[12264] | 1 | /** @file GS2LuceneDelete.java
|
---|
| 2 | *
|
---|
| 3 | * Provides a wrapper to the document deleting features of Lucene.
|
---|
| 4 | *
|
---|
| 5 | * This java application makes use of the existing Lucene class IndexModifier
|
---|
| 6 | * to access and make changes to the information stored about documents in a
|
---|
| 7 | * Lucene database. This is an essential component of the IncrementalBuilder
|
---|
| 8 | * PERL module, and endevours to make editing the text and metadata of
|
---|
| 9 | * documents without having to rebuild the entire collection a reality (in
|
---|
| 10 | * other words, true incremental/dynamic building).
|
---|
| 11 | *
|
---|
| 12 | * A component of the Greenstone digital library software from the New Zealand
|
---|
| 13 | * Digital Library Project at the University of Waikato, New Zealand.
|
---|
| 14 | *
|
---|
| 15 | * This program is free software; you can redistribute it and/or modify it
|
---|
| 16 | * under the terms of the GNU General Public License as published by the Free
|
---|
| 17 | * Software Foundation; either version 2 of the License, or (at your option)
|
---|
| 18 | * any later version.
|
---|
| 19 | *
|
---|
| 20 | * This program is distributed in the hope that it will be useful, but WITHOUT
|
---|
| 21 | * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
|
---|
| 22 | * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for
|
---|
| 23 | * more details.
|
---|
| 24 | *
|
---|
| 25 | * You should have received a copy of the GNU General Public License along
|
---|
| 26 | * with this program; if not, write to the Free Software Foundation, Inc., 675
|
---|
| 27 | * Mass Ave, Cambridge, MA 02139, USA.
|
---|
| 28 | *
|
---|
| 29 | * Copyright (c) 2006 DL Consulting Ltd., New Zealand
|
---|
| 30 | */
|
---|
| 31 |
|
---|
[13686] | 32 | package org.greenstone.LuceneWrapper;
|
---|
[12264] | 33 |
|
---|
| 34 | import java.io.IOException;
|
---|
| 35 | import org.apache.lucene.analysis.standard.StandardAnalyzer;
|
---|
[16429] | 36 | import org.apache.lucene.index.IndexWriter;
|
---|
| 37 | import org.apache.lucene.index.Term;
|
---|
[12264] | 38 |
|
---|
| 39 |
|
---|
| 40 | /** Contains methods for deleting a document that has previously been indexed
|
---|
| 41 | * into a Lucene database.
|
---|
| 42 | * @author John Thompson, DL Consulting Ltd. (unless stated otherwise)
|
---|
| 43 | */
|
---|
| 44 | public class GS2LuceneDelete
|
---|
| 45 | {
|
---|
| 46 | /** This is the main entry point to the deletor and is responsible for
|
---|
| 47 | * parsing the arguments and creating an instance of the deletor class.
|
---|
| 48 | *
|
---|
| 49 | * @param args The arguments passed into the application as a string
|
---|
| 50 | * array
|
---|
| 51 | * @return An integer describing the exit state of the application
|
---|
| 52 | * @throws Exception on any fatal error state
|
---|
| 53 | */
|
---|
| 54 | static public void main (String args[])
|
---|
| 55 | throws Exception
|
---|
| 56 | {
|
---|
| 57 | // Parse arguments
|
---|
[16429] | 58 | String index_path = "";
|
---|
[12264] | 59 | int node_id = -1;
|
---|
| 60 |
|
---|
| 61 | for (int i = 0; i < args.length; i += 2)
|
---|
[16429] | 62 | {
|
---|
| 63 | if (args[i].equals("--index"))
|
---|
| 64 | {
|
---|
| 65 | index_path = args[i + 1];
|
---|
| 66 | }
|
---|
| 67 | else if (args[i].equals("--nodeid"))
|
---|
| 68 | {
|
---|
| 69 | node_id = Integer.parseInt(args[i + 1]);
|
---|
| 70 | }
|
---|
| 71 | else
|
---|
| 72 | {
|
---|
| 73 | System.out.println("Error! Unknown argument: " + args[i]);
|
---|
| 74 | GS2LuceneDelete.printUsage();
|
---|
| 75 | System.exit(0);
|
---|
| 76 | }
|
---|
| 77 | }
|
---|
[12264] | 78 |
|
---|
| 79 | // Check arguments
|
---|
[16429] | 80 | if (index_path.equals(""))
|
---|
| 81 | {
|
---|
| 82 | System.out.println("Error! Missing index path");
|
---|
| 83 | GS2LuceneDelete.printUsage();
|
---|
| 84 | System.exit(0);
|
---|
| 85 | }
|
---|
| 86 | if (node_id == -1)
|
---|
| 87 | {
|
---|
| 88 | System.out.println("Error! Missing or invalid Node ID");
|
---|
| 89 | GS2LuceneDelete.printUsage();
|
---|
| 90 | System.exit(0);
|
---|
| 91 | }
|
---|
[12264] | 92 |
|
---|
| 93 | // Instantiate deletor, and perform the delete
|
---|
| 94 | GS2LuceneDelete deletor = new GS2LuceneDelete(index_path);
|
---|
| 95 | deletor.deleteDocument(node_id);
|
---|
| 96 | deletor.destroy();
|
---|
| 97 | deletor = null;
|
---|
| 98 | }
|
---|
| 99 |
|
---|
[16429] | 100 |
|
---|
| 101 | /** Display program usage message.
