org.chocosolver.parser.flatzinc.ast.FGoal Maven / Gradle / Ivy
Go to download
Show more of this group Show more artifacts with this name
Show all versions of choco-parsers Show documentation
Show all versions of choco-parsers Show documentation
Provide parsers from FlatZinc, XCSP3 and MSP to Choco-solver.
/*
* This file is part of choco-parsers, http://choco-solver.org/
*
* Copyright (c) 2024, IMT Atlantique. All rights reserved.
*
* Licensed under the BSD 4-clause license.
*
* See LICENSE file in the project root for full license information.
*/
package org.chocosolver.parser.flatzinc.ast;
import org.chocosolver.parser.ParserException;
import org.chocosolver.parser.flatzinc.ast.expression.EAnnotation;
import org.chocosolver.parser.flatzinc.ast.expression.EArray;
import org.chocosolver.parser.flatzinc.ast.expression.EIdentifier;
import org.chocosolver.parser.flatzinc.ast.expression.Expression;
import org.chocosolver.parser.flatzinc.ast.searches.Assignment;
import org.chocosolver.parser.flatzinc.ast.searches.IntSearch;
import org.chocosolver.parser.flatzinc.ast.searches.SetSearch;
import org.chocosolver.parser.flatzinc.ast.searches.VarChoice;
import org.chocosolver.solver.Model;
import org.chocosolver.solver.ResolutionPolicy;
import org.chocosolver.solver.search.limits.FailCounter;
import org.chocosolver.solver.search.strategy.strategy.AbstractStrategy;
import org.chocosolver.solver.search.strategy.strategy.StrategiesSequencer;
import org.chocosolver.solver.variables.IntVar;
import org.chocosolver.solver.variables.SetVar;
import java.util.Arrays;
import java.util.List;
import java.util.Objects;
/*
* User : CPRUDHOM
* Mail : cprudhom(a)emn.fr
* Date : 12 janv. 2010
* Since : Choco 2.1.1
*
* Class for solve goals definition based on flatzinc-like objects.
*
* A solve goal is defined with:
* 'solve annotations satisfy;'
* or 'solve annotations maximize expression;'
* /br> or 'solve annotations minimize expression;'
*/
@SuppressWarnings({"rawtypes", "unchecked"})
public class FGoal {
private enum Search {
seq_search,
int_search,
bool_search,
set_search,
warm_start_bool,
warm_start_int,
restart_luby,
restart_geometric,
restart_linear,
restart_constant,
restart_none
}
public static void define_goal(Model aModel, List annotations, ResolutionPolicy type, Expression expr) {
// First define solving process
if (type != ResolutionPolicy.SATISFACTION) {
IntVar obj = expr.intVarValue(aModel);
aModel.setObjective(type == ResolutionPolicy.MAXIMIZE, obj);
}
// Then define search goal
StringBuilder description = new StringBuilder();
// Always read the search strategies, if any
AbstractStrategy[] strategies = new AbstractStrategy[annotations.size()];
for (int a = 0; a < annotations.size(); a++) {
EAnnotation annotation = annotations.get(a);
if (annotation.id.value.equals("seq_search")) {
EArray earray = (EArray) annotation.exps.get(0);
AbstractStrategy[] substrategies = new AbstractStrategy[earray.what.size()];
for (int i = 0; i < substrategies.length; i++) {
substrategies[i] = readSearchAnnotation((EAnnotation) earray.getWhat_i(i), aModel, description);
}
strategies[a] = new StrategiesSequencer(aModel.getEnvironment(),
Arrays.stream(substrategies)
.filter(Objects::nonNull)
.toArray(AbstractStrategy[]::new));
} else {
strategies[a] = readSearchAnnotation(annotation, aModel, description);
}
}
strategies = Arrays.stream(strategies).filter(Objects::nonNull).toArray(AbstractStrategy[]::new);
if(strategies.length > 0){
aModel.getSolver().setSearch(strategies);
}
}
/**
* Read search annotation and build corresponding strategy
*
* @param e {@link org.chocosolver.parser.flatzinc.ast.expression.EAnnotation}
* @param model solver within the search is defined
* @return {@code true} if a search strategy is defined
*/
private static AbstractStrategy readSearchAnnotation(EAnnotation e, Model model, StringBuilder description) {
Expression[] exps = new Expression[e.exps.size()];
e.exps.toArray(exps);
Search search;
try {
search = Search.valueOf(e.id.value);
} catch (IllegalArgumentException ex) {
model.getSolver().log().printf("%% ignored search annotation: %s\n", e);
return null;
}
if (search == Search.seq_search) {
EArray eArray = (EArray) e.exps.get(0);
AbstractStrategy[] strats = new AbstractStrategy[eArray.what.size()];
for (int i = 0; i < strats.length; i++) {
strats[i] = readSearchAnnotation((EAnnotation) eArray.getWhat_i(i), model, description);
}
return org.chocosolver.solver.search.strategy.Search.sequencer(strats);
}
if (search == Search.warm_start_int || search == Search.warm_start_bool) {
IntVar[] scope = exps[0].toIntVarArray(model); // deal with set var?
int[] values = exps[1].toIntArray();
for (int i = 0; i < scope.length; i++) {
model.getSolver().addHint(scope[i], values[i]);
}
return null;
}
if (search.toString().startsWith("restart")) {
switch (search) {
case restart_luby:
int scale = exps[0].intValue();
model.getSolver().setLubyRestart(scale, new FailCounter(model, 0), Integer.MAX_VALUE);
break;
case restart_geometric:
float base = exps[0].floatValue();
int scale0 = exps[1].intValue();
model.getSolver().setGeometricalRestart(scale0, base, new FailCounter(model, 0), Integer.MAX_VALUE);
break;
case restart_linear:
int scale1 = exps[0].intValue();
model.getSolver().setLinearRestart(scale1, new FailCounter(model, 0), Integer.MAX_VALUE);
break;
case restart_constant:
int scale2 = exps[0].intValue();
model.getSolver().setConstantRestart(scale2, new FailCounter(model, 0), Integer.MAX_VALUE);
break;
case restart_none:
break;
}
return null;
}
VarChoice vchoice = VarChoice.valueOf(((EIdentifier) exps[1]).value);
description.append(vchoice).append(";");
Assignment assignment = Assignment.valueOf(((EIdentifier) exps[2]).value);
switch (search) {
case int_search:
case bool_search: {
IntVar[] scope = exps[0].toIntVarArray(model);
return IntSearch.build(scope, vchoice, assignment, model);
}
case set_search: {
SetVar[] scope = exps[0].toSetVarArray(model);
return SetSearch.build(scope, vchoice, assignment, model);
}
default:
System.err.println("Unknown search annotation " + e);
throw new ParserException();
}
}
}