eclipse} does not use several bits
%of a 'Prolog word' for a tag and so its data types are fully compatible
%with external software and it can represent numbers, pointers etc.
%without any restriction.
%
%\subsection{Few Limits}
%All {\eclipse} memory areas are automatically extended when necessary.
%There are no limits (other than the available memory) to
%the size of atoms or strings or their number,
%the length of integers,
%the arity of structures is unlimited
%and there are no limits on the complexity of compiled clauses.
%
%\subsection{Occur Check and Complete Search Rule}
%\index{occur check}
%\index{depth-first iterative deepening}
%\index{Horn clause logic}
%Most Prolog systems implement logic programming only incompletely,
%because they omit the occur check and use only the depth-first
%search. {\eclipse} offers a complete system for Horn clause logic
%by providing optional compiler support for
%the occur check and for the depth-first iterative deepening search rule.
%
%\subsection{Strings}
%\index{strings}
%{\eclipse} has the data type {\it string} whose representation
%is compact and compatible with strings in C, but it also
%supports lists of ASCII codes.
%
%\subsection{Arrays and Global Variables}
%\index{arrays}
%\index{global variables}
%For special purposes and for communication with external
%software it is possible to use (extra-logical) arrays of various types
%and global variables which can be modified with destructive
%assignment.
%For instance, the implementation of a counter using a global variable
%is much faster and uses less space than using \bipref{assert/1}{../bips/kernel/dynamic/assert-1.html}.
%The arrays map on C structures and they can be manipulated
%in external predicates.
%The global array and variable names are module sensitive.
%
%\subsection{Indexed Database}
%\index{records}
%{\eclipse} contains the {\bf record} family of predicates which allow
%to store several terms under a specified key.
%There a two sets of predicates for the indexed database.
%One provides {\it database references} as used in Cprolog, the other
%works with keys only.
%The indexed database keys are module sensitive.
%
%\subsection{Modules}
%{\eclipse} has a sophisticated module concept which makes it possible to
%build large applications, avoid name clashes and to hide information
%from unauthorised access.
%The items covered by the module system's visibility rules are predicates,
%global variables and arrays, indexed database keys, operators and
%macro transformations.
%
%\subsection{Event Handling}
%{\eclipse} can handle synchronous events ({\bf exceptions}) and
%asynchronous ones ({\bf interrupts}).
%An exception can be a wrong argument type, calling undefined predicate,
%reading past the file end, etc.
%When an event occurs, the corresponding {\it event handler} is called,
%which is a user-definable procedure.
%All built-in predicates raise an event when a nonstandard situation
%occurs, be it an error or a situation where several actions
%would be possible and the user has the possibility to influence
%it by defining an appropriate event handler.
%
%When an interrupt occurs, e.g., a signal, the current execution
%is interrupted and the signal is handled immediately in real time.
%This feature can be used to write real-time applications,
%for example graphics applications which make use of a mouse etc.
%
%\subsection{Coroutining}
%The user can express a condition under which a call to a specified
%procedure will be delayed.
%This condition can be expressed in a normal Prolog form using
%unification and predicate calls.
%The delayed predicates are resumed when one of the specified
%variables is bound and the delay condition is then verified again.
%This goal suspension can be used to preserve completeness
%e.g., when avoiding infinite loops, to increase efficiency
%by reordering subgoals, to have sound negation
%and to implement coroutines.
%
%The user can also process suspensions explicitly which makes it
%possible to implement almost any control strategy.
%
%\subsection{Debugger}
%{\eclipse} includes a built-in debugger which debugs
%compiled code without {\it metainterpreting} it
%(i.e., the debugger is not a Prolog program which would
%simulate the execution of another program) and so the
%debugged code is not considerably slower than the not debugged one
%as is the case in most other Prolog systems.
%It also allows tracing the special {\eclipse} features.
%It is able to trace suspended goals, skip to the place
%where a suspended goal has been woken, trace event handlers,
%etc. and it has a comfortable user interface which can be easily
%connected with a windowing environment.
%The user can define macros of debugger commands, repeat them and
%set various leash modes of ports and procedures.
%Debugger events can be used to define conditional spy points,
%manipulate the goal variables etc.
%
%\subsection{Stream I/O}
%The {\eclipse} I/O is based on the concept of {\it streams}
%which are mapped on the I/O channels of the underlying operating system.
%Apart from user-definable streams there are several predefined
%system streams like {\it toplevel_output}, {\it debug_input} etc.
%which make it easy to filter the data of a certain kind,
%connect it with a windowing environment or to use it in
%event handlers.
%Any input stream can be passed as input to the compiler.
%Furthermore {\eclipse} supports I/O in raw mode and I/O to and from a string.
%
%\subsection{Macros}
%{\eclipse} provides a very general macro mechanism that can be used to transform
%terms and program clauses while they are read into the system
%and before they are printed out.
%
%\subsection{Blocks}
%Blocks as suggested by the ISO draft \cite{ISO} are supported
%by {\eclipse}.
%It is a mechanism similar to LISP's catch\&throw, it allows
%to abandon the execution of a goal, return to a specified ancestor
%and call another procedure instead.
