[dna-commits] DNA SVN: r1319 - in trunk: dna-graph/src/main/java/org/jboss/dna/graph/query/process and 6 other directories.

Author: rhauch Date: 2009-11-16 18:24:06 -0500 (Mon, 16 Nov 2009) New Revision: 1319 Added: trunk/dna-graph/src/main/java/org/jboss/dna/graph/query/process/AbstractAccessComponent.java trunk/dna-search/src/main/java/org/jboss/dna/search/DualIndexStrategy.java trunk/dna-search/src/main/java/org/jboss/dna/search/IndexRules.java trunk/dna-search/src/main/java/org/jboss/dna/search/IndexStrategy.java trunk/dna-search/src/main/java/org/jboss/dna/search/KitchenSinkIndexStrategy.java trunk/dna-search/src/main/java/org/jboss/dna/search/SearchContext.java Removed: trunk/dna-graph/src/main/java/org/jboss/dna/graph/query/process/ValueCache.java trunk/dna-search/src/main/java/org/jboss/dna/search/IndexingRules.java trunk/dna-search/src/main/java/org/jboss/dna/search/IndexingStrategy.java trunk/dna-search/src/main/java/org/jboss/dna/search/LuceneQueryComponent.java trunk/dna-search/src/main/java/org/jboss/dna/search/LuceneQueryEngine.java trunk/dna-search/src/main/java/org/jboss/dna/search/StoreLittleIndexingStrategy.java Modified: trunk/dna-graph/src/main/java/org/jboss/dna/graph/query/QueryContext.java trunk/dna-graph/src/main/java/org/jboss/dna/graph/query/QueryEngine.java trunk/dna-graph/src/main/java/org/jboss/dna/graph/query/Queryable.java trunk/dna-graph/src/main/java/org/jboss/dna/graph/query/validate/ImmutableSchemata.java trunk/dna-graph/src/test/java/org/jboss/dna/graph/query/optimize/AddAccessNodesTest.java trunk/dna-graph/src/test/java/org/jboss/dna/graph/query/optimize/ChooseJoinAlgorithmTest.java trunk/dna-graph/src/test/java/org/jboss/dna/graph/query/optimize/PushSelectCriteriaTest.java trunk/dna-graph/src/test/java/org/jboss/dna/graph/query/optimize/ReplaceViewsTest.java trunk/dna-graph/src/test/java/org/jboss/dna/graph/query/optimize/RightOuterToLeftOuterJoinsTest.java trunk/dna-graph/src/test/java/org/jboss/dna/graph/query/optimize/RuleBasedOptimizerTest.java trunk/dna-graph/src/test/java/org/jboss/dna/graph/query/plan/CanonicalPlannerTest.java trunk/dna-graph/src/test/java/org/jboss/dna/graph/query/process/DistinctComponentTest.java trunk/dna-graph/src/test/java/org/jboss/dna/graph/query/process/QueryResultColumnsTest.java trunk/dna-graph/src/test/java/org/jboss/dna/graph/query/process/SortLocationsComponentTest.java trunk/dna-graph/src/test/java/org/jboss/dna/graph/query/process/SortValuesComponentTest.java trunk/dna-search/src/main/java/org/jboss/dna/search/IndexContext.java trunk/dna-search/src/main/java/org/jboss/dna/search/SearchEngine.java trunk/dna-search/src/main/java/org/jboss/dna/search/WorkspaceSearchEngine.java trunk/dna-search/src/test/java/org/jboss/dna/search/IndexingRulesTest.java trunk/dna-search/src/test/java/org/jboss/dna/search/SearchEngineTest.java trunk/dna-search/src/test/java/org/jboss/dna/search/WorkspaceSearchEngineTest.java Log: DNA-467 Additional refactoring to move toward a working search engine. Modified: trunk/dna-graph/src/main/java/org/jboss/dna/graph/query/QueryContext.java =================================================================== --- trunk/dna-graph/src/main/java/org/jboss/dna/graph/query/QueryContext.java 2009-11-16 23:23:02 UTC (rev 1318) +++ trunk/dna-graph/src/main/java/org/jboss/dna/graph/query/QueryContext.java 2009-11-16 23:24:06 UTC (rev 1319) @@ -26,6 +26,7 @@ import java.util.Collections; import java.util.HashMap; import java.util.Map; +import net.jcip.annotations.Immutable; import org.jboss.dna.common.collection.Problems; import org.jboss.dna.common.collection.SimpleProblems; import org.jboss.dna.common.util.CheckArg; @@ -35,64 +36,120 @@ import org.jboss.dna.graph.query.validate.Schemata; /** - * + * An immutable context in which queries are to be executed. Each query context defines the information that is available during + * query execution. */ -public final class QueryContext { +@Immutable +public class QueryContext { private final ExecutionContext context; private final PlanHints hints; private final Schemata schemata; private final Problems problems; private final Map<String, Object> variables; + /** + * Create a new context for query execution. + * + * @param context the execution context + * @param schemata the schemata + * @param hints the hints, or null if there are no hints + * @param problems the problems container, or null if a new problems container should be created + * @param variables the mapping of variables and values, or null if there are no such variables + * @throws IllegalArgumentException if the context or schmata are null + */ public QueryContext( ExecutionContext context, + Schemata schemata, PlanHints hints, - Schemata schemata, Problems problems, Map<String, Object> variables ) { CheckArg.isNotNull(context, "context"); + CheckArg.isNotNull(schemata, "schemata"); this.context = context; this.hints = hints != null ? hints : new PlanHints(); this.schemata = schemata; this.problems = problems != null ? problems : new SimpleProblems(); this.variables = variables != null ? Collections.<String, Object>unmodifiableMap(new HashMap<String, Object>(variables)) : Collections.<String, Object>emptyMap(); + assert this.context != null; + assert this.hints != null; + assert this.schemata != null; + assert this.problems != null; + assert this.variables != null; } + /** + * Create a new context for query execution. + * + * @param context the execution context + * @param schemata the schemata + * @param hints the hints, or null if there are no hints + * @param problems the problems container, or null if a new problems container should be created + * @throws IllegalArgumentException if the context or schmata are null + */ public QueryContext( ExecutionContext context, + Schemata schemata, PlanHints hints, - Schemata schemata, Problems problems ) { - this(context, hints, schemata, problems, null); + this(context, schemata, hints, problems, null); } + /** + * Create a new context for query execution. + * + * @param context the execution context + * @param schemata the schemata + * @param hints the hints, or null if there are no hints + * @throws IllegalArgumentException if the context or schmata are null + */ public QueryContext( ExecutionContext context, - PlanHints hints, + Schemata schemata, + PlanHints hints ) { + this(context, schemata, hints, null, null); + } + + /** + * Create a new context for query execution. + * + * @param context the execution context + * @param schemata the schemata + * @throws IllegalArgumentException if the context or schmata are null + */ + public QueryContext( ExecutionContext context, Schemata schemata ) { - this(context, hints, schemata, null, null); + this(context, schemata, null, null, null); } /** - * @return context + * Get the execution context available to this query context. + * + * @return the execution context; never null */ public final ExecutionContext getExecutionContext() { return context; } /** - * @return hints + * Get the plan hints. + * + * @return the plan hints; never null */ public final PlanHints getHints() { return hints; } /** - * @return problems + * Get the problem container used by this query context. Any problems that have been encountered will be accumlated in this + * container. + * + * @return the problem container; never null */ public final Problems getProblems() { return problems; } /** - * @return schemata + * Get the definition of the tables available within this query context. + * + * @return the schemata; never null */ public Schemata getSchemata() { return schemata; @@ -101,10 +158,66 @@ /** * Get the variables that are to be substituted into the {@link BindVariableName} used in the query. * - * @return immutable map of variable values keyed by their name; never null + * @return immutable map of variable values keyed by their name; never null but possibly empty */ public Map<String, Object> getVariables() { return variables; } + /** + * Obtain a copy of this context, except that the copy uses the supplied execution context. + * + * @param context the execution context that should be used in the new query context + * @return the new context; never null + * @throws IllegalArgumentException if the execution context reference is null + */ + public QueryContext with( ExecutionContext context ) { + CheckArg.isNotNull(context, "context"); + return new QueryContext(context, schemata, hints, problems, variables); + } + + /** + * Obtain a copy of this context, except that the copy uses the supplied schemata. + * + * @param schemata the schemata that should be used in the new context + * @return the new context; never null + * @throws IllegalArgumentException if the schemata reference is null + */ + public QueryContext with( Schemata schemata ) { + CheckArg.isNotNull(schemata, "schemata"); + return new QueryContext(context, schemata, hints, problems, variables); + } + + /** + * Obtain a copy of this context, except that the copy uses the supplied hints. + * + * @param hints the hints that should be used in the new context + * @return the new context; never null + * @throws IllegalArgumentException if the hints reference is null + */ + public QueryContext with( PlanHints hints ) { + CheckArg.isNotNull(hints, "hints"); + return new QueryContext(context, schemata, hints, problems, variables); + } + + /** + * Obtain a copy of this context, except that the copy uses the supplied problem container. + * + * @param problems the problems that should be used in the new context; may be null if a new problem container should be used + * @return the new context; never null + */ + public QueryContext with( Problems problems ) { + return new QueryContext(context, schemata, hints, problems, variables); + } + + /** + * Obtain a copy of this context, except that the copy uses the supplied variables. + * + * @param variables the variables that should be used in the new context; may be null if there are no such variables + * @return the new context; never null + */ + public QueryContext with( Map<String, Object> variables ) { + return new QueryContext(context, schemata, hints, problems, variables); + } + } Modified: trunk/dna-graph/src/main/java/org/jboss/dna/graph/query/QueryEngine.java =================================================================== --- trunk/dna-graph/src/main/java/org/jboss/dna/graph/query/QueryEngine.java 2009-11-16 23:23:02 UTC (rev 1318) +++ trunk/dna-graph/src/main/java/org/jboss/dna/graph/query/QueryEngine.java 2009-11-16 23:24:06 UTC (rev 1319) @@ -27,18 +27,15 @@ import java.util.concurrent.atomic.AtomicBoolean; import net.jcip.annotations.ThreadSafe; import org.jboss.dna.common.util.CheckArg; -import org.jboss.dna.graph.ExecutionContext; import org.jboss.dna.graph.query.QueryResults.Statistics; import org.jboss.dna.graph.query.model.Column; import org.jboss.dna.graph.query.model.Constraint; import org.jboss.dna.graph.query.model.FullTextSearch; import org.jboss.dna.graph.query.model.QueryCommand; -import org.jboss.dna.graph.query.model.SelectorName; import org.jboss.dna.graph.query.model.Visitors; import org.jboss.dna.graph.query.optimize.Optimizer; import org.jboss.dna.graph.query.optimize.RuleBasedOptimizer; import org.jboss.dna.graph.query.plan.CanonicalPlanner; -import org.jboss.dna.graph.query.plan.PlanHints; import org.jboss.dna.graph.query.plan.PlanNode; import org.jboss.dna.graph.query.plan.Planner; import org.jboss.dna.graph.query.plan.PlanNode.Property; @@ -54,17 +51,6 @@ @ThreadSafe public class QueryEngine implements Queryable { - /** - * A {@link Schemata} implementation that always returns null, meaning the table does not exist. - */ - private static final Schemata DEFAULT_SCHEMATA = new Schemata() { - public Table getTable( SelectorName name ) { - // This won't allow the query engine to do anything (or much of anything), - // but it is legal and will result in meaningful problems - return null; - } - }; - private final Planner planner; private final Optimizer optimizer; private final Processor processor; @@ -92,42 +78,25 @@ /** * {@inheritDoc} * - * @see org.jboss.dna.graph.query.Queryable#execute(org.jboss.dna.graph.ExecutionContext, - * org.jboss.dna.graph.query.model.QueryCommand, org.jboss.dna.graph.query.validate.Schemata) + * @see org.jboss.dna.graph.query.Queryable#execute(org.jboss.dna.graph.query.QueryContext, + * org.jboss.dna.graph.query.model.QueryCommand) */ - public QueryResults execute( ExecutionContext context, - QueryCommand query, - Schemata schemata ) { - return execute(context, query, schemata, new PlanHints()); - } - - /** - * {@inheritDoc} - * - * @see org.jboss.dna.graph.query.Queryable#execute(org.jboss.dna.graph.ExecutionContext, - * org.jboss.dna.graph.query.model.QueryCommand, org.jboss.dna.graph.query.validate.Schemata, - * org.jboss.dna.graph.query.plan.PlanHints) - */ - public QueryResults execute( ExecutionContext context, - QueryCommand query, - Schemata schemata, - PlanHints hints ) { + public QueryResults execute( QueryContext context, + QueryCommand query ) { CheckArg.isNotNull(context, "context"); CheckArg.isNotNull(query, "query"); - if (schemata == null) schemata = DEFAULT_SCHEMATA; - QueryContext queryContext = new QueryContext(context, hints, schemata); // Create the canonical plan ... long start = System.nanoTime(); - PlanNode plan = planner.createPlan(queryContext, query); + PlanNode plan = planner.createPlan(context, query); long duration = System.nanoTime() - start; Statistics stats = new Statistics(duration); QueryResultColumns resultColumns = QueryResultColumns.empty(); - if (!queryContext.getProblems().hasErrors()) { + if (!context.getProblems().hasErrors()) { // Optimize the plan ... start = System.nanoTime(); - PlanNode optimizedPlan = optimizer.optimize(queryContext, plan); + PlanNode optimizedPlan = optimizer.optimize(context, plan); duration = System.nanoTime() - start; stats = stats.withOptimizationTime(duration); @@ -137,11 +106,11 @@ duration = System.nanoTime() - start; stats = stats.withOptimizationTime(duration); - if (!queryContext.getProblems().hasErrors()) { + if (!context.getProblems().hasErrors()) { // Execute the plan ... try { start = System.nanoTime(); - return processor.execute(queryContext, query, stats, optimizedPlan); + return processor.execute(context, query, stats, optimizedPlan); } finally { duration = System.nanoTime() - start; stats = stats.withOptimizationTime(duration); @@ -149,7 +118,7 @@ } } // There were problems somewhere ... - return new org.jboss.dna.graph.query.process.QueryResults(queryContext, query, resultColumns, stats); + return new org.jboss.dna.graph.query.process.QueryResults(context, query, resultColumns, stats); } protected QueryResultColumns determineQueryResultColumns( PlanNode optimizedPlan ) { Modified: trunk/dna-graph/src/main/java/org/jboss/dna/graph/query/Queryable.java =================================================================== --- trunk/dna-graph/src/main/java/org/jboss/dna/graph/query/Queryable.java 2009-11-16 23:23:02 UTC (rev 1318) +++ trunk/dna-graph/src/main/java/org/jboss/dna/graph/query/Queryable.java 2009-11-16 23:24:06 UTC (rev 1319) @@ -23,10 +23,7 @@ */ package org.jboss.dna.graph.query; -import org.jboss.dna.graph.ExecutionContext; import org.jboss.dna.graph.query.model.QueryCommand; -import org.jboss.dna.graph.query.plan.PlanHints; -import org.jboss.dna.graph.query.validate.Schemata; /** * An interface defining the ability to submit a query and obtain results. @@ -38,26 +35,9 @@ * * @param context the context in which the query should be executed * @param query the query that is to be executed - * @param schemata the schemata that should be used to validate the query * @return the query results; never null * @throws IllegalArgumentException if the context or query references are null */ - QueryResults execute( ExecutionContext context, - QueryCommand query, - Schemata schemata ); - - /** - * Execute the supplied query by planning, optimizing, and then processing it. - * - * @param context the context in which the query should be executed - * @param query the query that is to be executed - * @param schemata the schemata that should be used to validate the query - * @param hints the hints for the execution; may be null if there are no hints - * @return the query results; never null - * @throws IllegalArgumentException if the context or query references are null - */ - QueryResults execute( ExecutionContext context, - QueryCommand query, - Schemata schemata, - PlanHints hints ); + QueryResults execute( QueryContext context, + QueryCommand query ); } Added: trunk/dna-graph/src/main/java/org/jboss/dna/graph/query/process/AbstractAccessComponent.java =================================================================== --- trunk/dna-graph/src/main/java/org/jboss/dna/graph/query/process/AbstractAccessComponent.java (rev 0) +++ trunk/dna-graph/src/main/java/org/jboss/dna/graph/query/process/AbstractAccessComponent.java 2009-11-16 23:24:06 UTC (rev 1319) @@ -0,0 +1,120 @@ +/* + * JBoss DNA (http://www.jboss.org/dna) + * See the COPYRIGHT.txt file distributed with this work for information + * regarding copyright ownership. Some portions may be licensed + * to Red Hat, Inc. under one or more contributor license agreements. + * See the AUTHORS.txt file in the distribution for a full listing of + * individual contributors. + * + * JBoss DNA is free software. Unless otherwise indicated, all code in JBoss DNA + * is licensed to you under the terms of the GNU Lesser General Public License as + * published by the Free Software Foundation; either version 2.1 of + * the License, or (at your option) any later version. + * + * JBoss DNA is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public + * License along with this software; if not, write to the Free + * Software Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA + * 02110-1301 USA, or see the FSF site: http://www.fsf.org. + */ +package org.jboss.dna.graph.query.process; + +import java.util.ArrayList; +import java.util.List; +import