Source code

001package org.apache.archiva.repository.scanner;
002
003/*
004 * Licensed to the Apache Software Foundation (ASF) under one
005 * or more contributor license agreements.  See the NOTICE file
006 * distributed with this work for additional information
007 * regarding copyright ownership.  The ASF licenses this file
008 * to you under the Apache License, Version 2.0 (the
009 * "License"); you may not use this file except in compliance
010 * with the License.  You may obtain a copy of the License at
011 *
012 *  http://www.apache.org/licenses/LICENSE-2.0
013 *
014 * Unless required by applicable law or agreed to in writing,
015 * software distributed under the License is distributed on an
016 * "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
017 * KIND, either express or implied.  See the License for the
018 * specific language governing permissions and limitations
019 * under the License.
020 */
021
022import org.apache.archiva.consumers.InvalidRepositoryContentConsumer;
023import org.apache.archiva.consumers.KnownRepositoryContentConsumer;
024import org.apache.archiva.repository.ManagedRepository;
025
026import java.util.List;
027import java.util.Set;
028
029import static org.apache.archiva.indexer.ArchivaIndexManager.DEFAULT_INDEX_PATH;
030import static org.apache.archiva.indexer.ArchivaIndexManager.DEFAULT_PACKED_INDEX_PATH;
031
032/**
033 * RepositoryScanner
034 *
035 *
036 */
037public interface RepositoryScanner
038{
039    /**
040     * The value to pass to {@link #scan(ManagedRepository, long)} to have the scan
041     * operate in a fresh fashion, with no check on changes based on timestamp.
042     */
043    long FRESH_SCAN = 0;
044
045    /**
046     * <p>
047     * Typical Ignorable Content patterns.
048     * </p>
049     * <p><strong>
050     * NOTE: Do not use for normal webapp or task driven repository scanning.
051     * </strong></p>     
052     * <p>
053     * These patterns are only valid for archiva-cli and archiva-converter use.
054     * </p>
055     */
056    static final String[] IGNORABLE_CONTENT =
057        { "bin/**", "reports/**", DEFAULT_PACKED_INDEX_PATH, ".reports/**", ".maven/**", "**/.svn/**", "**/*snapshot-version",
058            "*/website/**", "*/licences/**", "**/.htaccess", "**/*.html", "**/*.txt", "**/README*", "**/CHANGELOG*",
059            "**/KEYS*", DEFAULT_INDEX_PATH };
060
061    /**
062     * Scan the repository for content changes.
063     * <p>
064     * Internally, this will use the as-configured known and invalid consumer lists.
065     *
066     * @param repository   the repository to change.
067     * @param changesSince the timestamp to use as a threshold on what is considered new or changed.
068     *                     (To have all content be taken into consideration regardless of timestamp,
069     *                     use the {@link #FRESH_SCAN} constant)
070     * @return the statistics for this scan.
071     * @throws RepositoryScannerException if there was a fundamental problem with getting the discoverer started.
072     */
073    RepositoryScanStatistics scan( ManagedRepository repository, long changesSince )
074        throws RepositoryScannerException;
075
076    /**
077     * Scan the repository for content changes.
078     * <p>
079     * Internally, this will use the as-configured known and invalid consumer lists.
080     *
081     * @param repository              the repository to change.
082     * @param knownContentConsumers   the list of consumers that follow the {@link KnownRepositoryContentConsumer}
083     *                                interface that should be used for this scan.
084     * @param invalidContentConsumers the list of consumers that follow the {@link InvalidRepositoryContentConsumer}
085     *                                interface that should be used for this scan.
086     * @param ignoredContentPatterns  list of patterns that should be ignored and not sent to any consumer.
087     * @param changesSince            the timestamp to use as a threshold on what is considered new or changed.
088     *                                (To have all content be taken into consideration regardless of timestamp,
089     *                                use the {@link #FRESH_SCAN} constant)
090     * @return the statistics for this scan.
091     * @throws RepositoryScannerException if there was a fundamental problem with getting the discoverer started.
092     */
093    RepositoryScanStatistics scan( ManagedRepository repository,
094                                   List<KnownRepositoryContentConsumer> knownContentConsumers,
095                                   List<InvalidRepositoryContentConsumer> invalidContentConsumers,
096                                   List<String> ignoredContentPatterns, long changesSince )
097        throws RepositoryScannerException;
098
099    Set<RepositoryScannerInstance> getInProgressScans();
100}