%
%%\subsection{Sound and Constructive Negation}
%%In addition to the usual {\it negation by failure},
%%{\eclipse} contains the {\it sound negation}, similar to that provided by MU-Prolog
%%and also {\it constructive negation}.
%%Constructive negation can be used to generate solutions
%%for negated goals and thus it is logically complete and sound.
%%It uses delayed goals to express constraints on
%%the variables in a negated query, so that e.g.,
%%when a fact {\it p(a)} exists, the goal
%%{\it ?-neg p(X)} will yield a solution {\it X $\neq$ a}
%%\footnote{
%%For impatient users: before trying out this example
%%consult page \pageref{cnpage} how to load the appropriate library.}.
%
%\subsection{External Predicates}
%{\eclipse} can interface
%to any external function written in C and load it dynamically
%if required\footnote{The dynamic loading feature may not be available
%on certain machines},
%the C function can manipulate Prolog data, or it might
%be completely independent of Prolog structures.
%All C data can be mapped on Prolog data, in particular C structures
%correspond to {\eclipse} arrays.
%As in other Prolog systems,
%external procedures may succeed or fail,
%but {\eclipse} also offers the ability to backtrack, delay and
%call Prolog procedures from an external one.
%These features make the external predicates a powerful
%tool for the integration of non-Prolog data and control into {\eclipse}.
%
%\subsection{Compatibility}
%{\eclipse}'s flexible syntax and the extensive use of its event handling
%capabilities allows it to build compatibility packages for programs
%written in other Prolog dialects.
%Currently packages are provided for C-Prolog, Quintus Prolog
%SICStus Prolog and the old BSI standard draft.
%Many of the system parameters can be modified by changing
%internal flags.
%
%\subsection{Autoloading Libraries}
%\index{autoloading}
%{\eclipse} supports autoloading libraries of Prolog predicates.
%When a predicate is declared as autoloading, it is loaded
%from the corresponding file in the library path when it is
%called for the first time.
%User libraries are supported, the library path can contain
%any user directories.
%
%\subsection{Profiling Tools}
%\index{profiling}
%Two tools are provided to analyse program runs and extract
%information for performance improvements.
%One is a port-model based analyser which produces statistics
%about how many times a program passed every debugger port.
%The other is a true timing profiler which measures the time spent in
%every predicate of a program. The latter works even on optimized compiled code.
%
%\subsection{Portability}
%\sepia and {\eclipse} have been developed on a Sun-4 and
%ported to VAXes,
%Apollo, Bull DPX machines and Siemens MX-300 and MX-500.
%Apart from the
%software implementation of \sepia, the Computer Architecture Group at ECRC
%developed a hardware implementation termed the {\it Knowledge Crunching Machine}
%(KCM) \cite{benker89}.
%
%\subsection{On-line Documentation}
%The {\eclipse} Prolog Built-in Predicate Reference Manual
%(BIP book \cite{sepiabipbook}) is available on-line.
%Calling {\tt help(PredSpec)} will display the appropriate manual page,
%containing a detailed description of the specified built-in predicate.
%
%\section{Extensions}
%{\eclipse} has been designed as a basis for various extensions
%%of Prolog which may even go beyond the scope of logic programming.
%%\index{OPIUM}
%The various Constraints Logic Programming libraries
%are examples of successful systems based on {\eclipse}'s extensibility:
%\begin{description}
%\item [library(fd)] finite-domain constraint solver
%\item [library(r)] linear rational arithmetic constraint solver
%\item [library(chr)] constraint handling rules system
%\item [library(propia)] generalized propagation system
%\item [library(conjunto)] finite sets constraint solver
%\end{description}

\section{Further Information}
{\eclipse} was initially developed at the European Computer-Industry
Research Centre (ECRC) in Munich, and then at IC-Parc, Imperial College in London
until the end of 2005. It is now an open-source project, with the
support of Cisco Systems.
Up-to-date information can be obtained from the {\eclipse} web site
\begin{quote}
\begin{verbatim}
http://www.eclipseclp.org
\end{verbatim}
\end{quote}
or from the Sourceforge site under the project name {\bf eclipse-clp}
\begin{quote}
\begin{verbatim}
http://www.sourceforge.net/projects/eclipse-clp
\end{verbatim}
\end{quote}
which also hosts the main source repository.
There you can also subscribe to the {\eclipse} user group mailing list
or access its archives.%
\index{user group}%
\index{mailing list}
\begin{quote}
\begin{verbatim}
eclipse-clp-users@lists.sf.net
\end{verbatim}
\end{quote}

\section{Reporting Problems}
In order to make {\eclipse} as useful and reliable as possible,
we would like to encourage users to submit problem reports via the web site
\begin{quote}
\begin{verbatim}
http://eclipseclp.org/bugs.html
\end{verbatim}
\end{quote}
or by e-mail to
\begin{quote}
\begin{verbatim}
eclipse-clp-bugs@lists.sf.net
\end{verbatim}
\end{quote}
\index{bug reports}
%HEVEA\cutend