org.jboss.dna.graph.Location; +import org.jboss.dna.graph.property.Name; +import org.jboss.dna.graph.property.NameFactory; +import org.jboss.dna.graph.query.QueryContext; +import org.jboss.dna.graph.query.QueryResults.Columns; +import org.jboss.dna.graph.query.model.AllNodes; +import org.jboss.dna.graph.query.model.And; +import org.jboss.dna.graph.query.model.Column; +import org.jboss.dna.graph.query.model.Constraint; +import org.jboss.dna.graph.query.model.Limit; +import org.jboss.dna.graph.query.model.SelectorName; +import org.jboss.dna.graph.query.plan.PlanNode; +import org.jboss.dna.graph.query.plan.PlanNode.Property; +import org.jboss.dna.graph.query.plan.PlanNode.Type; +import org.jboss.dna.graph.query.validate.Schemata; + +/** + * A reusable base class for {@link ProcessingComponent} implementations that does everything except obtain the correct + * {@link Location} objects for the query results. + */ +public abstract class AbstractAccessComponent extends ProcessingComponent { + + protected final PlanNode accessNode; + protected final SelectorName sourceName; + protected final List<Column> projectedColumns; + protected final Constraint constraint; + protected final Limit limit; + + protected AbstractAccessComponent( QueryContext context, + Columns columns, + PlanNode accessNode ) { + super(context, columns); + this.accessNode = accessNode; + + // Find the table name; should be + PlanNode source = accessNode.findAtOrBelow(Type.SOURCE); + if (source != null) { + this.sourceName = source.getProperty(Property.SOURCE_NAME, SelectorName.class); + if (!AllNodes.ALL_NODES_NAME.equals(this.sourceName)) { + throw new IllegalArgumentException(); + } + } else { + throw new IllegalArgumentException(); + } + + // Find the project ... + PlanNode project = accessNode.findAtOrBelow(Type.PROJECT); + if (project != null) { + List<Column> projectedColumns = project.getPropertyAsList(Property.PROJECT_COLUMNS, Column.class); + if (projectedColumns != null) { + this.projectedColumns = projectedColumns; + } else { + // Get the columns from the source columns ... + List<Schemata.Column> schemataColumns = source.getPropertyAsList(Property.SOURCE_COLUMNS, Schemata.Column.class); + this.projectedColumns = new ArrayList<Column>(schemataColumns.size()); + NameFactory nameFactory = context.getExecutionContext().getValueFactories().getNameFactory(); + for (Schemata.Column schemataColumn : schemataColumns) { + String columnName = schemataColumn.getName(); + // PropertyType type = schemataColumn.getPropertyType(); + Name propertyName = nameFactory.create(columnName); + Column column = new Column(sourceName, propertyName, columnName); + this.projectedColumns.add(column); + } + } + } else { + throw new IllegalArgumentException(); + } + + // Add the criteria ... + Constraint constraint = null; + for (PlanNode select : accessNode.findAllAtOrBelow(Type.SELECT)) { + Constraint selectConstraint = select.getProperty(Property.SELECT_CRITERIA, Constraint.class); + if (constraint != null) { + constraint = new And(constraint, selectConstraint); + } else { + constraint = selectConstraint; + } + } + this.constraint = constraint; + + // Find the limit ... + Limit limit = Limit.NONE; + PlanNode limitNode = accessNode.findAtOrBelow(Type.LIMIT); + if (limitNode != null) { + Integer count = limitNode.getProperty(Property.LIMIT_COUNT, Integer.class); + if (count != null) limit = limit.withRowLimit(count.intValue()); + Integer offset = limitNode.getProperty(Property.LIMIT_OFFSET, Integer.class); + if (offset != null) limit = limit.withOffset(offset.intValue()); + } + this.limit = limit; + } + +} Property changes on: trunk/dna-graph/src/main/java/org/jboss/dna/graph/query/process/AbstractAccessComponent.java ___________________________________________________________________ Name: svn:keywords + Id Revision Name: svn:eol-style + LF Deleted: trunk/dna-graph/src/main/java/org/jboss/dna/graph/query/process/ValueCache.java =================================================================== --- trunk/dna-graph/src/main/java/org/jboss/dna/graph/query/process/ValueCache.java 2009-11-16 23:23:02 UTC (rev 1318) +++ trunk/dna-graph/src/main/java/org/jboss/dna/graph/query/process/ValueCache.java 2009-11-16 23:24:06 UTC (rev 1319) @@ -1,44 +0,0 @@ -/* - * JBoss DNA (http://www.jboss.org/dna) - * See the COPYRIGHT.txt file distributed with this work for information - * regarding copyright ownership. Some portions may be licensed - * to Red Hat, Inc. under one or more contributor license agreements. - * See the AUTHORS.txt file in the distribution for a full listing of - * individual contributors. - * - * JBoss DNA is free software. Unless otherwise indicated, all code in JBoss DNA - * is licensed to you under the terms of the GNU Lesser General Public License as - * published by the Free Software Foundation; either version 2.1 of - * the License, or (at your option) any later version. - * - * JBoss DNA is distributed in the hope that it will be useful, - * but WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU - * Lesser General Public License for more details. - * - * You should have received a copy of the GNU Lesser General Public - * License along with this software; if not, write to the Free - * Software Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA - * 02110-1301 USA, or see the FSF site: http://www.fsf.org. - */ -package org.jboss.dna.graph.query.process; - -import org.jboss.dna.graph.Location; -import org.jboss.dna.graph.property.Name; -import org.jboss.dna.graph.query.QueryResults; - -/** - * Interface representing a cache of property values used in a {@link QueryResults}. - */ -public interface ValueCache { - - /** - * Get the value of the named property on the node at the supplied location. - * - * @param location the location of the node; may not be null - * @param name the property name - * @return the value of the property, or null if there is no such value - */ - Object getProperty( Location location, - Name name ); -} Modified: trunk/dna-graph/src/main/java/org/jboss/dna/graph/query/validate/ImmutableSchemata.java =================================================================== --- trunk/dna-graph/src/main/java/org/jboss/dna/graph/query/validate/ImmutableSchemata.java 2009-11-16 23:23:02 UTC (rev 1318) +++ trunk/dna-graph/src/main/java/org/jboss/dna/graph/query/validate/ImmutableSchemata.java 2009-11-16 23:24:06 UTC (rev 1319) @@ -45,7 +45,6 @@ import org.jboss.dna.graph.query.parse.InvalidQueryException; import org.jboss.dna.graph.query.parse.SqlQueryParser; import org.jboss.dna.graph.query.plan.CanonicalPlanner; -import org.jboss.dna.graph.query.plan.PlanHints; import org.jboss.dna.graph.query.plan.PlanNode; import org.jboss.dna.graph.query.plan.PlanNode.Property; import org.jboss.dna.graph.query.plan.PlanNode.Type; @@ -394,7 +393,7 @@ for (SelectorName name : viewNames) { QueryCommand command = definitions.get(name); // Create the canonical plan for the definition ... - QueryContext queryContext = new QueryContext(context, new PlanHints(), schemata); + QueryContext queryContext = new QueryContext(context, schemata); CanonicalPlanner planner = new CanonicalPlanner(); PlanNode plan = planner.createPlan(queryContext, command); if (queryContext.getProblems().hasErrors()) continue; Modified: trunk/dna-graph/src/test/java/org/jboss/dna/graph/query/optimize/AddAccessNodesTest.java =================================================================== --- trunk/dna-graph/src/test/java/org/jboss/dna/graph/query/optimize/AddAccessNodesTest.java 2009-11-16 23:23:02 UTC (rev 1318) +++ trunk/dna-graph/src/test/java/org/jboss/dna/graph/query/optimize/AddAccessNodesTest.java 2009-11-16 23:24:06 UTC (rev 1319) @@ -31,9 +31,6 @@ import org.jboss.dna.graph.ExecutionContext; import org.jboss.dna.graph.query.AbstractQueryTest; import org.jboss.dna.graph.query.QueryContext; -import org.jboss.dna.graph.query.optimize.AddAccessNodes; -import org.jboss.dna.graph.query.optimize.OptimizerRule; -import org.jboss.dna.graph.query.plan.PlanHints; import org.jboss.dna.graph.query.plan.PlanNode; import org.jboss.dna.graph.query.plan.PlanNode.Type; import org.jboss.dna.graph.query.validate.Schemata; @@ -50,7 +47,7 @@ @Before public void beforeEach() { - context = new QueryContext(new ExecutionContext(), new PlanHints(), mock(Schemata.class)); + context = new QueryContext(new ExecutionContext(), mock(Schemata.class)); rule = AddAccessNodes.INSTANCE; } Modified: trunk/dna-graph/src/test/java/org/jboss/dna/graph/query/optimize/ChooseJoinAlgorithmTest.java =================================================================== --- trunk/dna-graph/src/test/java/org/jboss/dna/graph/query/optimize/ChooseJoinAlgorithmTest.java 2009-11-16 23:23:02 UTC (rev 1318) +++ trunk/dna-graph/src/test/java/org/jboss/dna/graph/query/optimize/ChooseJoinAlgorithmTest.java 2009-11-16 23:24:06 UTC (rev 1319) @@ -35,10 +35,7 @@ import org.jboss.dna.graph.query.model.DescendantNodeJoinCondition; import org.jboss.dna.graph.query.model.JoinCondition; import org.jboss.dna.graph.query.model.JoinType; -import org.jboss.dna.graph.query.optimize.ChooseJoinAlgorithm; -import org.jboss.dna.graph.query.optimize.OptimizerRule; import org.jboss.dna.graph.query.plan.JoinAlgorithm; -import org.jboss.dna.graph.query.plan.PlanHints; import org.jboss.dna.graph.query.plan.PlanNode; import org.jboss.dna.graph.query.plan.PlanNode.Property; import org.jboss.dna.graph.query.plan.PlanNode.Type; @@ -57,7 +54,7 @@ @Before public void beforeEach() { - context = new QueryContext(new ExecutionContext(), new PlanHints(), mock(Schemata.class)); + context = new QueryContext(new ExecutionContext(), mock(Schemata.class)); bestRule = ChooseJoinAlgorithm.USE_BEST_JOIN_ALGORITHM; nestedRule = ChooseJoinAlgorithm.USE_ONLY_NESTED_JOIN_ALGORITHM; } Modified: trunk/dna-graph/src/test/java/org/jboss/dna/graph/query/optimize/PushSelectCriteriaTest.java =================================================================== --- trunk/dna-graph/src/test/java/org/jboss/dna/graph/query/optimize/PushSelectCriteriaTest.java 2009-11-16 23:23:02 UTC (rev 1318) +++ trunk/dna-graph/src/test/java/org/jboss/dna/graph/query/optimize/PushSelectCriteriaTest.java 2009-11-16 23:24:06 UTC (rev 1319) @@ -32,9 +32,6 @@ import org.jboss.dna.graph.query.AbstractQueryTest; import org.jboss.dna.graph.query.QueryContext; import org.jboss.dna.graph.query.model.JoinType; -import org.jboss.dna.graph.query.optimize.OptimizerRule; -import org.jboss.dna.graph.query.optimize.PushSelectCriteria; -import org.jboss.dna.graph.query.plan.PlanHints; import org.jboss.dna.graph.query.plan.PlanNode; import org.jboss.dna.graph.query.plan.PlanNode.Property; import org.jboss.dna.graph.query.plan.PlanNode.Type; @@ -52,7 +49,7 @@ @Before public void beforeEach() { - context = new QueryContext(new ExecutionContext(), new PlanHints(), mock(Schemata.class)); + context = new QueryContext(new ExecutionContext(), mock(Schemata.class)); rule = PushSelectCriteria.INSTANCE; } Modified: trunk/dna-graph/src/test/java/org/jboss/dna/graph/query/optimize/ReplaceViewsTest.java =================================================================== --- trunk/dna-graph/src/test/java/org/jboss/dna/graph/query/optimize/ReplaceViewsTest.java 2009-11-16 23:23:02 UTC (rev 1318) +++ trunk/dna-graph/src/test/java/org/jboss/dna/graph/query/optimize/ReplaceViewsTest.java 2009-11-16 23:24:06 UTC (rev 1319) @@ -38,7 +38,6 @@ import org.jboss.dna.graph.query.model.Operator; import org.jboss.dna.graph.query.model.PropertyValue; import org.jboss.dna.graph.query.model.SelectorName; -import org.jboss.dna.graph.query.plan.PlanHints; import org.jboss.dna.graph.query.plan.PlanNode; import org.jboss.dna.graph.query.plan.PlanNode.Property; import org.jboss.dna.graph.query.plan.PlanNode.Type; @@ -67,7 +66,7 @@ builder.addView("v1", "SELECT c11, c12 FROM t1 WHERE c13 < CAST('3' AS LONG)"); builder.addView("v2", "SELECT t1.c11, t1.c12, t2.c23 FROM t1 JOIN t2 ON t1.c11 = t2.c21"); schemata = builder.build(); - context = new QueryContext(execContext, new PlanHints(), schemata); + context = new QueryContext(execContext, schemata); } /** Modified: trunk/dna-graph/src/test/java/org/jboss/dna/graph/query/optimize/RightOuterToLeftOuterJoinsTest.java =================================================================== --- trunk/dna-graph/src/test/java/org/jboss/dna/graph/query/optimize/RightOuterToLeftOuterJoinsTest.java 2009-11-16 23:23:02 UTC (rev 1318) +++ trunk/dna-graph/src/test/java/org/jboss/dna/graph/query/optimize/RightOuterToLeftOuterJoinsTest.java 2009-11-16 23:24:06 UTC (rev 1319) @@ -32,9 +32,6 @@ import org.jboss.dna.graph.query.AbstractQueryTest; import org.jboss.dna.graph.query.QueryContext; import org.jboss.dna.graph.query.model.JoinType; -import org.jboss.dna.graph.query.optimize.OptimizerRule; -import org.jboss.dna.graph.query.optimize.RightOuterToLeftOuterJoins; -import org.jboss.dna.graph.query.plan.PlanHints; import org.jboss.dna.graph.query.plan.PlanNode; import org.jboss.dna.graph.query.plan.PlanNode.Property; import org.jboss.dna.graph.query.plan.PlanNode.Type; @@ -52,7 +49,7 @@ @Before public void beforeEach() { - context = new QueryContext(new ExecutionContext(), new PlanHints(), mock(Schemata.class)); + context = new QueryContext(new ExecutionContext(), mock(Schemata.class)); rule = RightOuterToLeftOuterJoins.INSTANCE; } Modified: trunk/dna-graph/src/test/java/org/jboss/dna/graph/query/optimize/RuleBasedOptimizerTest.java =================================================================== --- trunk/dna-graph/src/test/java/org/jboss/dna/graph/query/optimize/RuleBasedOptimizerTest.java 2009-11-16 23:23:02 UTC (rev 1318) +++ trunk/dna-graph/src/test/java/org/jboss/dna/graph/query/optimize/RuleBasedOptimizerTest.java 2009-11-16 23:24:06 UTC (rev 1319) @@ -86,7 +86,7 @@ builder.addView("type2", "SELECT all.a3, all.a4 FROM all WHERE all.primaryType IN ('t2','t0') AND all.mixins IN ('t4','t5')"); Schemata schemata = builder.build(); - context = new QueryContext(execContext, new PlanHints(), schemata); + context = new QueryContext(execContext, schemata); node = new PlanNode(Type.ACCESS); Modified: trunk/dna-graph/src/test/java/org/jboss/dna/graph/query/plan/CanonicalPlannerTest.java =================================================================== --- trunk/dna-graph/src/test/java/org/jboss/dna/graph/query/plan/CanonicalPlannerTest.java 2009-11-16 23:23:02 UTC (rev 1318) +++ trunk/dna-graph/src/test/java/org/jboss/dna/graph/query/plan/CanonicalPlannerTest.java 2009-11-16 23:24:06 UTC (rev 1319) @@ -112,7 +112,7 @@ public void shouldProducePlanForSelectStarFromTable() { schemata = schemataBuilder.addTable("__ALLNODES__", "column1", "column2", "column3").build(); query = builder.selectStar().fromAllNodes().query(); - queryContext = new QueryContext(context, hints, schemata, problems); + queryContext = new QueryContext(context, schemata, hints, problems); plan = planner.createPlan(queryContext, query); assertThat(problems.isEmpty(), is(true)); assertProjectNode(plan, "column1", "column2", "column3"); @@ -128,7 +128,7 @@ public void shouldProduceErrorWhenSelectingNonExistantTable() { schemata = schemataBuilder.addTable("someTable", "column1", "column2", "column3").build(); query = builder.selectStar().fromAllNodes().query(); - queryContext = new QueryContext(context, hints, schemata, problems); + queryContext = new QueryContext(context, schemata, hints, problems); plan = planner.createPlan(queryContext, query); assertThat(problems.hasErrors(), is(true)); } @@ -137,7 +137,7 @@ public void shouldProduceErrorWhenSelectingNonExistantColumnOnExistingTable() { schemata = schemataBuilder.addTable("someTable", "column1", "column2", "column3").build(); query = builder.select("column1", "column4").from("someTable").query(); - queryContext = new QueryContext(context, hints, schemata, problems); + queryContext = new QueryContext(context, schemata, hints, problems); plan = planner.createPlan(queryContext, query); assertThat(problems.hasErrors(), is(true)); } @@ -146,7 +146,7 @@ public void shouldProducePlanWhenSelectingAllColumnsOnExistingTable() { schemata = schemataBuilder.addTable("someTable", "column1", "column2", "column3").build(); query = builder.selectStar().from("someTable").query(); - queryContext = new QueryContext(context, hints, schemata, problems); + queryContext = new QueryContext(context, schemata, hints, problems); plan = planner.createPlan(queryContext, query); System.out.println(plan); assertThat(problems.hasErrors(), is(false)); Modified: trunk/dna-graph/src/test/java/org/jboss/dna/graph/query/process/DistinctComponentTest.java =================================================================== --- trunk/dna-graph/src/test/java/org/jboss/dna/graph/query/process/DistinctComponentTest.java 2009-11-16 23:23:02 UTC (rev 1318) +++ trunk/dna-graph/src/test/java/org/jboss/dna/graph/query/process/DistinctComponentTest.java 2009-11-16 23:24:06 UTC (rev 1319) @@ -31,9 +31,6 @@ import org.jboss.dna.graph.ExecutionContext; import org.jboss.dna.graph.query.QueryContext; import org.jboss.dna.graph.query.QueryResults.Columns; -import org.jboss.dna.graph.query.plan.PlanHints; -import org.jboss.dna.graph.query.process.DistinctComponent; -import org.jboss.dna.graph.query.process.ProcessingComponent; import org.jboss.dna.graph.query.validate.Schemata; import org.junit.Before; import org.junit.Test; @@ -50,7 +47,7 @@ @Before public void beforeEach() { - context = new QueryContext(new ExecutionContext(), new PlanHints(), mock(Schemata.class)); + context = new QueryContext(new ExecutionContext(), mock(Schemata.class)); inputTuples = new ArrayList<Object[]>(); // Define the columns for the results ... columns = resultColumns("Selector1", "ColA", "ColB", "ColC"); Modified: trunk/dna-graph/src/test/java/org/jboss/dna/graph/query/process/QueryResultColumnsTest.java =================================================================== --- trunk/dna-graph/src/test/java/org/jboss/dna/graph/query/process/QueryResultColumnsTest.java 