|
---|
| 102 | */
|
---|
| 103 | static public void printUsage()
|
---|
| 104 | {
|
---|
| 105 | System.out.println("usage: GS2LuceneDelete --index <path> --nodeid <int>");
|
---|
| 106 | System.out.println("");
|
---|
| 107 | System.out.println("where:");
|
---|
| 108 | System.out.println(" index - is the full path to the directory containing the directory");
|
---|
| 109 | System.out.println(" to edit, including the level (ie didx, sidx)");
|
---|
| 110 | System.out.println(" nodeid - the unique identifier of the document to delete. This is the");
|
---|
| 111 | System.out.println(" same as the docnum in the GDBM");
|
---|
| 112 | System.out.println("");
|
---|
| 113 | }
|
---|
| 114 |
|
---|
| 115 |
|
---|
[12264] | 116 | /** **/
|
---|
| 117 | private boolean debug = true;
|
---|
| 118 |
|
---|
| 119 | /** **/
|
---|
[16429] | 120 | private IndexWriter index_writer = null;
|
---|
[12264] | 121 |
|
---|
[16429] | 122 |
|
---|
[12264] | 123 | /** Constructor which takes the path to the Lucene index to be edited.
|
---|
| 124 | *
|
---|
| 125 | * @param index_path The full path to the index directory as a String
|
---|
| 126 | */
|
---|
| 127 | public GS2LuceneDelete(String index_path)
|
---|
| 128 | throws IOException
|
---|
| 129 | {
|
---|
[16429] | 130 | index_writer = new IndexWriter(index_path, new StandardAnalyzer());
|
---|
[12264] | 131 | }
|
---|
| 132 |
|
---|
[16429] | 133 |
|
---|
[12264] | 134 | /** When called prints a debug message but only if debugging is enabled.
|
---|
| 135 | */
|
---|
| 136 | public void debug(String message)
|
---|
| 137 | {
|
---|
[16429] | 138 | if (debug)
|
---|
| 139 | {
|
---|
| 140 | System.err.println(message);
|
---|
| 141 | }
|
---|
[12264] | 142 | }
|
---|
| 143 |
|
---|
[16429] | 144 |
|
---|
[12264] | 145 | /** Destructor which unallocates connection to Lucene.
|
---|
| 146 | */
|
---|
| 147 | public void destroy()
|
---|
| 148 | throws IOException
|
---|
| 149 | {
|
---|
[16429] | 150 | index_writer.close();
|
---|
| 151 | index_writer = null;
|
---|
[12264] | 152 | }
|
---|
| 153 |
|
---|
[16429] | 154 |
|
---|
[12264] | 155 | /** Delete the indicated document from the Lucene index. This process is
|
---|
| 156 | * very similar to the initial step of index editing.
|
---|
| 157 | *
|
---|
| 158 | * @param node_id The unique identifier of a Lucene document as an
|
---|
| 159 | * integer
|
---|
| 160 | */
|
---|
| 161 | public void deleteDocument(int node_id)
|
---|
| 162 | throws IOException
|
---|
| 163 | {
|
---|
| 164 | debug("GS2LuceneDelete.deleteDocument(" + node_id + ")");
|
---|
[16429] | 165 | debug("- Initial number of documents in index: " + index_writer.docCount());
|
---|
| 166 | index_writer.deleteDocuments(new Term("nodeid", "" + node_id));
|
---|
| 167 | debug("- Final number of documents in index: " + index_writer.docCount());
|
---|
[12264] | 168 | }
|
---|
| 169 | }
|
---|