Source Code Cross Referenced for TokenizerFactory.java in  » Search-Engine » apache-solr-1.2.0 » org » apache » solr » analysis » Java Source Code / Java DocumentationJava Source Code and Java Documentation

01:        /**
02:         * Licensed to the Apache Software Foundation (ASF) under one or more
03:         * contributor license agreements.  See the NOTICE file distributed with
04:         * this work for additional information regarding copyright ownership.
05:         * The ASF licenses this file to You under the Apache License, Version 2.0
06:         * (the "License"); you may not use this file except in compliance with
07:         * the License.  You may obtain a copy of the License at
08:         *
09:         *     http://www.apache.org/licenses/LICENSE-2.0
10:         *
11:         * Unless required by applicable law or agreed to in writing, software
12:         * distributed under the License is distributed on an "AS IS" BASIS,
13:         * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
14:         * See the License for the specific language governing permissions and
15:         * limitations under the License.
16:         */package org.apache.solr.analysis;
17:
18:        import java.io.*;
19:        import java.util.Map;
20:
21:        import org.apache.lucene.analysis.*;
22:
23:        /**
24:         * A <code>TokenizerFactory</code> breaks up a stream of characters 
25:         * into tokens.
26:         *
27:         * <p>
28:         * TokenizerFactories are registered for <code>FieldType</code>s with the
29:         * <code>IndexSchema</code> through the <code>schema.xml</code> file.
30:         * </p>
31:         * <p>
32:         * Example <code>schema.xml</code> entry to register a TokenizerFactory 
33:         * implementation to tokenize fields of type "cool"
34:         * </p>
35:         * <pre>
36:         *  &lt;fieldtype name="cool" class="solr.TextField"&gt;
37:         *      &lt;analyzer&gt;
38:         *      &lt;tokenizer class="solr.StandardTokenizerFactory"/&gt;
39:         *      ...
40:         * </pre>
41:         * <p>
42:         * A single instance of any registered TokenizerFactory is created
43:         * via the default constructor and is reused for each FieldType.
44:         * </p>
45:         * @author yonik
46:         * @version $Id: TokenizerFactory.java 472574 2006-11-08 18:25:52Z yonik $
47:         */
48:        public interface TokenizerFactory {
49:            /** <code>init</code> will be called just once, immediately after creation.
50:             * <p>The args are user-level initialization parameters that
51:             * may be specified when declaring a the factory in the
52:             * schema.xml
53:             */
54:            public void init(Map<String, String> args);
55:
56:            /**
57:             * Accessor method for reporting the args used to initialize this factory.
58:             * <p>
59:             * Implementations are <strong>strongly</strong> encouraged to return 
60:             * the contents of the Map passed to to the init method
61:             * </p>
62:             */
63:            public Map<String, String> getArgs();
64:
65:            /** Creates a TokenStream of the specified input */
66:            public TokenStream create(Reader input);
67:        }