2009-11-16 23:23:02 UTC (rev 1318) +++ trunk/dna-graph/src/test/java/org/jboss/dna/graph/query/process/QueryResultColumnsTest.java 2009-11-16 23:24:06 UTC (rev 1319) @@ -37,9 +37,6 @@ import org.jboss.dna.graph.query.QueryResults.Statistics; import org.jboss.dna.graph.query.model.Column; import org.jboss.dna.graph.query.model.QueryCommand; -import org.jboss.dna.graph.query.plan.PlanHints; -import org.jboss.dna.graph.query.process.QueryResultColumns; -import org.jboss.dna.graph.query.process.QueryResults; import org.jboss.dna.graph.query.validate.Schemata; import org.junit.Before; import org.junit.Test; @@ -66,7 +63,7 @@ @Before public void beforeEach() { MockitoAnnotations.initMocks(this); - context = new QueryContext(executionContext, new PlanHints(), schemata); + context = new QueryContext(executionContext, schemata); columnList = new ArrayList<Column>(); columnList.add(new Column(selector("table1"), name("colA"), "colA")); columnList.add(new Column(selector("table1"), name("colB"), "colB")); Modified: trunk/dna-graph/src/test/java/org/jboss/dna/graph/query/process/SortLocationsComponentTest.java =================================================================== --- trunk/dna-graph/src/test/java/org/jboss/dna/graph/query/process/SortLocationsComponentTest.java 2009-11-16 23:23:02 UTC (rev 1318) +++ trunk/dna-graph/src/test/java/org/jboss/dna/graph/query/process/SortLocationsComponentTest.java 2009-11-16 23:24:06 UTC (rev 1319) @@ -31,9 +31,6 @@ import org.jboss.dna.graph.ExecutionContext; import org.jboss.dna.graph.query.QueryContext; import org.jboss.dna.graph.query.QueryResults.Columns; -import org.jboss.dna.graph.query.plan.PlanHints; -import org.jboss.dna.graph.query.process.ProcessingComponent; -import org.jboss.dna.graph.query.process.SortLocationsComponent; import org.jboss.dna.graph.query.validate.Schemata; import org.junit.Before; import org.junit.Test; @@ -50,7 +47,7 @@ @Before public void beforeEach() { - context = new QueryContext(new ExecutionContext(), new PlanHints(), mock(Schemata.class)); + context = new QueryContext(new ExecutionContext(), mock(Schemata.class)); inputTuples = new ArrayList<Object[]>(); // Define the columns for the results ... columns = resultColumns("Selector1", "ColA", "ColB", "ColC"); Modified: trunk/dna-graph/src/test/java/org/jboss/dna/graph/query/process/SortValuesComponentTest.java =================================================================== --- trunk/dna-graph/src/test/java/org/jboss/dna/graph/query/process/SortValuesComponentTest.java 2009-11-16 23:23:02 UTC (rev 1318) +++ trunk/dna-graph/src/test/java/org/jboss/dna/graph/query/process/SortValuesComponentTest.java 2009-11-16 23:24:06 UTC (rev 1319) @@ -32,7 +32,6 @@ import org.jboss.dna.graph.query.QueryContext; import org.jboss.dna.graph.query.QueryResults.Columns; import org.jboss.dna.graph.query.model.Ordering; -import org.jboss.dna.graph.query.plan.PlanHints; import org.jboss.dna.graph.query.validate.Schemata; import org.junit.Before; import org.junit.Test; @@ -56,7 +55,7 @@ columns = resultColumns("Selector1", "ColA", "ColB", "ColC"); schemata = schemataFor(columns, PropertyType.STRING, PropertyType.LONG, PropertyType.STRING); // Define the context ... - context = new QueryContext(new ExecutionContext(), new PlanHints(), schemata); + context = new QueryContext(new ExecutionContext(), schemata); inputTuples = new ArrayList<Object[]>(); // And define the delegating component ... delegate = new ProcessingComponent(context, columns) { Copied: trunk/dna-search/src/main/java/org/jboss/dna/search/DualIndexStrategy.java (from rev 1318, trunk/dna-search/src/main/java/org/jboss/dna/search/StoreLittleIndexingStrategy.java) =================================================================== --- trunk/dna-search/src/main/java/org/jboss/dna/search/DualIndexStrategy.java (rev 0) +++ trunk/dna-search/src/main/java/org/jboss/dna/search/DualIndexStrategy.java 2009-11-16 23:24:06 UTC (rev 1319) @@ -0,0 +1,399 @@ +/* + * JBoss DNA (http://www.jboss.org/dna) + * See the COPYRIGHT.txt file distributed with this work for information + * regarding copyright ownership. Some portions may be licensed + * to Red Hat, Inc. under one or more contributor license agreements. + * See the AUTHORS.txt file in the distribution for a full listing of + * individual contributors. + * + * JBoss DNA is free software. Unless otherwise indicated, all code in JBoss DNA + * is licensed to you under the terms of the GNU Lesser General Public License as + * published by the Free Software Foundation; either version 2.1 of + * the License, or (at your option) any later version. + * + * JBoss DNA is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public + * License along with this software; if not, write to the Free + * Software Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA + * 02110-1301 USA, or see the FSF site: http://www.fsf.org. + */ +package org.jboss.dna.search; + +import java.io.FileNotFoundException; +import java.io.IOException; +import java.text.DateFormat; +import java.text.SimpleDateFormat; +import java.util.LinkedList; +import java.util.List; +import java.util.UUID; +import net.jcip.annotations.ThreadSafe; +import org.apache.lucene.analysis.Analyzer; +import org.apache.lucene.analysis.standard.StandardAnalyzer; +import org.apache.lucene.document.Document; +import org.apache.lucene.document.Field; +import org.apache.lucene.document.FieldSelector; +import org.apache.lucene.document.FieldSelectorResult; +import org.apache.lucene.index.IndexReader; +import org.apache.lucene.index.Term; +import org.apache.lucene.queryParser.ParseException; +import org.apache.lucene.queryParser.QueryParser; +import org.apache.lucene.search.IndexSearcher; +import org.apache.lucene.search.PrefixQuery; +import org.apache.lucene.search.Query; +import org.apache.lucene.search.ScoreDoc; +import org.apache.lucene.search.TermQuery; +import org.apache.lucene.search.TopDocs; +import org.apache.lucene.util.Version; +import org.jboss.dna.common.text.NoOpEncoder; +import org.jboss.dna.common.text.TextEncoder; +import org.jboss.dna.common.util.Logger; +import org.jboss.dna.graph.Location; +import org.jboss.dna.graph.Node; +import org.jboss.dna.graph.property.Binary; +import org.jboss.dna.graph.property.DateTime; +import org.jboss.dna.graph.property.DateTimeFactory; +import org.jboss.dna.graph.property.Name; +import org.jboss.dna.graph.property.Path; +import org.jboss.dna.graph.property.Property; +import org.jboss.dna.graph.property.ValueFactory; +import org.jboss.dna.graph.query.QueryContext; +import org.jboss.dna.graph.query.QueryEngine; +import org.jboss.dna.graph.query.QueryResults; +import org.jboss.dna.graph.query.QueryResults.Columns; +import org.jboss.dna.graph.query.model.QueryCommand; +import org.jboss.dna.graph.query.optimize.Optimizer; +import org.jboss.dna.graph.query.optimize.OptimizerRule; +import org.jboss.dna.graph.query.optimize.RuleBasedOptimizer; +import org.jboss.dna.graph.query.plan.CanonicalPlanner; +import org.jboss.dna.graph.query.plan.PlanHints; +import org.jboss.dna.graph.query.plan.PlanNode; +import org.jboss.dna.graph.query.plan.Planner; +import org.jboss.dna.graph.query.process.ProcessingComponent; +import org.jboss.dna.graph.query.process.QueryProcessor; +import org.jboss.dna.search.IndexRules.Rule; + +/** + * A simple {@link IndexStrategy} implementation that relies upon two separate indexes: one for the node content and a second + * one for paths and UUIDs. + */ +@ThreadSafe +abstract class DualIndexStrategy implements IndexStrategy { + + static class PathIndex { + public static final String PATH = "path"; + public static final String UUID = "uuid"; + } + + static class ContentIndex { + public static final String UUID = PathIndex.UUID; + public static final String FULL_TEXT = "fts"; + } + + /** + * The number of results that should be returned when performing queries while deleting entire branches of content. The + * current value is {@value} . + */ + protected static final int SIZE_OF_DELETE_BATCHES = 1000; + + private ThreadLocal<DateFormat> dateFormatter = new ThreadLocal<DateFormat>() { + @Override + protected DateFormat initialValue() { + return new SimpleDateFormat("yyyyMMdd'T'HH:mm:ss"); + } + }; + + /** + * Obtain an immutable {@link FieldSelector} instance that accesses the UUID field. + */ + protected static final FieldSelector UUID_FIELD_SELECTOR = new FieldSelector() { + private static final long serialVersionUID = 1L; + + public FieldSelectorResult accept( String fieldName ) { + return PathIndex.UUID.equals(fieldName) ? FieldSelectorResult.LOAD_AND_BREAK : FieldSelectorResult.NO_LOAD; + } + }; + + private final IndexRules rules; + private final Logger logger; + private final QueryEngine queryEngine; + + /** + * Create a new indexing strategy instance. + * + * @param rules the indexing rules that govern how properties are to be index; may not be null + */ + protected DualIndexStrategy( IndexRules rules ) { + assert rules != null; + this.rules = rules; + this.logger = Logger.getLogger(getClass()); + // Create the query engine ... + Planner planner = new CanonicalPlanner(); + Optimizer optimizer = new RuleBasedOptimizer() { + /** + * {@inheritDoc} + * + * @see org.jboss.dna.graph.query.optimize.RuleBasedOptimizer#populateRuleStack(java.util.LinkedList, + * org.jboss.dna.graph.query.plan.PlanHints) + */ + @Override + protected void populateRuleStack( LinkedList<OptimizerRule> ruleStack, + PlanHints hints ) { + super.populateRuleStack(ruleStack, hints); + // Add any custom rules here, either at the front of the stack or at the end + } + }; + QueryProcessor processor = new QueryProcessor() { + /** + * {@inheritDoc} + * + * @see org.jboss.dna.graph.query.process.QueryProcessor#createAccessComponent(org.jboss.dna.graph.query.QueryContext, + * org.jboss.dna.graph.query.plan.PlanNode, org.jboss.dna.graph.query.QueryResults.Columns, + * org.jboss.dna.graph.query.process.SelectComponent.Analyzer) + */ + @Override + protected ProcessingComponent createAccessComponent( QueryContext context, + PlanNode accessNode, + Columns resultColumns, + org.jboss.dna.graph.query.process.SelectComponent.Analyzer analyzer ) { + return DualIndexStrategy.this.createAccessComponent((SearchContext)context, accessNode, resultColumns, analyzer); + } + }; + + this.queryEngine = new QueryEngine(planner, optimizer, processor); + } + + protected abstract ProcessingComponent createAccessComponent( SearchContext context, + PlanNode accessNode, + Columns resultColumns, + org.jboss.dna.graph.query.process.SelectComponent.Analyzer analyzer ); + + /** + * Utility method to obtain a {@link DateFormat} instance that can be used safely within a single thread. + * + * @return the date formatter; never null + */ + protected final DateFormat dateFormatter() { + return dateFormatter.get(); + } + + /** + * {@inheritDoc} + * + * @see org.jboss.dna.search.IndexStrategy#getNamespaceEncoder() + */ + public TextEncoder getNamespaceEncoder() { + return new NoOpEncoder(); + } + + /** + * {@inheritDoc} + * + * @see org.jboss.dna.search.IndexStrategy#getChangeCountForAutomaticOptimization() + */ + public int getChangeCountForAutomaticOptimization() { + return 0; + } + + /** + * {@inheritDoc} + * + * @see org.jboss.dna.search.IndexStrategy#createAnalyzer() + */ + public Analyzer createAnalyzer() { + return new StandardAnalyzer(Version.LUCENE_CURRENT); + } + + /** + * {@inheritDoc} + * <p> + * Because this strategy uses multiple indexes, and since there's no correlation between the documents in those indexes, we + * need to perform the delete in multiple steps. First, we need to perform a query to find out which nodes exist below a + * certain path. Then, we need to delete those nodes from the paths index. Finally, we need to delete the corresponding + * documents in the content index that represent those same nodes. + * </p> + * <p> + * Since we don't know how many documents there will be, we perform these steps in batches, where each batch limits the number + * of results to a maximum number. We repeat batches as long as we find more results. This approach has the advantage that + * we'll never bring in a large number of results, and it allows us to delete the documents from the content node using a + * query. + * </p> + * + * @see org.jboss.dna.search.IndexStrategy#deleteBelow(Path, IndexContext) + */ + public int deleteBelow( Path path, + IndexContext indexes ) throws IOException { + // Perform a query using the reader to find those nodes at/below the path ... + try { + IndexReader pathReader = indexes.getPathsReader(); + IndexSearcher pathSearcher = new IndexSearcher(pathReader); + String pathStr = indexes.stringFactory().create(path) + "/"; + PrefixQuery query = new PrefixQuery(new Term(PathIndex.PATH, pathStr)); + int numberDeleted = 0; + while (true) { + // Execute the query and get the results ... + TopDocs results = pathSearcher.search(query, SIZE_OF_DELETE_BATCHES); + int numResultsInBatch = results.scoreDocs.length; + // Walk the results, delete the doc, and add to the query that we'll use against the content index ... + IndexReader contentReader = indexes.getContentReader(); + for (ScoreDoc result : results.scoreDocs) { + int docId = result.doc; + // Find the UUID of the node ... + Document doc = pathReader.document(docId, UUID_FIELD_SELECTOR); + String uuid = doc.get(PathIndex.UUID); + // Delete the document from the paths index ... + pathReader.deleteDocument(docId); + // Delete the corresponding document from the content index ... + contentReader.deleteDocuments(new Term(ContentIndex.UUID, uuid)); + } + numberDeleted += numResultsInBatch; + if (numResultsInBatch < SIZE_OF_DELETE_BATCHES) break; + } + indexes.commit(); + return numberDeleted; + } catch (FileNotFoundException e) { + // There are no index files yet, so nothing to delete ... + return 0; + } + } + + /** + * {@inheritDoc} + * + * @see org.jboss.dna.search.IndexStrategy#index(Node, IndexContext) + */ + public void index( Node node, + IndexContext indexes ) throws IOException { + ValueFactory<String> strings = indexes.stringFactory(); + Location location = node.getLocation(); + UUID uuid = location.getUuid(); + if (uuid == null) uuid = UUID.randomUUID(); + Path path = location.getPath(); + String pathStr = path.isRoot() ? "/" : strings.create(location.getPath()) + "/"; + String uuidStr = uuid.toString(); + + if (logger.isTraceEnabled()) { + logger.trace("indexing {0}", pathStr); + } + + // Create a separate document for the path, which makes it easier to handle moves since the path can + // be changed without changing any other content fields ... + Document doc = new Document(); + doc.add(new Field(PathIndex.PATH, pathStr, Field.Store.YES, Field.Index.NOT_ANALYZED)); + doc.add(new Field(PathIndex.UUID, uuidStr, Field.Store.YES, Field.Index.NOT_ANALYZED)); + indexes.getPathsWriter().addDocument(doc); + + // Create the document for the content (properties) ... + doc = new Document(); + doc.add(new Field(ContentIndex.UUID, uuidStr, Field.Store.YES, Field.Index.NOT_ANALYZED)); + String stringValue = null; + StringBuilder fullTextSearchValue = null; + for (Property property : node.getProperties()) { + Name name = property.getName(); + Rule rule = rules.getRule(name); + if (rule.isSkipped()) continue; + String nameString = strings.create(name); + if (rule.isDate()) { + DateTimeFactory dateFactory = indexes.dateFactory(); + for (Object value : property) { + if (value == null) continue; + DateTime dateValue = dateFactory.create(value); + stringValue = dateFormatter().format(dateValue.toDate()); + // Add a separate field for each property value ... + doc.add(new Field(nameString, stringValue, rule.getStoreOption(), rule.getIndexOption())); + // Dates are not added to the full-text search field (since this wouldn't make sense) + } + continue; + } + for (Object value : property) { + if (value == null) continue; + if (value instanceof Binary) { + // don't include binary values as individual fields but do include them in the full-text search ... + // TODO : add to full-text search ... + continue; + } + stringValue = strings.create(value); + // Add a separate field for each property value ... + doc.add(new Field(nameString, stringValue, rule.getStoreOption(), rule.getIndexOption())); + // And add to the full-text field ... + if (rule.isFullText()) { + if (fullTextSearchValue == null) { + fullTextSearchValue = new StringBuilder(); + } else { + fullTextSearchValue.append(' '); + } + fullTextSearchValue.append(stringValue); + } + } + } + // Add the full-text-search field ... + if (fullTextSearchValue != null) { + doc.add(new Field(ContentIndex.FULL_TEXT, fullTextSearchValue.toString(), Field.Store.NO, Field.Index.ANALYZED)); + } + indexes.getContentWriter().addDocument(doc); + } + + /** + * {@inheritDoc} + * + * @see org.jboss.dna.search.IndexStrategy#search(String, int, int, IndexContext, List) + */ + public void search( String fullTextString, + int maxResults, + int offset, + IndexContext indexes, + List<Location> results ) throws IOException, ParseException { + assert fullTextString != null; + assert fullTextString.length() > 0; + assert offset >= 0; + assert maxResults > 0; + assert indexes != null; + assert results != null; + + // Parse the full-text search and search against the 'fts' field ... + QueryParser parser = new QueryParser(ContentIndex.FULL_TEXT, createAnalyzer()); + Query query = parser.parse(fullTextString); + TopDocs docs = indexes.getContentSearcher().search(query, maxResults + offset); + + // Collect the results ... + IndexReader contentReader = indexes.getContentReader(); + IndexReader pathReader = indexes.getPathsReader(); + IndexSearcher pathSearcher = indexes.getPathsSearcher(); + ScoreDoc[] scoreDocs = docs.scoreDocs; + int numberOfResults = scoreDocs.length; + if (numberOfResults > offset) { + // There are enough results to satisfy the offset ... + for (int i = offset, num = scoreDocs.length; i != num; ++i) { + ScoreDoc result = scoreDocs[i]; + int docId = result.doc; + // Find the UUID of the node (this UUID might be artificial, so we have to find the path) ... + Document doc = contentReader.document(docId, UUID_FIELD_SELECTOR); + String uuid = doc.get(ContentIndex.UUID); + // Find the path for this node (is there a better way to do this than one search per UUID?) ... + TopDocs pathDocs = pathSearcher.search(new TermQuery(new Term(PathIndex.UUID, uuid)), 1); + if (pathDocs.scoreDocs.length < 1) { + // No path record found ... + continue; + } + Document pathDoc = pathReader.document(pathDocs.scoreDocs[0].doc); + Path path = indexes.pathFactory().create(pathDoc.get(PathIndex.PATH)); + // Now add the location ... + results.add(Location.create(path, UUID.fromString(uuid))); + } + } + } + + /** + * {@inheritDoc} + * + * @see org.jboss.dna.search.IndexStrategy#query(org.jboss.dna.search.SearchContext, + * org.jboss.dna.graph.query.model.QueryCommand) + */ + public QueryResults query( SearchContext context, + QueryCommand query ) { + return this.queryEngine.execute(context, query); + } +} Property changes on: trunk/dna-search/src/main/java/org/jboss/dna/search/DualIndexStrategy.java ___________________________________________________________________ Name: svn:keywords + Id Revision Name: svn:eol-style + LF Modified: trunk/dna-search/src/main/java/org/jboss/dna/search/IndexContext.java =================================================================== --- trunk/dna-search/src/main/java/org/jboss/dna/search/IndexContext.java 2009-11-16 23:23:02 UTC (rev 1318) +++ trunk/dna-search/src/main/java/org/jboss/dna/search/IndexContext.java 2009-11-16 23:24:06 UTC (rev 1319) @@ -31,16 +31,17 @@ import org.apache.lucene.index.IndexWriter.MaxFieldLength; import org.apache.lucene.search.IndexSearcher; import org.apache.lucene.store.Directory; +import org.jboss.dna.common.util.CheckArg; import org.jboss.dna.graph.ExecutionContext; import org.jboss.dna.graph.property.DateTimeFactory; import org.jboss.dna.graph.property.PathFactory; import org.jboss.dna.graph.property.ValueFactory; /** - * A set of index readers and writers. + * A context for working with the index readers and writers. */ @NotThreadSafe -public final class IndexContext { +final class IndexContext { private final ExecutionContext context; private final Directory pathsIndexDirectory; @@ -278,4 +279,16 @@ if (runtimeError != null) throw runtimeError; } + /** + * Create a copy of this index context, except that it uses the supplied execution context. + * + * @param context the new execution context that should be used in the copy + * @return the new context; never null + * @throws IllegalArgumentException if the context is null + */ + public IndexContext with( ExecutionContext context ) { + CheckArg.isNotNull(context, "context"); + return new IndexContext(context, pathsIndexDirectory, contentIndexDirectory, analyzer, overwrite, readOnly); + } + } Copied: trunk/dna-search/src/main/java/org/jboss/dna/search/IndexRules.java (from rev 1318, trunk/dna-search/src/main/java/org/jboss/dna/search/IndexingRules.java) =================================================================== --- trunk/dna-search/src/main/java/org/jboss/dna/search/IndexRules.java (rev 0) +++ trunk/dna-search/src/main/java/org/jboss/dna/search/IndexRules.java 2009-11-16 23:24:06 UTC (rev 1319) @@ -0,0 +1,628 @@ +/* + * JBoss DNA (http://www.jboss.org/dna) + * See the COPYRIGHT.txt file distributed with this work for information + * regarding copyright ownership. Some portions may be licensed + * to Red Hat, Inc. under one or more contributor license agreements. + * See the AUTHORS.txt file in the distribution for a full listing of + * individual contributors. + * + * JBoss DNA is free software. Unless otherwise indicated, all code in JBoss DNA + * is licensed to you under the terms of the GNU Lesser General Public License as + * published by the Free Software Foundation; either version 2.1 of + * the License, or (at your option) any later version. + * + * JBoss DNA is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public + * License along with this software; if not, write to the Free + * Software Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA + * 02110-1301 USA, or see the FSF site: http://www.fsf.org. + */ +package org.jboss.dna.search; + +import java.util.Collections; +import java.util.HashMap; +import java.util.Map; +import java.util.concurrent.ConcurrentHashMap; +import net.jcip.annotations.Immutable; +import net.jcip.annotations.NotThreadSafe; +import org.apache.lucene.document.Field; +import org.apache.lucene.document.Field.Index; +import org.apache.lucene.document.Field.Store; +import org.jboss.dna.common.util.CheckArg; +import org.jboss.dna.graph.property.Name; + +/** + * The set of rules that dictate how properties should be indexed. + */ +@Immutable +class IndexRules { + + public static final int INDEX = 2 << 0; + public static final int ANALYZE = 2 << 1; + public static final int STORE = 2 << 2; + public static final int STORE_COMPRESSED = 2 << 3; + public static final int ANALYZED_WITHOUT_NORMS = 2 << 4; + public static final int FULL_TEXT = 2 << 5; + public static final int TREAT_AS_DATE = 2 << 6; + + /** + * A single rule that dictates how a single property should be indexed. + * + * @see IndexRules#getRule(Name) + */ + @Immutable + public static interface Rule { + boolean isIncluded(); + + boolean isSkipped(); + + boolean isAnalyzed(); + + boolean isAnalyzedWithoutNorms(); + + boolean isStored(); + + boolean isStoredCompressed(); + + boolean isFullText(); + + boolean isDate(); + + int getMask(); + + Field.Store getStoreOption(); + + Field.Index getIndexOption(); + } + + public static final Rule SKIP = new SkipRule(); + + @Immutable + protected static class SkipRule implements Rule { + /** + * {@inheritDoc} + * + * @see org.jboss.dna.search.IndexRules.Rule#getMask() + */ + public int getMask() { + return 0; + } + + /** + * {@inheritDoc} + * + * @see org.jboss.dna.search.IndexRules.Rule#isAnalyzed() + */ + public boolean isAnalyzed() { + return false; + } + + /** + * {@inheritDoc} + * + * @see org.jboss.dna.search.IndexRules.Rule#isAnalyzedWithoutNorms() + */ + public boolean isAnalyzedWithoutNorms() { + return false; + } + + /** + * {@inheritDoc} + * + * @see org.jboss.dna.search.IndexRules.Rule#isFullText() + */ + public boolean isFullText() { + return false; + } + + /** + * {@inheritDoc} + * + * @see org.jboss.dna.search.IndexRules.Rule#isIncluded() + */ + public boolean isIncluded() { + return false; + } + + /** + * {@inheritDoc} + * + * @see org.jboss.dna.search.IndexRules.Rule#isSkipped() + */ + public boolean isSkipped() { + return true; + } + + /** + * {@inheritDoc} + * + * @see org.jboss.dna.search.IndexRules.Rule#isStored() + */ + public boolean isStored() { + return false; + } + + /** + * {@inheritDoc} + * + * @see org.jboss.dna.search.IndexRules.Rule#isStoredCompressed() + */ + public boolean isStoredCompressed() { + return false; + } + + /** + * {@inheritDoc} + * + * @see org.jboss.dna.search.IndexRules.Rule#isDate() + */ + public boolean isDate() { + return false; + } + + /** + * {@inheritDoc} + * + * @see org.jboss.dna.search.IndexRules.Rule#getIndexOption() + */ + public Index getIndexOption() { + return Field.Index.NO; + } + + /** + * {@inheritDoc} + * + * @see org.jboss.dna.search.IndexRules.Rule#getStoreOption() + */ + public Store getStoreOption() { + return Field.Store.NO; + } + } + + @Immutable + public static final class GeneralRule implements Rule { + private final int value; + private final Field.Store store; + private final Field.Index index; + + protected GeneralRule( int value ) { + this.value = value; + this.index = isAnalyzed() ? Field.Index.ANALYZED : Field.Index.NOT_ANALYZED; + this.store = isStored() ? Field.Store.YES : Field.Store.NO; + } + + /** + * {@inheritDoc} + * + * @see org.jboss.dna.search.IndexRules.Rule#getMask() + */ + public int getMask() { + return value; + } + + /** + * {@inheritDoc} + * + * @see org.jboss.dna.search.IndexRules.Rule#isAnalyzed() + */ + public boolean isAnalyzed() { + return (value & ANALYZE) == ANALYZE; + } + + /** + * {@inheritDoc} + * + * @see org.jboss.dna.search.IndexRules.Rule#isAnalyzedWithoutNorms() + */ + public boolean isAnalyzedWithoutNorms() { + return (value & ANALYZED_WITHOUT_NORMS) == ANALYZED_WITHOUT_NORMS; + } + + /** + * {@inheritDoc} + * + * @see org.jboss.dna.search.IndexRules.Rule#isFullText() + */ + public boolean isFullText() { + return (value & FULL_TEXT) == FULL_TEXT; + } + + /** + * {@inheritDoc} + * + * @see org.jboss.dna.search.IndexRules.Rule#isIncluded() + */ + public boolean isIncluded() { + return true; + } + + /** + * {@inheritDoc} + * + * @see org.jboss.dna.search.IndexRules.Rule#isSkipped() + */ + public boolean isSkipped() { + return false; + } + + /** + * {@inheritDoc} + * + * @see org.jboss.dna.search.IndexRules.Rule#isStored() + */ + public boolean isStored() { + return (value & STORE) == STORE; + } + + /** + * {@inheritDoc} + * + * @see org.jboss.dna.search.IndexRules.Rule#isStoredCompressed() + */ + public boolean isStoredCompressed() { + return (value & STORE_COMPRESSED) == STORE_COMPRESSED; + } + + /** + * {@inheritDoc} + * + * @see org.jboss.dna.search.IndexRules.Rule#isDate() + */ + public boolean isDate() { + return (value & TREAT_AS_DATE) == TREAT_AS_DATE; + } + + protected Rule with( int options ) { + return createRule(value | options); + } + + /** + * {@inheritDoc} + * + * @see org.jboss.dna.search.IndexRules.Rule#getIndexOption() + */ + public Index getIndexOption() { + return index; + } + + /** + * {@inheritDoc} + * + * @see org.jboss.dna.search.IndexRules.Rule#getStoreOption() + */ + public Store getStoreOption() { + return store; + } + } + + private static final ConcurrentHashMap<Integer, Rule> CACHE = new ConcurrentHashMap<Integer, Rule>(); + + protected static Rule createRule( int value ) { + if (value <= 0) { + return SKIP; + } + Integer key = new Integer(value); + Rule rule = CACHE.get(key); + if (rule == null) { + Rule newRule = new GeneralRule(value); + rule = CACHE.putIfAbsent(value, newRule); + if (rule == null) rule = newRule; + } + return rule; + } + + private final Map<Name, Rule> rulesByName; + private final Rule defaultRule; + + protected IndexRules( Map<Name, Rule> rulesByName, + Rule defaultRule ) { + this.rulesByName = rulesByName; + this.defaultRule = defaultRule != null ? defaultRule : SKIP; + assert this.defaultRule != null; + } + + /** + * Get the rule associated with the given property name. + * + * @param name the property name, or null if the default rule is to be returned + * @return the rule; never null + */ + public Rule getRule( Name name ) { + Rule result = rulesByName.get(name); + return result != null ? result : this.defaultRule; + } + + /** + * Return a new builder that can be used to create {@link IndexRules} objects. + * + * @return a builder; never null + */ + public static Builder createBuilder() { + return new Builder(new HashMap<Name, Rule>()); + } + + /** + * Return a new builder that can be used to create {@link IndexRules} objects. + * + * @param initialRules the rules that the builder should start with + * @return a builder; never null + * @throws IllegalArgumentException if the initial rules reference is null + */ + public static Builder createBuilder( IndexRules initialRules ) { + CheckArg.isNotNull(initialRules, "initialRules"); + return new Builder(initialRules.rulesByName).defaultTo(initialRules.defaultRule); + } + + /** + * A builder of immutable {@link IndexRules} objects. + */ + @NotThreadSafe + public static class Builder { + private final Map<Name, Rule> rulesByName; + private Rule defaultRule; + + Builder( Map<Name, Rule> rulesByName ) { + assert rulesByName != null; + this.rulesByName = rulesByName; + } + + /** + * Set the default rules. + * + * @param rule the default rule to use + * @return this builder for convenience and method chaining; never null + * @throws IllegalArgumentException if the rule mask is negative + */ + public Builder defaultTo( Rule rule ) { + CheckArg.isNotNull(rule, "rule"); + defaultRule = rule; + return this; + } + + /** + * Set the default rules. + * + * @param ruleMask the bitmask of rule to use + * @return this builder for convenience and method chaining; never null + * @throws IllegalArgumentException if the rule mask is negative + */ + public Builder defaultTo( int ruleMask ) { + CheckArg.isNonNegative(ruleMask, "options"); + if (ruleMask == 0) { + defaultRule = SKIP; + } else { + // Make sure the index flag is set ... + ruleMask |= INDEX; + defaultRule = createRule(ruleMask); + } + return this; + } + + /** + * Mark the properties with the supplied names to be skipped from indexing. + * + * @param namesToIndex the names of the properties that are to be skipped + * @return this builder for convenience and method chaining; never null + */ + public Builder skip( Name... namesToIndex ) { + if (namesToIndex != null) { + for (Name name : namesToIndex) { + rulesByName.put(name, SKIP); + } + } + return this; + } + + /** + * Set the properties with the supplied names to use the supplied rules. + * + * @param ruleMask the bitmask of rules to use + * @param namesToIndex the names of the properties that are to be skipped + * @return this builder for convenience and method chaining; never null + * @throws IllegalArgumentException if the rule mask is negative + */ + public Builder set( int ruleMask, + Name... namesToIndex ) { + CheckArg.isNonNegative(ruleMask, "options"); + if (namesToIndex != null) { + if (ruleMask > 0) { + skip(namesToIndex); + } else { + // Make sure the index flag is set ... + ruleMask |= INDEX; + Rule rule = createRule(ruleMask); + for (Name name : namesToIndex) { + rulesByName.put(name, rule); + } + } + } + return this; + } + + /** + * Mark the properties with the supplied names to use the supplied rule mask. This does not remove any other rules for + * these properties. + * + * @param ruleMask the bitmask of rules to add + * @param namesToIndex the names of the properties that are to be skipped + * @return this builder for convenience and method chaining; never null + * @throws IllegalArgumentException if the rule mask is negative + */ + public Builder add( int ruleMask, + Name... namesToIndex ) { + CheckArg.isNonNegative(ruleMask, "options"); + if (namesToIndex != null) { + for (Name name : namesToIndex) { + add(name, ruleMask); + } + } + return this; + } + + /** + * Mark the properties with the supplied names to be indexed. This does not remove any other rules for these properties. + * + * @param namesToIndex the names of the properties that are to be indexed + * @return this builder for convenience and method chaining; never null + */ + public Builder index( Name... namesToIndex ) { + if (namesToIndex != null) { + for (Name name : namesToIndex) { + add(name, INDEX); + } + } + return this; + } + + /** + * Mark the properties with the supplied names to be analyzed (and obviously indexed). This does not remove any other + * rules for these properties. + * + * @param namesToIndex the names of the properties that are to be analyzed + * @return this builder for convenience and method chaining; never null + */ + public Builder analyze( Name... namesToIndex ) { + if (namesToIndex != null) { + for (Name name : namesToIndex) { + add(name, ANALYZE | INDEX); + } + } + return this; + } + + /** + * Mark the properties with the supplied names to be stored (and obviously indexed). This does not remove any other rules + * for these properties. + * + * @param namesToIndex the names of the properties that are to be stored + * @return this builder for convenience and method chaining; never null + */ + public Builder store( Name... namesToIndex ) { + if (namesToIndex != null) { + for (Name name : namesToIndex) { + add(name, STORE | INDEX); + } + } + return this; + } + + /** + * Mark the properties with the supplied names to be included in full-text searches (and obviously indexed). This does not + * remove any other rules for these properties. + * + * @param namesToIndex the names of the properties that are to be included in full-text searches + * @return this builder for convenience and method chaining; never null + */ + public Builder fullText( Name... namesToIndex ) { + if (namesToIndex != null) { + for (Name name : namesToIndex) { + add(name, FULL_TEXT | INDEX); + } + } + return this; + } + + /** + * Mark the properties with the supplied names to be treated as dates (and obviously indexed). This does not remove any + * other rules for these properties. + * + * @param namesToIndex the names of the properties that are to be included in full-text searches + * @return this builder for convenience and method chaining; never null + */ + public Builder treatAsDates( Name... namesToIndex ) { + if (namesToIndex != null) { + for (Name name : namesToIndex) { + add(name, TREAT_AS_DATE | INDEX); + } + } + return this; + } + + /** + * Mark the properties with the supplied names to be indexed, analyzed and stored. This does not remove any other rules + * for these properties. + * + * @param namesToIndex the names of the properties that are to be indexed, analyzed and stored + * @return this builder for convenience and method chaining; never null + */ + public Builder analyzeAndStore( Name... namesToIndex ) { + if (namesToIndex != null) { + for (Name name : namesToIndex) { + add(name, INDEX | ANALYZE | STORE); + } + } + return this; + } + + /** + * Mark the properties with the supplied names to be indexed, analyzed, stored and included in full-text searches. This + * does not remove any other rules for these properties. + * + * @param namesToIndex the names of the properties that are to be indexed, analyzed, stored and included in full-text + * searches + * @return this builder for convenience and method chaining; never null + */ + public Builder analyzeAndStoreAndFullText( Name... namesToIndex ) { + if (namesToIndex != null) { + for (Name name : namesToIndex) { + add(name, INDEX | ANALYZE | STORE | FULL_TEXT); + } + } + return this; + } + + /** + * Mark the properties with the supplied names to be indexed, analyzed and included in full-text searches. This does not + * remove any other rules for these properties. + * + * @param namesToIndex the names of the properties that are to be indexed, analyzed and included in full-text searches + * @return this builder for convenience and method chaining; never null + */ + public Builder analyzeAndFullText( Name... namesToIndex ) { + if (namesToIndex != null) { + for (Name name : namesToIndex) { + add(name, INDEX | ANALYZE | FULL_TEXT); + } + } + return this; + } + + /** + * Mark the properties with the supplied names to be indexed, stored and included in full-text searches. This does not + * remove any other rules for these properties. + * + * @param namesToIndex the names of the properties that are to be indexed, stored and included in full-text searches + * @return this builder for convenience and method chaining; never null + */ + public Builder storeAndFullText( Name... namesToIndex ) { + if (namesToIndex != null) { + for (Name name : namesToIndex) { + add(name, INDEX | STORE | FULL_TEXT); + } + } + return this; + } + + protected void add( Name name, + int option ) { + Rule rule = rulesByName.get(name); + if (rule != null) { + option |= rule.getMask(); + } + rulesByName.put(name, createRule(option)); + } + + /** + * Build the indexing rules. + * + * @return the immutable indexing rules. + */ + public IndexRules build() { + return new IndexRules(Collections.unmodifiableMap(new HashMap<Name, Rule>(rulesByName)), defaultRule); + } + } +} Property changes on: trunk/dna-search/src/main/java/org/jboss/dna/search/IndexRules.java ___________________________________________________________________ Name: svn:keywords + Id Revision Name: svn:eol-style + LF Copied: trunk/dna-search/src/main/java/org/jboss/dna/search/IndexStrategy.java (from rev 1318, trunk/dna-search/src/main/java/org/jboss/dna/search/IndexingStrategy.java) =================================================================== --- trunk/dna-search/src/main/java/org/jboss/dna/search/IndexStrategy.java (rev 0) +++ trunk/dna-search/src/main/java/org/jboss/dna/search/IndexStrategy.java 2009-11-16 23:24:06 UTC (rev 1319) @@ -0,0 +1,133 @@ +/* + * JBoss DNA (http://www.jboss.org/dna) + * See the COPYRIGHT.txt file distributed with this work for information + * regarding copyright ownership. Some portions may be licensed + * to Red Hat, Inc. under one or more contributor license agreements. + * See the AUTHORS.txt file in the distribution for a full listing of + * individual contributors. + * + * JBoss DNA is free software. Unless otherwise indicated, all code in JBoss DNA + * is licensed to you under the terms of the GNU Lesser General Public License as + * published by the Free Software Foundation; either version 2.1 of + * the License, or (at your option) any later version. + * + * JBoss DNA is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public + * License along with this software; if not, write to the Free + * Software Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA + * 02110-1301 USA, or see the FSF site: http://www.fsf.org. + */ +package org.jboss.dna.search; + +import java.io.IOException; +import java.util.List; +import org.apache.lucene.analysis.Analyzer; +import org.apache.lucene.index.IndexWriter; +import org.apache.lucene.queryParser.ParseException; +import org.jboss.dna.common.text.TextEncoder; +import org.jboss.dna.graph.Location; +import org.jboss.dna.graph.Node; +import org.jboss.dna.graph.property.Path; +import org.jboss.dna.graph.query.QueryResults; +import org.jboss.dna.graph.query.model.QueryCommand; +import org.jboss.dna.graph.query.validate.Schemata; +import org.jboss.dna.graph.request.ChangeRequest; + +/** + * Interface defining the behaviors associated with indexing graph content. + */ +interface IndexStrategy { + + /** + * Get the number of changes that are allowed before optimization is automatically run. + * + * @return a positive number denoting the minimum number of changes between automatic optimization operations, or a + * non-positive number if automatic optimization should never be run + */ + int getChangeCountForAutomaticOptimization(); + + /** + * Get the {@link TextEncoder} that should be used to encode the namespace URIs. + * + * @return the encoder; may not be null + */ + TextEncoder getNamespaceEncoder(); + + /** + * Index the node given the index writers. Note that implementors should simply just use the writers to add documents to the + * index(es), and should never call any of the writer lifecycle methods (e.g., {@link IndexWriter#commit()}, + * {@link IndexWriter#rollback()}, etc.). + * + * @param node the node to be indexed; never null + * @param indexes the set of index readers and writers; never null + * @throws IOException if there is a problem indexing or using the writers + */ + void index( Node node, + IndexContext indexes ) throws IOException; + + /** + * Update the indexes to reflect the supplied changes to the graph content. Note that implementors should simply just use the + * writers to add documents to the index(es), and should never call any of the writer lifecycle methods (e.g., + * {@link IndexWriter#commit()}, {@link IndexWriter#rollback()}, etc.). + * + * @param changes the set of changes to the content + * @param indexes the set of index readers and writers; never null + * @return the (approximate) number of nodes that were affected by the changes + * @throws IOException if there is a problem indexing or using the writers + */ + int apply( Iterable<ChangeRequest> changes, + IndexContext indexes ) throws IOException; + + /** + * Remove from the index(es) all of the information pertaining to the nodes at or below the supplied path. Note that + * implementors should simply just use the writers to add documents to the index(es), and should never call any of the writer + * lifecycle methods (e.g., {@link IndexWriter#commit()}, {@link IndexWriter#rollback()}, etc.). + * + * @param path the path identifying the graph content that is to be removed; never null + * @param indexes the set of index readers and writers; never null + * @return the (approximate) number of nodes that were affected by the changes + * @throws IOException if there is a problem indexing or using the writers + */ + int deleteBelow( Path path, + IndexContext indexes ) throws IOException; + + /** + * Create the analyzer that is used for reading and updating the indexes. + * + * @return the analyzer; may not be null + */ + Analyzer createAnalyzer(); + + /** + * Perform a full-text search given the supplied query. + * + * @param fullTextString the full-text query; never null or blank + * @param maxResults the maximum number of results that are to be returned; always positive + * @param offset the number of initial results to skip, or 0 if the first results are to be returned + * @param indexes the set of index readers and writers; never null + * @param results the list where the results should be accumulated; never null + * @throws IOException if there is a problem indexing or using the writers + * @throws ParseException if there is a problem parsing the query + */ + void search( String fullTextString, + int maxResults, + int offset, + IndexContext indexes, + List<Location> results ) throws IOException, ParseException; + + /** + * Perform a query of the content. The {@link QueryCommand query} is supplied in the form of the Abstract Query Model, with + * the {@link Schemata} that defines the tables and views that are available to the query, and the set of index readers (and + * writers) that should be used. + * + * @param context the context in which the query should be executed; never null + * @param query the query; never null + * @return the results of the query + */ + QueryResults query( SearchContext context, + QueryCommand query ); +} Property changes on: trunk/dna-search/src/main/java/org/jboss/dna/search/IndexStrategy.java ___________________________________________________________________ Name: svn:keywords + Id Revision Name: svn:eol-style + LF Deleted: trunk/dna-search/src/main/java/org/jboss/dna/search/IndexingRules.java =================================================================== --- trunk/dna-search/src/main/java/org/jboss/dna/search/IndexingRules.java 2009-11-16 23:23:02 UTC (rev 1318) +++ trunk/dna-search/src/main/java/org/jboss/dna/search/IndexingRules.java 2009-11-16 23:24:06 UTC (rev 1319) @@ -1,628 +0,0 @@ -/* - * JBoss DNA (http://www.jboss.org/dna) - * See the COPYRIGHT.txt file distributed with this work for information - * regarding copyright ownership. Some portions may be licensed - * to Red Hat, Inc. under one or more contributor license agreements. - * See the AUTHORS.txt file in the distribution for a full listing of - * individual contributors. - * - * JBoss DNA is free software. Unless otherwise indicated, all code in JBoss DNA - * is licensed to you under the terms of the GNU Lesser General Public License as - * published by the Free Software Foundation; either version 2.1 of - * the License, or (at your option) any later version. - * - * JBoss DNA is distributed in the hope that it will be useful, - * but WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU - * Lesser General Public License for more details. - * - * You should have received a copy of the GNU Lesser General Public - * License along with this software; if not, write to the Free - * Software Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA - * 02110-1301 USA, or see the FSF site: http://www.fsf.org. - */ -package org.jboss.dna.search; - -import java.util.Collections; -import java.util.HashMap; -import java.util.Map; -import java.util.concurrent.ConcurrentHashMap; -import net.jcip.annotations.Immutable; -import net.jcip.annotations.NotThreadSafe; -import org.apache.lucene.document.Field; -import org.apache.lucene.document.Field.Index; -import org.apache.lucene.document.Field.Store; -import org.jboss.dna.common.util.CheckArg; -import org.jboss.dna.graph.property.Name; - -/** - * The set of rules that dictate how properties should be indexed. - */ -@Immutable -public class IndexingRules { - - public static final int INDEX = 2 << 0; - public static final int ANALYZE = 2 << 1; - public static final int STORE = 2 << 2; - public static final int STORE_COMPRESSED = 2 << 3; - public static final int ANALYZED_WITHOUT_NORMS = 2 << 4; - public static final int FULL_TEXT = 2 << 5; - public static final int TREAT_AS_DATE = 2 << 6; - - /** - * A single rule that dictates how a single property should be indexed. - * - * @see IndexingRules#getRule(Name) - */ - @Immutable - public static interface Rule { - boolean isIncluded(); - - boolean isSkipped(); - - boolean isAnalyzed(); - - boolean isAnalyzedWithoutNorms(); - - boolean isStored(); - - boolean isStoredCompressed(); - - boolean isFullText(); - - boolean isDate(); - - int getMask(); - - Field.Store getStoreOption(); - - Field.Index getIndexOption(); - } - - public static final Rule SKIP = new SkipRule(); - - @Immutable - protected static class SkipRule implements Rule { - /** - * {@inheritDoc} - * - * @see org.jboss.dna.search.IndexingRules.Rule#getMask() - */ - public int getMask() { - return 0; - } - - /** - * {@inheritDoc} - * - * @see org.jboss.dna.search.IndexingRules.Rule#isAnalyzed() - */ - public boolean isAnalyzed() { - return false; - } - - /** - * {@inheritDoc} - * - * @see org.jboss.dna.search.IndexingRules.Rule#isAnalyzedWithoutNorms() - */ - public boolean isAnalyzedWithoutNorms() { - return false; - } - - /** - * {@inheritDoc} - * - * @see org.jboss.dna.search.IndexingRules.Rule#isFullText() - */ - public boolean isFullText() { - return false; - } - - /** - * {@inheritDoc} - * - * @see org.jboss.dna.search.IndexingRules.Rule#isIncluded() - */ - public boolean isIncluded() { - return false; - } - - /** - * {@inheritDoc} - * - * @see org.jboss.dna.search.IndexingRules.Rule#isSkipped() - */ - public boolean isSkipped() { - return true; - } - - /** - * {@inheritDoc} - * - * @see org.jboss.dna.search.IndexingRules.Rule#isStored() - */ - public boolean isStored() { - return false; - } - - /** - * {@inheritDoc} - * - * @see org.jboss.dna.search.IndexingRules.Rule#isStoredCompressed() - */ - public boolean isStoredCompressed() { - return false; - } - - /** - * {@inheritDoc} - * - * @see org.jboss.dna.search.IndexingRules.Rule#isDate() - */ - public boolean isDate() { - return false; - } - - /** - * {@inheritDoc} - * - * @see org.jboss.dna.search.IndexingRules.Rule#getIndexOption() - */ - public Index getIndexOption() { - return Field.Index.NO; - } - - /** - * {@inheritDoc} - * - * @see org.jboss.dna.search.IndexingRules.Rule#getStoreOption() - */ - public Store getStoreOption() { - return Field.Store.NO; - } - } - - @Immutable - public static final class GeneralRule implements Rule { - private final int value; - private final Field.Store store; - private final Field.Index index; - - protected GeneralRule( int value ) { - this.value = value; - this.index = isAnalyzed() ? Field.Index.ANALYZED : Field.Index.NOT_ANALYZED; - this.store = isStored() ? Field.Store.YES : Field.Store.NO; - } - - /** - * {@inheritDoc} - * - * @see org.jboss.dna.search.IndexingRules.Rule#getMask() - */ - public int getMask() { - return value; - } - - /** - * {@inheritDoc} - * - * @see org.jboss.dna.search.IndexingRules.Rule#isAnalyzed() - */ - public boolean isAnalyzed() { - return (value & ANALYZE) == ANALYZE; - } - - /** - * {@inheritDoc} - * - * @see org.jboss.dna.search.IndexingRules.Rule#isAnalyzedWithoutNorms() - */ - public boolean isAnalyzedWithoutNorms() { - return (value & ANALYZED_WITHOUT_NORMS) == ANALYZED_WITHOUT_NORMS; - } - - /** - * {@inheritDoc} - * - * @see org.jboss.dna.search.IndexingRules.Rule#isFullText() - */ - public boolean isFullText() { - return (value & FULL_TEXT) == FULL_TEXT; - } - - /** - * {@inheritDoc} - * - * @see org.jboss.dna.search.IndexingRules.Rule#isIncluded() - */ - public boolean isIncluded() { - return true; - } - - /** - * {@inheritDoc} - * - * @see org.jboss.dna.search.IndexingRules.Rule#isSkipped() - */ - public boolean isSkipped() { - return false; - } - - /** - * {@inheritDoc} - * - * @see org.jboss.dna.search.IndexingRules.Rule#isStored() - */ - public boolean isStored() { - return (value & STORE) == STORE; - } - - /** - * {@inheritDoc} - * - * @see org.jboss.dna.search.IndexingRules.Rule#isStoredCompressed() - */ - public boolean isStoredCompressed() { - return (value & STORE_COMPRESSED) == STORE_COMPRESSED; - } - - /** - * {@inheritDoc} - * - * @see org.jboss.dna.search.IndexingRules.Rule#isDate() - */ - public boolean isDate() { - return (value & TREAT_AS_DATE) == TREAT_AS_DATE; - } - - protected Rule with( int options ) { - return createRule(value | options); - } - - /** - * {@inheritDoc} - * - * @see org.jboss.dna.search.IndexingRules.Rule#getIndexOption() - */ - public Index getIndexOption() { - return index; - } - - /** - * {@inheritDoc} - * - * @see org.jboss.dna.search.IndexingRules.Rule#getStoreOption() - */ - public Store getStoreOption() { - return store; - } - } - - private static final ConcurrentHashMap<Integer, Rule> CACHE = new ConcurrentHashMap<Integer, Rule>(); - - protected static Rule createRule( int value ) { - if (value <= 0) { - return SKIP; - } - Integer key = new Integer(value); - Rule rule = CACHE.get(key); - if (rule == null) { - Rule newRule = new GeneralRule(value); - rule = CACHE.putIfAbsent(value, newRule); - if (rule == null) rule = newRule; - } - return rule; - } - - private final Map<Name, Rule> rulesByName; - private final Rule defaultRule; - - protected IndexingRules( Map<Name, Rule> rulesByName, - Rule defaultRule ) { - this.rulesByName = rulesByName; - this.defaultRule = defaultRule != null ? defaultRule : SKIP; - assert this.defaultRule != null; - } - - /** - * Get the rule associated with the given property name. - * - * @param name the property name, or null if the default rule is to be returned - * @return the rule; never null - */ - public Rule getRule( Name name ) { - Rule result = rulesByName.get(name); - return result != null ? result : this.defaultRule; - } - - /** - * Return a new builder that can be used to create {@link IndexingRules} objects. - * - * @return a builder; never null - */ - public static Builder createBuilder() { - return new Builder(new HashMap<Name, Rule>()); - } - - /** - * Return a new builder that can be used to create {@link IndexingRules} objects. - * - * @param initialRules the rules that the builder should start with - * @return a builder; never null - * @throws IllegalArgumentException if the initial rules reference is null - */ - public static Builder createBuilder( IndexingRules initialRules ) { - CheckArg.isNotNull(initialRules, "initialRules"); - return new Builder(initialRules.rulesByName).defaultTo(initialRules.defaultRule); - } - - /** - * A builder of immutable {@link IndexingRules} objects. - */ - @NotThreadSafe - public static class Builder { - private final Map<Name, Rule> rulesByName; - private Rule defaultRule; - - Builder( Map<Name, Rule> rulesByName ) { - assert rulesByName != null; - this.rulesByName = rulesByName; - } - - /** - * Set the default rules. - * - * @param rule the default rule to use - * @return this builder for convenience and method chaining; never null - * @throws IllegalArgumentException if the rule mask is negative - */ - public Builder defaultTo( Rule rule ) { - CheckArg.isNotNull(rule, "rule"); - defaultRule = rule; - return this; - } - - /** - * Set the default rules. - * - * @param ruleMask the bitmask of rule to use - * @return this builder for convenience and method chaining; never null - * @throws IllegalArgumentException if the rule mask is negative - */ - public Builder defaultTo( int ruleMask ) { - CheckArg.isNonNegative(ruleMask, "options"); - if (ruleMask == 0) { - defaultRule = SKIP; - } else { - // Make sure the index flag is set ... - ruleMask |= INDEX; - defaultRule = createRule(ruleMask); - } - return this; - } - - /** - * Mark the properties with the supplied names to be skipped from indexing. - * - * @param namesToIndex the names of the properties that are to be skipped - * @return this builder for convenience and method chaining; never null - */ - public Builder skip( Name... namesToIndex ) { - if (namesToIndex != null) { - for (Name name : namesToIndex) { - rulesByName.put(name, SKIP); - } - } - return this; - } - - /** - * Set the properties with the supplied names to use the supplied rules. - * - * @param ruleMask the bitmask of rules to use - * @param namesToIndex the names of the properties that are to be skipped - * @return this builder for convenience and method chaining; never null - * @throws IllegalArgumentException if the rule mask is negative - */ - public Builder set( int ruleMask, - Name... namesToIndex ) { - CheckArg.isNonNegative(ruleMask, "options"); - if (namesToIndex != null) { - if (ruleMask > 0) { - skip(namesToIndex); - } else { - // Make sure the index flag is set ... - ruleMask |= INDEX; - Rule rule = createRule(ruleMask); - for (Name name : namesToIndex) { - rulesByName.put(name, rule); - } - } - } - return this; - } - - /** - * Mark the properties with the supplied names to use the supplied rule mask. This does not remove any other rules for - * these properties. - * - * @param ruleMask the bitmask of rules to add - * @param namesToIndex the names of the properties that are to be skipped - * @return this builder for convenience and method chaining; never null - * @throws IllegalArgumentException if the rule mask is negative - */ - public Builder add( int ruleMask, - Name... namesToIndex ) { - CheckArg.isNonNegative(ruleMask, "options"); - if (namesToIndex != null) { - for (Name name : namesToIndex) { - add(name, ruleMask); - } - } - return this; - } - - /** - * Mark the properties with the supplied names to be indexed. This does not remove any other rules for these properties. - * - * @param namesToIndex the names of the properties that are to be indexed - * @return this builder for convenience and method chaining; never null - */ - public Builder index( Name... namesToIndex ) { - if (namesToIndex != null) { - for (Name name : namesToIndex) { - add(name, INDEX); - } - } - return this; - } - - /** - * Mark the properties with the supplied names to be analyzed (and obviously indexed). This does not remove any other - * rules for these properties. - * - * @param namesToIndex the names of the properties that are to be analyzed - * @return this builder for convenience and method chaining; never null - */ - public Builder analyze( Name... namesToIndex ) { - if (namesToIndex != null) { - for (Name name : namesToIndex) { - add(name, ANALYZE | INDEX); - } - } - return this; - } - - /** - * Mark the properties with the supplied names to be stored (and obviously indexed). This does not remove any other rules - * for these properties. - * - * @param namesToIndex the names of the properties that are to be stored - * @return this builder for convenience and method chaining; never null - */ - public Builder store( Name... namesToIndex ) { - if (namesToIndex != null) { - for (Name name : namesToIndex) { - add(name, STORE | INDEX); - } - } - return this; - } - - /** - * Mark the properties with the supplied names to be included in full-text searches (and obviously indexed). This does not - * remove any other rules for these properties. - * - * @param namesToIndex the names of the properties that are to be included in full-text searches - * @return this builder for convenience and method chaining; never null - */ - public Builder fullText( Name... namesToIndex ) { - if (namesToIndex != null) { - for (Name name : namesToIndex) { - add(name, FULL_TEXT | INDEX); - } - } - return this; - } - - /** - * Mark the properties with the supplied names to be treated as dates (and obviously indexed). This does not remove any - * other rules for these properties. - * - * @param namesToIndex the names of the properties that are to be included in full-text searches - * @return this builder for convenience and method chaining; never null - */ - public Builder treatAsDates( Name... namesToIndex ) { - if (namesToIndex != null) { - for (Name name : namesToIndex) { - add(name, TREAT_AS_DATE | INDEX); - } - } - return this; - } - - /** - * Mark the properties with the supplied names to be indexed, analyzed and stored. This does not remove any other rules - * for these properties. - * - * @param namesToIndex the names of the properties that are to be indexed, analyzed and stored - * @return this builder for convenience and method chaining; never null - */ - public Builder analyzeAndStore( Name... namesToIndex ) { - if (namesToIndex != null) { - for (Name name : namesToIndex) { - add(name, INDEX | ANALYZE | STORE); - } - } - return this; - } - - /** - * Mark the properties with the supplied names to be indexed, analyzed, stored and included in full-text searches. This - * does not remove any other rules for these properties. - * - * @param namesToIndex the names of the properties that are to be indexed, analyzed, stored and included in full-text - * searches - * @return this builder for convenience and method chaining; never null - */ - public Builder analyzeAndStoreAndFullText( Name... namesToIndex ) { - if (namesToIndex != null) { - for (Name name : namesToIndex) { - add(name, INDEX | ANALYZE | STORE | FULL_TEXT); - } - } - return this; - } - - /** - * Mark the properties with the supplied names to be indexed, analyzed and included in full-text searches. This does not - * remove any other rules for these properties. - * - * @param namesToIndex the names of the properties that are to be indexed, analyzed and included in full-text searches - * @return this builder for convenience and method chaining; never null - */ - public Builder analyzeAndFullText( Name... namesToIndex ) { - if (namesToIndex != null) { - for (Name name : namesToIndex) { - add(name, INDEX | ANALYZE | FULL_TEXT); - } - } - return this; - } - - /** - * Mark the properties with the supplied names to be indexed, stored and included in full-text searches. This does not - * remove any other rules for these properties. - * - * @param namesToIndex the names of the properties that are to be indexed, stored and included in full-text searches - * @return this builder for convenience and method chaining; never null - */ - public Builder storeAndFullText( Name... namesToIndex ) { - if (namesToIndex != null) { - for (Name name : namesToIndex) { - add(name, INDEX | STORE | FULL_TEXT); - } - } - return this; - } - - protected void add( Name name, - int option ) { - Rule rule = rulesByName.get(name); - if (rule != null) { - option |= rule.getMask(); - } - rulesByName.put(name, createRule(option)); - } - - /** - * Build the indexing rules. - * - * @return the immutable indexing rules. - */ - public IndexingRules build() { - return new IndexingRules(Collections.unmodifiableMap(new HashMap<Name, Rule>(rulesByName)), defaultRule); - } - } -} Deleted: trunk/dna-search/src/main/java/org/jboss/dna/search/IndexingStrategy.java =================================================================== --- trunk/dna-search/src/main/java/org/jboss/dna/search/IndexingStrategy.java 2009-11-16 23:23:02 UTC (rev 1318) +++ trunk/dna-search/src/main/java/org/jboss/dna/search/IndexingStrategy.java 2009-11-16 23:24:06 UTC (rev 1319) @@ -1,126 +0,0 @@ -/* - * JBoss DNA (http://www.jboss.org/dna) - * See the COPYRIGHT.txt file distributed with this work for information - * regarding copyright ownership. Some portions may be licensed - * to Red Hat, Inc. under one or more contributor license agreements. - * See the AUTHORS.txt file in the distribution for a full listing of - * individual contributors. - * - * JBoss DNA is free software. Unless otherwise indicated, all code in JBoss DNA - * is licensed to you under the terms of the GNU Lesser General Public License as - * published by the Free Software Foundation; either version 2.1 of - * the License, or (at your option) any later version. - * - * JBoss DNA is distributed in the hope that it will be useful, - * but WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU - * Lesser General Public License for more details. - * - * You should have received a copy of the GNU Lesser General Public - * License along with this software; if not, write to the Free - * Software Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA - * 02110-1301 USA, or see the FSF site: http://www.fsf.org. - */ -package org.jboss.dna.search; - -import java.io.IOException; -import java.util.List; -import org.apache.lucene.analysis.Analyzer; -import org.apache.lucene.index.IndexWriter; -import org.apache.lucene.queryParser.ParseException; -import org.jboss.dna.common.text.TextEncoder; -import org.jboss.dna.graph.Location; -import org.jboss.dna.graph.Node; -import org.jboss.dna.graph.property.Path; -import org.jboss.dna.graph.query.QueryResults; -import org.jboss.dna.graph.query.model.QueryCommand; -import org.jboss.dna.graph.query.validate.Schemata; -import org.jboss.dna.graph.request.ChangeRequest; - -/** - * Interface defining the behaviors associated with indexing graph content. - */ -public interface IndexingStrategy { - - int getChangeCountForAutomaticOptimization(); - - TextEncoder getNamespaceEncoder(); - - /** - * Index the node given the index writers. Note that implementors should simply just use the writers to add documents to the - * index(es), and should never call any of the writer lifecycle methods (e.g., {@link IndexWriter#commit()}, - * {@link IndexWriter#rollback()}, etc.). - * - * @param node the node to be indexed; never null - * @param indexes the set of index readers and writers; never null - * @throws IOException if there is a problem indexing or using the writers - */ - void index( Node node, - IndexContext indexes ) throws IOException; - - /** - * Update the indexes to reflect the supplied changes to the graph content. Note that implementors should simply just use the - * writers to add documents to the index(es), and should never call any of the writer lifecycle methods (e.g., - * {@link IndexWriter#commit()}, {@link IndexWriter#rollback()}, etc.). - * - * @param changes the set of changes to the content - * @param indexes the set of index readers and writers; never null - * @return the (approximate) number of nodes that were affected by the changes - * @throws IOException if there is a problem indexing or using the writers - */ - int apply( Iterable<ChangeRequest> changes, - IndexContext indexes ) throws IOException; - - /** - * Remove from the index(es) all of the information pertaining to the nodes at or below the supplied path. Note that - * implementors should simply just use the writers to add documents to the index(es), and should never call any of the writer - * lifecycle methods (e.g., {@link IndexWriter#commit()}, {@link IndexWriter#rollback()}, etc.). - * - * @param path the path identifying the graph content that is to be removed; never null - * @param indexes the set of index readers and writers; never null - * @return the (approximate) number of nodes that were affected by the changes - * @throws IOException if there is a problem indexing or using the writers - */ - int deleteBelow( Path path, - IndexContext indexes ) throws IOException; - - /** - * Create the analyzer that is used for reading and updating the indexes. - * - * @return the analyzer; may not be null - */ - Analyzer createAnalyzer(); - - /** - * Perform a full-text search given the supplied query. - * - * @param fullTextString the full-text query; never null or blank - * @param maxResults the maximum number of results that are to be returned; always positive - * @param offset the number of initial results to skip, or 0 if the first results are to be returned - * @param indexes the set of index readers and writers; never null - * @param results the list where the results should be accumulated; never null - * @throws IOException if there is a problem indexing or using the writers - * @throws ParseException if there is a problem parsing the query - */ - void performQuery( String fullTextString, - int maxResults, - int offset, - IndexContext indexes, - List<Location> results ) throws IOException, ParseException; - - /** - * Perform a query of the content. The {@link QueryCommand query} is supplied in the form of the Abstract Query Model, with - * the {@link Schemata} that defines the tables and views that are available to the query, and the set of index readers (and - * writers) that should be used. - * - * @param query the query; never null - * @param schemata the definition of the tables used in the query; never null - * @param indexes the set of index readers and writers; never null - * @return the results of the query - * @throws IOException if there is a problem indexing or using the writers - * @throws ParseException if there is a problem parsing the query - */ - QueryResults performQuery( QueryCommand query, - Schemata schemata, - IndexContext indexes ) throws IOException, ParseException; -} Added: trunk/dna-search/src/main/java/org/jboss/dna/search/KitchenSinkIndexStrategy.java =================================================================== --- trunk/dna-search/src/main/java/org/jboss/dna/search/KitchenSinkIndexStrategy.java (rev 0) +++ trunk/dna-search/src/main/java/org/jboss/dna/search/KitchenSinkIndexStrategy.java 2009-11-16 23:24:06 UTC (rev 1319) @@ -0,0 +1,115 @@ +/* + * JBoss DNA (http://www.jboss.org/dna) + * See the COPYRIGHT.txt file distributed with this work for information + * regarding copyright ownership. Some portions may be licensed + * to Red Hat, Inc. under one or more contributor license agreements. + * See the AUTHORS.txt file in the distribution for a full listing of + * individual contributors. + * + * JBoss DNA is free software. Unless otherwise indicated, all code in JBoss DNA + * is licensed to you under the terms of the GNU Lesser General Public License as + * published by the Free Software Foundation; either version 2.1 of + * the License, or (at your option) any later version. + * + * JBoss DNA is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public + * License along with this software; if not, write to the Free + * Software Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA + * 02110-1301 USA, or see the FSF site: http://www.fsf.org. + */ +package org.jboss.dna.search; + +import java.util.List; +import net.jcip.annotations.ThreadSafe; +import org.jboss.dna.graph.DnaLexicon; +import org.jboss.dna.graph.JcrLexicon; +import org.jboss.dna.graph.query.QueryResults.Columns; +import org.jboss.dna.graph.query.plan.PlanNode; +import org.jboss.dna.graph.query.process.AbstractAccessComponent; +import org.jboss.dna.graph.query.process.ProcessingComponent; +import org.jboss.dna.graph.query.process.SelectComponent.Analyzer; +import org.jboss.dna.graph.request.ChangeRequest; + +/** + * An {@link IndexStrategy} implementation that stores all content within a set of two indexes: one for the node content and a + * second one for paths and UUIDs. + */ +@ThreadSafe +class KitchenSinkIndexStrategy extends DualIndexStrategy { + + /** + * The default set of {@link IndexRules} used by {@link KitchenSinkIndexStrategy} instances when no rules are provided. + */ + public static final IndexRules DEFAULT_RULES; + + static { + IndexRules.Builder builder = IndexRules.createBuilder(); + // Configure the default behavior ... + builder.defaultTo(IndexRules.INDEX | IndexRules.ANALYZE); + // Configure the UUID properties to be just indexed (not stored, not analyzed, not included in full-text) ... + builder.index(JcrLexicon.UUID, DnaLexicon.UUID); + // Configure the properties that we'll treat as dates ... + builder.treatAsDates(JcrLexicon.CREATED, JcrLexicon.LAST_MODIFIED); + DEFAULT_RULES = builder.build(); + } + + /** + * Create a new indexing strategy instance uses the {@link #DEFAULT_RULES default indexing rules}. + */ + public KitchenSinkIndexStrategy() { + this(null); + } + + /** + * Create a new indexing strategy instance. + * + * @param rules the indexing rules that govern how properties are to be index, or null if the {@link #DEFAULT_RULES default + * rules} are to be used + */ + public KitchenSinkIndexStrategy( IndexRules rules ) { + super(rules != null ? rules : DEFAULT_RULES); + } + + /** + * {@inheritDoc} + * + * @see org.jboss.dna.search.DualIndexStrategy#createAccessComponent(org.jboss.dna.search.SearchContext, + * org.jboss.dna.graph.query.plan.PlanNode, org.jboss.dna.graph.query.QueryResults.Columns, + * org.jboss.dna.graph.query.process.SelectComponent.Analyzer) + */ + @Override + protected ProcessingComponent createAccessComponent( final SearchContext context, + PlanNode accessNode, + Columns resultColumns, + Analyzer analyzer ) { + // Create a processing component for this access query ... + return new AbstractAccessComponent(context, resultColumns, accessNode) { + /** + * {@inheritDoc} + * + * @see org.jboss.dna.graph.query.process.ProcessingComponent#execute() + */ + @Override + public List<Object[]> execute() { + return null; + } + }; + } + + /** + * {@inheritDoc} + * + * @see org.jboss.dna.search.IndexStrategy#apply(Iterable, IndexContext) + */ + public int apply( Iterable<ChangeRequest> changes, + IndexContext indexes ) /*throws IOException*/{ + for (ChangeRequest change : changes) { + if (change != null) continue; + } + return 0; + } +} Property changes on: trunk/dna-search/src/main/java/org/jboss/dna/search/KitchenSinkIndexStrategy.java ___________________________________________________________________ Name: svn:keywords + Id Revision Name: svn:eol-style + LF Deleted: trunk/dna-search/src/main/java/org/jboss/dna/search/LuceneQueryComponent.java =================================================================== --- trunk/dna-search/src/main/java/org/jboss/dna/search/LuceneQueryComponent.java 2009-11-16 23:23:02 UTC (rev 1318) +++ trunk/dna-search/src/main/java/org/jboss/dna/search/LuceneQueryComponent.java 2009-11-16 23:24:06 UTC (rev 1319) @@ -1,57 +0,0 @@ -/* - * JBoss DNA (http://www.jboss.org/dna) - * See the COPYRIGHT.txt file distributed with this work for information - * regarding copyright ownership. Some portions may be licensed - * to Red Hat, Inc. under one or more contributor license agreements. - * See the AUTHORS.txt file in the distribution for a full listing of - * individual contributors. - * - * JBoss DNA is free software. Unless otherwise indicated, all code in JBoss DNA - * is licensed to you under the terms of the GNU Lesser General Public License as - * published by the Free Software Foundation; either version 2.1 of - * the License, or (at your option) any later version. - * - * JBoss DNA is distributed in the hope that it will be useful, - * but WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU - * Lesser General Public License for more details. - * - * You should have received a copy of the GNU Lesser General Public - * License along with this software; if not, write to the Free - * Software Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA - * 02110-1301 USA, or see the FSF site: http://www.fsf.org. - */ -package org.jboss.dna.search; - -import java.util.List; -import org.jboss.dna.graph.query.QueryContext; -import org.jboss.dna.graph.query.QueryResults.Columns; -import org.jboss.dna.graph.query.plan.PlanNode; -import org.jboss.dna.graph.query.process.ProcessingComponent; - -/** - * A {@link ProcessingComponent} implementation that is used by the {@link LuceneQueryEngine.LuceneProcessor} to perform atomic - * queries against the Lucene indexes. - */ -class LuceneQueryComponent extends ProcessingComponent { - - private final PlanNode accessNode; - - LuceneQueryComponent( QueryContext context, - Columns columns, - PlanNode accessNode ) { - super(context, columns); - this.accessNode = accessNode; - } - - /** - * {@inheritDoc} - * - * @see org.jboss.dna.graph.query.process.ProcessingComponent#execute() - */ - @Override - public List<Object[]> execute() { - if (accessNode != null) return null; - return null; - } -} Deleted: trunk/dna-search/src/main/java/org/jboss/dna/search/LuceneQueryEngine.java =================================================================== --- trunk/dna-search/src/main/java/org/jboss/dna/search/LuceneQueryEngine.java 2009-11-16 23:23:02 UTC (rev 1318) +++ trunk/dna-search/src/main/java/org/jboss/dna/search/LuceneQueryEngine.java 2009-11-16 23:24:06 UTC (rev 1319) @@ -1,114 +0,0 @@ -/* - * JBoss DNA (http://www.jboss.org/dna) - * See the COPYRIGHT.txt file distributed with this work for information - * regarding copyright ownership. Some portions may be licensed - * to Red Hat, Inc. under one or more contributor license agreements. - * See the AUTHORS.txt file in the distribution for a full listing of - * individual contributors. - * - * JBoss DNA is free software. Unless otherwise indicated, all code in JBoss DNA - * is licensed to you under the terms of the GNU Lesser General Public License as - * published by the Free Software Foundation; either version 2.1 of - * the License, or (at your option) any later version. - * - * JBoss DNA is distributed in the hope that it will be useful, - * but WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU - * Lesser General Public License for more details. - * - * You should have received a copy of the GNU Lesser General Public - * License along with this software; if not, write to the Free - * Software Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA - * 02110-1301 USA, or see the FSF site: http://www.fsf.org. - */ -package org.jboss.dna.search; - -import java.io.IOException; -import java.util.LinkedList; -import org.apache.lucene.queryParser.ParseException; -import org.jboss.dna.graph.query.QueryContext; -import org.jboss.dna.graph.query.QueryEngine; -import org.jboss.dna.graph.query.QueryResults; -import org.jboss.dna.graph.query.QueryResults.Columns; -import org.jboss.dna.graph.query.model.QueryCommand; -import org.jboss.dna.graph.query.optimize.Optimizer; -import org.jboss.dna.graph.query.optimize.OptimizerRule; -import org.jboss.dna.graph.query.optimize.RuleBasedOptimizer; -import org.jboss.dna.graph.query.plan.CanonicalPlanner; -import org.jboss.dna.graph.query.plan.PlanHints; -import org.jboss.dna.graph.query.plan.PlanNode; -import org.jboss.dna.graph.query.process.ProcessingComponent; -import org.jboss.dna.graph.query.process.QueryProcessor; -import org.jboss.dna.graph.query.process.SelectComponent.Analyzer; -import org.jboss.dna.graph.query.validate.Schemata; - -/** - * - */ -class LuceneQueryEngine { - - private QueryEngine engine; - - public LuceneQueryEngine() { - engine = new QueryEngine(new CanonicalPlanner(), new LuceneOptimizer(), new LuceneProcessor()); - } - - /** - * Execute the supplied query by planning, optimizing, and then processing it. - * - * @param query the query that is to be executed - * @param schemata the schemata that defines the tables used in the query - * @param indexes the indexes that should be used to execute the query; never null - * @return the query results; never null - * @throws IllegalArgumentException if the context or query references are null - * @throws IOException if there is a problem indexing or using the writers - * @throws ParseException if there is a problem parsing the query - */ - public QueryResults execute( QueryCommand query, - Schemata schemata, - IndexContext indexes ) throws IOException, ParseException { - return engine.execute(indexes.context(), query, schemata, new PlanHints()); - } - - /** - * An {@link Optimizer} implementation that specializes the {@link RuleBasedOptimizer} by using custom rules. - */ - protected static class LuceneOptimizer extends RuleBasedOptimizer { - - /** - * {@inheritDoc} - * - * @see org.jboss.dna.graph.query.optimize.RuleBasedOptimizer#populateRuleStack(java.util.LinkedList, - * org.jboss.dna.graph.query.plan.PlanHints) - */ - @Override - protected void populateRuleStack( LinkedList<OptimizerRule> ruleStack, - PlanHints hints ) { - super.populateRuleStack(ruleStack, hints); - // Add any custom rules here, either at the front of the stack or at the end - } - } - - /** - * A query processor that operates against Lucene indexes. All functionality is inherited from the {@link QueryProcessor}, - * except for the creation of the {@link ProcessingComponent} that does the low-level atomic queries (against the Lucene - * indexes). - */ - protected static class LuceneProcessor extends QueryProcessor { - - /** - * {@inheritDoc} - * - * @see org.jboss.dna.graph.query.process.QueryProcessor#createAccessComponent(org.jboss.dna.graph.query.QueryContext, - * org.jboss.dna.graph.query.plan.PlanNode, org.jboss.dna.graph.query.QueryResults.Columns, - * org.jboss.dna.graph.query.process.SelectComponent.Analyzer) - */ - @Override - protected ProcessingComponent createAccessComponent( QueryContext context, - PlanNode accessNode, - Columns resultColumns, - Analyzer analyzer ) { - return new LuceneQueryComponent(context, resultColumns, accessNode); - } - } -} Added: trunk/dna-search/src/main/java/org/jboss/dna/search/SearchContext.java =================================================================== --- trunk/dna-search/src/main/java/org/jboss/dna/search/SearchContext.java (rev 0) +++ trunk/dna-search/src/main/java/org/jboss/dna/search/SearchContext.java 2009-11-16 23:24:06 UTC (rev 1319) @@ -0,0 +1,161 @@ +/* + * JBoss DNA (http://www.jboss.org/dna) + * See the COPYRIGHT.txt file distributed with this work for information + * regarding copyright ownership. Some portions may be licensed + * to Red Hat, Inc. under one or more contributor license agreements. + * See the AUTHORS.txt file in the distribution for a full listing of + * individual contributors. + * + * JBoss DNA is free software. Unless otherwise indicated, all code in JBoss DNA + * is licensed to you under the terms of the GNU Lesser General Public License as + * published by the Free Software Foundation; either version 2.1 of + * the License, or (at your option) any later version. + * + * JBoss DNA is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public + * License along with this software; if not, write to the Free + * Software Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA + * 02110-1301 USA, or see the FSF site: http://www.fsf.org. + */ +package org.jboss.dna.search; + +import java.util.Map; +import org.jboss.dna.common.collection.Problems; +import org.jboss.dna.common.util.CheckArg; +import org.jboss.dna.graph.ExecutionContext; +import org.jboss.dna.graph.query.QueryContext; +import org.jboss.dna.graph.query.plan.PlanHints; +import org.jboss.dna.graph.query.validate.Schemata; + +/** + * + */ +class SearchContext extends QueryContext { + + private final IndexContext indexes; + + /** + * Create a new context for searching and querying. + * + * @param indexes the indexes that should be used + * @param schemata the definition of the tables available to this query + */ + public SearchContext( IndexContext indexes, + Schemata schemata ) { + super(indexes.context(), schemata); + this.indexes = indexes; + assert this.indexes != null; + } + + /** + * Create a new context for searching and querying. + * + * @param queryContext + * @param indexes + */ + public SearchContext( QueryContext queryContext, + IndexContext indexes ) { + super(queryContext.getExecutionContext(), queryContext.getSchemata(), queryContext.getHints(), + queryContext.getProblems(), queryContext.getVariables()); + this.indexes = indexes; + assert this.indexes != null; + } + + /** + * Create a new context for searching and querying. + * + * @param context the execution context + * @param schemata the schemata + * @param hints the hints, or null if there are no hints + * @param problems the problems container, or null if a new problems container should be created + * @param variables the mapping of variables and values, or null if there are no such variables + * @throws IllegalArgumentException if the context or schmata are null + */ + public SearchContext( IndexContext context, + Schemata schemata, + PlanHints hints, + Problems problems, + Map<String, Object> variables ) { + super(context.context(), schemata, hints, problems, variables); + this.indexes = context; + assert this.indexes != null; + } + + /** + * Get the {@link IndexContext} for this query context. + * + * @return the index context; never null + */ + public IndexContext getIndexes() { + return indexes; + } + + /** + * Obtain a copy of this context, except that the copy uses the supplied index context. + * + * @param context the index context that should be used in the new query context + * @return the new context; never null + * @throws IllegalArgumentException if the index context reference is null + */ + public SearchContext with( IndexContext context ) { + CheckArg.isNotNull(context, "context"); + return new SearchContext(context, getSchemata(), getHints(), getProblems(), getVariables()); + } + + /** + * {@inheritDoc} + * + * @see org.jboss.dna.graph.query.QueryContext#with(org.jboss.dna.graph.ExecutionContext) + */ + @Override + public SearchContext with( ExecutionContext context ) { + CheckArg.isNotNull(context, "context"); + return new SearchContext(indexes.with(context), getSchemata(), getHints(), getProblems(), getVariables()); + } + + /** + * {@inheritDoc} + * + * @see org.jboss.dna.graph.query.QueryContext#with(org.jboss.dna.graph.query.validate.Schemata) + */ + @Override + public SearchContext with( Schemata schemata ) { + CheckArg.isNotNull(schemata, "schemata"); + return new SearchContext(indexes, schemata, getHints(), getProblems(), getVariables()); + } + + /** + * {@inheritDoc} + * + * @see org.jboss.dna.graph.query.QueryContext#with(org.jboss.dna.graph.query.plan.PlanHints) + */ + @Override + public SearchContext with( PlanHints hints ) { + CheckArg.isNotNull(hints, "hints"); + return new SearchContext(indexes, getSchemata(), hints, getProblems(), getVariables()); + } + + /** + * {@inheritDoc} + * + * @see org.jboss.dna.graph.query.QueryContext#with(org.jboss.dna.common.collection.Problems) + */ + @Override + public SearchContext with( Problems problems ) { + return new SearchContext(indexes, getSchemata(), getHints(), problems, getVariables()); + } + + /** + * {@inheritDoc} + * + * @see org.jboss.dna.graph.query.QueryContext#with(java.util.Map) + */ + @Override + public SearchContext with( Map<String, Object> variables ) { + return new SearchContext(indexes, getSchemata(), getHints(), getProblems(), variables); + } +} Property changes on: trunk/dna-search/src/main/java/org/jboss/dna/search/SearchContext.java ___________________________________________________________________ Name: svn:keywords + Id Revision Name: svn:eol-style + LF Modified: trunk/dna-search/src/main/java/org/jboss/dna/search/SearchEngine.java =================================================================== --- trunk/dna-search/src/main/java/org/jboss/dna/search/SearchEngine.java 2009-11-16 23:23:02 UTC (rev 1318) +++ trunk/dna-search/src/main/java/org/jboss/dna/search/SearchEngine.java 2009-11-16 23:24:06 UTC (rev 1319) @@ -61,7 +61,7 @@ private final String sourceName; private final RepositoryConnectionFactory connectionFactory; private final DirectoryConfiguration directoryFactory; - private final IndexingStrategy indexingStrategy; + private final IndexStrategy indexStrategy; private final PathFactory pathFactory; @GuardedBy( "workspaceEngineLock" ) private final Map<String, WorkspaceSearchEngine> workspaceEnginesByName; @@ -76,15 +76,12 @@ * @param sourceName the name of the {@link RepositorySource} * @param connectionFactory the connection factory * @param directoryFactory the factory for Lucene {@link Directory directories} - * @param indexingStrategy the indexing strategy that governs how properties are to be indexed; or null if the default - * strategy should be used * @throws IllegalArgumentException if any of the parameters (other than indexing strategy) are null */ public SearchEngine( ExecutionContext context, String sourceName, RepositoryConnectionFactory connectionFactory, - DirectoryConfiguration directoryFactory, - IndexingStrategy indexingStrategy ) { + DirectoryConfiguration directoryFactory ) { CheckArg.isNotNull(context, "context"); CheckArg.isNotNull(sourceName, "sourceName"); CheckArg.isNotNull(connectionFactory, "connectionFactory"); @@ -95,7 +92,7 @@ this.context = context; this.pathFactory = context.getValueFactories().getPathFactory(); this.workspaceEnginesByName = new HashMap<String, WorkspaceSearchEngine>(); - this.indexingStrategy = indexingStrategy != null ? indexingStrategy : new StoreLittleIndexingStrategy(); + this.indexStrategy = new KitchenSinkIndexStrategy(); } /** @@ -153,7 +150,7 @@ engine = workspaceEnginesByName.get(workspaceName); if (engine == null) { // Create the engine and register it ... - engine = new WorkspaceSearchEngine(context, directoryFactory, indexingStrategy, sourceName, workspaceName, + engine = new WorkspaceSearchEngine(context, directoryFactory, indexStrategy, sourceName, workspaceName, connectionFactory); workspaceEnginesByName.put(workspaceName, engine); } Deleted: trunk/dna-search/src/main/java/org/jboss/dna/search/StoreLittleIndexingStrategy.java =================================================================== --- trunk/dna-search/src/main/java/org/jboss/dna/search/StoreLittleIndexingStrategy.java 2009-11-16 23:23:02 UTC (rev 1318) +++ trunk/dna-search/src/main/java/org/jboss/dna/search/StoreLittleIndexingStrategy.java 2009-11-16 23:24:06 UTC (rev 1319) @@ -1,372 +0,0 @@ -/* - * JBoss DNA (http://www.jboss.org/dna) - * See the COPYRIGHT.txt file distributed with this work for information - * regarding copyright ownership. Some portions may be licensed - * to Red Hat, Inc. under one or more contributor license agreements. - * See the AUTHORS.txt file in the distribution for a full listing of - * individual contributors. - * - * JBoss DNA is free software. Unless otherwise indicated, all code in JBoss DNA - * is licensed to you under the terms of the GNU Lesser General Public License as - * published by the Free Software Foundation; either version 2.1 of - * the License, or (at your option) any later version. - * - * JBoss DNA is distributed in the hope that it will be useful, - * but WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU - * Lesser General Public License for more details. - * - * You should have received a copy of the GNU Lesser General Public - * License along with this software; if not, write to the Free - * Software Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA - * 02110-1301 USA, or see the FSF site: http://www.fsf.org. - */ -package org.jboss.dna.search; - -import java.io.FileNotFoundException; -import java.io.IOException; -import java.text.DateFormat; -import java.text.SimpleDateFormat; -import java.util.List; -import java.util.UUID; -import net.jcip.annotations.ThreadSafe; -import org.apache.lucene.analysis.Analyzer; -import org.apache.lucene.analysis.standard.StandardAnalyzer; -import org.apache.lucene.document.Document; -import org.apache.lucene.document.Field; -import org.apache.lucene.document.FieldSelector; -import org.apache.lucene.document.FieldSelectorResult; -import org.apache.lucene.index.IndexReader; -import org.apache.lucene.index.Term; -import org.apache.lucene.queryParser.ParseException; -import org.apache.lucene.queryParser.QueryParser; -import org.apache.lucene.search.IndexSearcher; -import org.apache.lucene.search.PrefixQuery; -import org.apache.lucene.search.Query; -import org.apache.lucene.search.ScoreDoc; -import org.apache.lucene.search.TermQuery; -import org.apache.lucene.search.TopDocs; -import org.apache.lucene.util.Version; -import org.jboss.dna.common.text.NoOpEncoder; -import org.jboss.dna.common.text.TextEncoder; -import org.jboss.dna.common.util.Logger; -import org.jboss.dna.graph.DnaLexicon; -import org.jboss.dna.graph.JcrLexicon; -import org.jboss.dna.graph.Location; -import org.jboss.dna.graph.Node; -import org.jboss.dna.graph.property.Binary; -import org.jboss.dna.graph.property.DateTime; -import org.jboss.dna.graph.property.DateTimeFactory; -import org.jboss.dna.graph.property.Name; -import org.jboss.dna.graph.property.Path; -import org.jboss.dna.graph.property.Property; -import org.jboss.dna.graph.property.ValueFactory; -import org.jboss.dna.graph.query.QueryResults; -import org.jboss.dna.graph.query.model.QueryCommand; -import org.jboss.dna.graph.query.validate.Schemata; -import org.jboss.dna.graph.request.ChangeRequest; -import org.jboss.dna.search.IndexingRules.Rule; - -/** - * A simple {@link IndexingStrategy} implementation that relies upon very few fields to be stored in the indexes. - */ -@ThreadSafe -class StoreLittleIndexingStrategy implements IndexingStrategy { - - static class PathIndex { - public static final String PATH = "path"; - public static final String UUID = "uuid"; - } - - static class ContentIndex { - public static final String UUID = PathIndex.UUID; - public static final String FULL_TEXT = "fts"; - } - - public static final int SIZE_OF_DELETE_BATCHES = 100; - - private ThreadLocal<DateFormat> dateFormatter = new ThreadLocal<DateFormat>() { - @Override - protected DateFormat initialValue() { - return new SimpleDateFormat("yyyyMMdd'T'HH:mm:ss"); - } - }; - - private static final FieldSelector UUID_FIELD_SELECTOR = new FieldSelector() { - private static final long serialVersionUID = 1L; - - public FieldSelectorResult accept( String fieldName ) { - return PathIndex.UUID.equals(fieldName) ? FieldSelectorResult.LOAD_AND_BREAK : FieldSelectorResult.NO_LOAD; - } - }; - - /** - * The default set of {@link IndexingRules} used by {@link StoreLittleIndexingStrategy} instances when no rules are provided. - */ - public static final IndexingRules DEFAULT_RULES; - - static { - IndexingRules.Builder builder = IndexingRules.createBuilder(); - // Configure the default behavior ... - builder.defaultTo(IndexingRules.INDEX | IndexingRules.ANALYZE); - // Configure the UUID properties to be just indexed (not stored, not analyzed, not included in full-text) ... - builder.index(JcrLexicon.UUID, DnaLexicon.UUID); - // Configure the properties that we'll treat as dates ... - builder.treatAsDates(JcrLexicon.CREATED, JcrLexicon.LAST_MODIFIED); - DEFAULT_RULES = builder.build(); - } - - private final IndexingRules rules; - private final Logger logger; - private final LuceneQueryEngine queryEngine; - - /** - * Create a new indexing strategy instance that does not support queries. - */ - public StoreLittleIndexingStrategy() { - this(null); - } - - /** - * Create a new indexing strategy instance. - * - * @param rules the indexing rules that govern how properties are to be index, or null if the {@link #DEFAULT_RULES default - * rules} are to be used - */ - public StoreLittleIndexingStrategy( IndexingRules rules ) { - this.rules = rules != null ? rules : DEFAULT_RULES; - this.logger = Logger.getLogger(getClass()); - this.queryEngine = new LuceneQueryEngine(); - } - - /** - * {@inheritDoc} - * - * @see org.jboss.dna.search.IndexingStrategy#getNamespaceEncoder() - */ - public TextEncoder getNamespaceEncoder() { - return new NoOpEncoder(); - } - - /** - * {@inheritDoc} - * - * @see org.jboss.dna.search.IndexingStrategy#getChangeCountForAutomaticOptimization() - */ - public int getChangeCountForAutomaticOptimization() { - return 0; - } - - /** - * {@inheritDoc} - * - * @see org.jboss.dna.search.IndexingStrategy#createAnalyzer() - */ - public Analyzer createAnalyzer() { - return new StandardAnalyzer(Version.LUCENE_CURRENT); - } - - /** - * {@inheritDoc} - * <p> - * Because this strategy uses multiple indexes, and since there's no correlation between the documents in those indexes, we - * need to perform the delete in multiple steps. First, we need to perform a query to find out which nodes exist below a - * certain path. Then, we need to delete those nodes from the paths index. Finally, we need to delete the corresponding - * documents in the content index that represent those same nodes. - * </p> - * <p> - * Since we don't know how many documents there will be, we perform these steps in batches, where each batch limits the number - * of results to a maximum number. We repeat batches as long as we find more results. This approach has the advantage that - * we'll never bring in a large number of results, and it allows us to delete the documents from the content node using a - * query. - * </p> - * - * @see org.jboss.dna.search.IndexingStrategy#deleteBelow(Path, IndexContext) - */ - public int deleteBelow( Path path, - IndexContext indexes ) throws IOException { - // Perform a query using the reader to find those nodes at/below the path ... - try { - IndexReader pathReader = indexes.getPathsReader(); - IndexSearcher pathSearcher = new IndexSearcher(pathReader); - String pathStr = indexes.stringFactory().create(path) + "/"; - PrefixQuery query = new PrefixQuery(new Term(PathIndex.PATH, pathStr)); - int numberDeleted = 0; - while (true) { - // Execute the query and get the results ... - TopDocs results = pathSearcher.search(query, SIZE_OF_DELETE_BATCHES); - int numResultsInBatch = results.scoreDocs.length; - // Walk the results, delete the doc, and add to the query that we'll use against the content index ... - IndexReader contentReader = indexes.getContentReader(); - for (ScoreDoc result : results.scoreDocs) { - int docId = result.doc; - // Find the UUID of the node ... - Document doc = pathReader.document(docId, UUID_FIELD_SELECTOR); - String uuid = doc.get(PathIndex.UUID); - // Delete the document from the paths index ... - pathReader.deleteDocument(docId); - // Delete the corresponding document from the content index ... - contentReader.deleteDocuments(new Term(ContentIndex.UUID, uuid)); - } - numberDeleted += numResultsInBatch; - if (numResultsInBatch < SIZE_OF_DELETE_BATCHES) break; - } - indexes.commit(); - return numberDeleted; - } catch (FileNotFoundException e) { - // There are no index files yet, so nothing to delete ... - return 0; - } - } - - /** - * {@inheritDoc} - * - * @see org.jboss.dna.search.IndexingStrategy#index(Node, IndexContext) - */ - public void index( Node node, - IndexContext indexes ) throws IOException { - ValueFactory<String> strings = indexes.stringFactory(); - Location location = node.getLocation(); - UUID uuid = location.getUuid(); - if (uuid == null) uuid = UUID.randomUUID(); - Path path = location.getPath(); - String pathStr = path.isRoot() ? "/" : strings.create(location.getPath()) + "/"; - String uuidStr = uuid.toString(); - - if (logger.isTraceEnabled()) { - logger.trace("indexing {0}", pathStr); - } - - // Create a separate document for the path, which makes it easier to handle moves since the path can - // be changed without changing any other content fields ... - Document doc = new Document(); - doc.add(new Field(PathIndex.PATH, pathStr, Field.Store.YES, Field.Index.NOT_ANALYZED)); - doc.add(new Field(PathIndex.UUID, uuidStr, Field.Store.YES, Field.Index.NOT_ANALYZED)); - indexes.getPathsWriter().addDocument(doc); - - // Create the document for the content (properties) ... - doc = new Document(); - doc.add(new Field(ContentIndex.UUID, uuidStr, Field.Store.YES, Field.Index.NOT_ANALYZED)); - String stringValue = null; - StringBuilder fullTextSearchValue = null; - for (Property property : node.getProperties()) { - Name name = property.getName(); - Rule rule = rules.getRule(name); - if (rule.isSkipped()) continue; - String nameString = strings.create(name); - if (rule.isDate()) { - DateTimeFactory dateFactory = indexes.dateFactory(); - for (Object value : property) { - if (value == null) continue; - DateTime dateValue = dateFactory.create(value); - stringValue = dateFormatter.get().format(dateValue.toDate()); - // Add a separate field for each property value ... - doc.add(new Field(nameString, stringValue, rule.getStoreOption(), rule.getIndexOption())); - // Dates are not added to the full-text search field (since this wouldn't make sense) - } - continue; - } - for (Object value : property) { - if (value == null) continue; - if (value instanceof Binary) { - // don't include binary values as individual fields but do include them in the full-text search ... - // TODO : add to full-text search ... - continue; - } - stringValue = strings.create(value); - // Add a separate field for each property value ... - doc.add(new Field(nameString, stringValue, rule.getStoreOption(), rule.getIndexOption())); - // And add to the full-text field ... - if (rule.isFullText()) { - if (fullTextSearchValue == null) { - fullTextSearchValue = new StringBuilder(); - } else { - fullTextSearchValue.append(' '); - } - fullTextSearchValue.append(stringValue); - } - } - } - // Add the full-text-search field ... - if (fullTextSearchValue != null) { - doc.add(new Field(ContentIndex.FULL_TEXT, fullTextSearchValue.toString(), Field.Store.NO, Field.Index.ANALYZED)); - } - indexes.getContentWriter().addDocument(doc); - } - - /** - * {@inheritDoc} - * - * @see org.jboss.dna.search.IndexingStrategy#performQuery(String, int, int, IndexContext, List) - */ - public void performQuery( String fullTextString, - int maxResults, - int offset, - IndexContext indexes, - List<Location> results ) throws IOException, ParseException { - assert fullTextString != null; - assert fullTextString.length() > 0; - assert offset >= 0; - assert maxResults > 0; - assert indexes != null; - assert results != null; - - // Parse the full-text search and search against the 'fts' field ... - QueryParser parser = new QueryParser(ContentIndex.FULL_TEXT, createAnalyzer()); - Query query = parser.parse(fullTextString); - TopDocs docs = indexes.getContentSearcher().search(query, maxResults + offset); - - // Collect the results ... - IndexReader contentReader = indexes.getContentReader(); - IndexReader pathReader = indexes.getPathsReader(); - IndexSearcher pathSearcher = indexes.getPathsSearcher(); - ScoreDoc[] scoreDocs = docs.scoreDocs; - int numberOfResults = scoreDocs.length; - if (numberOfResults > offset) { - // There are enough results to satisfy the offset ... - for (int i = offset, num = scoreDocs.length; i != num; ++i) { - ScoreDoc result = scoreDocs[i]; - int docId = result.doc; - // Find the UUID of the node (this UUID might be artificial, so we have to find the path) ... - Document doc = contentReader.document(docId, UUID_FIELD_SELECTOR); - String uuid = doc.get(ContentIndex.UUID); - // Find the path for this node (is there a better way to do this than one search per UUID?) ... - TopDocs pathDocs = pathSearcher.search(new TermQuery(new Term(PathIndex.UUID, uuid)), 1); - if (pathDocs.scoreDocs.length < 1) { - // No path record found ... - continue; - } - Document pathDoc = pathReader.document(pathDocs.scoreDocs[0].doc); - Path path = indexes.pathFactory().create(pathDoc.get(PathIndex.PATH)); - // Now add the location ... - results.add(Location.create(path, UUID.fromString(uuid))); - } - } - } - - /** - * {@inheritDoc} - * - * @see org.jboss.dna.search.IndexingStrategy#performQuery(org.jboss.dna.graph.query.model.QueryCommand, - * org.jboss.dna.graph.query.validate.Schemata, org.jboss.dna.search.IndexContext) - */ - public QueryResults performQuery( QueryCommand query, - Schemata schemata, - IndexContext indexes ) throws IOException, ParseException { - return this.queryEngine.execute(query, schemata, indexes); - } - - /** - * {@inheritDoc} - * - * @see org.jboss.dna.search.IndexingStrategy#apply(Iterable, IndexContext) - */ - public int apply( Iterable<ChangeRequest> changes, - IndexContext indexes ) /*throws IOException*/{ - for (ChangeRequest change : changes) { - if (change != null) continue; - } - return 0; - } -} Modified: trunk/dna-search/src/main/java/org/jboss/dna/search/WorkspaceSearchEngine.java =================================================================== --- trunk/dna-search/src/main/java/org/jboss/dna/search/WorkspaceSearchEngine.java 2009-11-16 23:23:02 UTC (rev 1318) +++ trunk/dna-search/src/main/java/org/jboss/dna/search/WorkspaceSearchEngine.java 2009-11-16 23:24:06 UTC (rev 1319) @@ -55,7 +55,7 @@ * A search engine dedicated to a single workspace. */ @ThreadSafe -public class WorkspaceSearchEngine { +class WorkspaceSearchEngine { protected static final String PATHS_INDEX_NAME = "paths"; protected static final String CONTENT_INDEX_NAME = "content"; @@ -67,7 +67,7 @@ private final String sourceName; private final String workspaceName; private final RepositoryConnectionFactory connectionFactory; - private final IndexingStrategy indexingStrategy; + private final IndexStrategy indexingStrategy; protected final AtomicInteger modifiedNodesSinceLastOptimize = new AtomicInteger(0); /** @@ -86,7 +86,7 @@ */ protected WorkspaceSearchEngine( ExecutionContext context, DirectoryConfiguration directoryFactory, - IndexingStrategy indexingStrategy, + IndexStrategy indexingStrategy, String sourceName, String workspaceName, RepositoryConnectionFactory connectionFactory ) throws SearchEngineException { @@ -168,7 +168,7 @@ return context.getValueFactories().getStringFactory().create(path); } - final IndexingStrategy strategy() { + final IndexStrategy strategy() { return indexingStrategy; } @@ -523,7 +523,7 @@ final List<Location> results = new ArrayList<Location>(maxResults); return new Search() { public void execute( IndexContext indexes ) throws IOException, ParseException { - strategy().performQuery(fullTextSearch, maxResults, offset, indexes, results); + strategy().search(fullTextSearch, maxResults, offset, indexes, results); } public String messageFor( Throwable error ) { @@ -550,9 +550,11 @@ final Schemata schemata ) { return new Query() { private QueryResults results = null; + private SearchContext context = null; - public void execute( IndexContext indexes ) throws IOException, ParseException { - results = strategy().performQuery(query, schemata, indexes); + public void execute( IndexContext indexes ) { + context = new SearchContext(indexes, schemata); + results = strategy().query(context, query); } public String messageFor( Throwable error ) { Modified: trunk/dna-search/src/test/java/org/jboss/dna/search/IndexingRulesTest.java =================================================================== --- trunk/dna-search/src/test/java/org/jboss/dna/search/IndexingRulesTest.java 2009-11-16 23:23:02 UTC (rev 1318) +++ trunk/dna-search/src/test/java/org/jboss/dna/search/IndexingRulesTest.java 2009-11-16 23:24:06 UTC (rev 1319) @@ -25,7 +25,7 @@ import static org.hamcrest.core.Is.is; import static org.junit.Assert.assertThat; -import org.jboss.dna.search.IndexingRules.Builder; +import org.jboss.dna.search.IndexRules.Builder; import org.junit.Before; import org.junit.Test; @@ -35,23 +35,23 @@ public class IndexingRulesTest { private Builder builder; - private IndexingRules rules; + private IndexRules rules; @Before public void beforeEach() { - builder = IndexingRules.createBuilder(); + builder = IndexRules.createBuilder(); rules = builder.build(); } @Test public void shouldBuildValidRulesFromBuilderThatIsNotInvoked() { - builder = IndexingRules.createBuilder(); + builder = IndexRules.createBuilder(); rules = builder.build(); } @Test public void shouldBuildValidRulesFromBuilderAfterJustSettingDefaultRules() { - builder.defaultTo(IndexingRules.FULL_TEXT); + builder.defaultTo(IndexRules.FULL_TEXT); rules = builder.build(); assertThat(rules.getRule(null).isFullText(), is(true)); } Modified: trunk/dna-search/src/test/java/org/jboss/dna/search/SearchEngineTest.java =================================================================== --- trunk/dna-search/src/test/java/org/jboss/dna/search/SearchEngineTest.java 2009-11-16 23:23:02 UTC (rev 1318) +++ trunk/dna-search/src/test/java/org/jboss/dna/search/SearchEngineTest.java 2009-11-16 23:24:06 UTC (rev 1319) @@ -48,7 +48,6 @@ private InMemoryRepositorySource source; private RepositoryConnectionFactory connectionFactory; private DirectoryConfiguration directoryFactory; - private IndexingStrategy indexingStrategy; private Graph content; @Before @@ -81,15 +80,9 @@ } }; - // Set up the indexing strategy ... - IndexingRules rules = IndexingRules.createBuilder(StoreLittleIndexingStrategy.DEFAULT_RULES) - .defaultTo(IndexingRules.INDEX | IndexingRules.ANALYZE | IndexingRules.FULL_TEXT) - .build(); - indexingStrategy = new StoreLittleIndexingStrategy(rules); - // Now set up the search engine ... directoryFactory = DirectoryConfigurations.inMemory(); - engine = new SearchEngine(context, sourceName, connectionFactory, directoryFactory, indexingStrategy); + engine = new SearchEngine(context, sourceName, connectionFactory, directoryFactory); } protected Path path( String string ) { Modified: trunk/dna-search/src/test/java/org/jboss/dna/search/WorkspaceSearchEngineTest.java =================================================================== --- trunk/dna-search/src/test/java/org/jboss/dna/search/WorkspaceSearchEngineTest.java 2009-11-16 23:23:02 UTC (rev 1318) +++ trunk/dna-search/src/test/java/org/jboss/dna/search/WorkspaceSearchEngineTest.java 2009-11-16 23:24:06 UTC (rev 1319) @@ -47,7 +47,7 @@ private InMemoryRepositorySource source; private RepositoryConnectionFactory connectionFactory; private DirectoryConfiguration directoryFactory; - private IndexingStrategy indexingStrategy; + private IndexStrategy indexingStrategy; private Graph content; @Before @@ -74,10 +74,10 @@ }; // Set up the indexing strategy ... - IndexingRules rules = IndexingRules.createBuilder(StoreLittleIndexingStrategy.DEFAULT_RULES) - .defaultTo(IndexingRules.INDEX | IndexingRules.ANALYZE | IndexingRules.FULL_TEXT) + IndexRules rules = IndexRules.createBuilder(KitchenSinkIndexStrategy.DEFAULT_RULES) + .defaultTo(IndexRules.INDEX | IndexRules.ANALYZE | IndexRules.FULL_TEXT) .build(); - indexingStrategy = new StoreLittleIndexingStrategy(rules); + indexingStrategy = new KitchenSinkIndexStrategy(rules); // Now set up the search engine ... directoryFactory = DirectoryConfigurations.inMemory();