1 package org.apache.lucene.analysis;
4 * Licensed to the Apache Software Foundation (ASF) under one or more
5 * contributor license agreements. See the NOTICE file distributed with
6 * this work for additional information regarding copyright ownership.
7 * The ASF licenses this file to You under the Apache License, Version 2.0
8 * (the "License"); you may not use this file except in compliance with
9 * the License. You may obtain a copy of the License at
11 * http://www.apache.org/licenses/LICENSE-2.0
13 * Unless required by applicable law or agreed to in writing, software
14 * distributed under the License is distributed on an "AS IS" BASIS,
15 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
16 * See the License for the specific language governing permissions and
17 * limitations under the License.
20 import java.io.IOException;
21 import java.io.Closeable;
22 import java.lang.reflect.Modifier;
24 import org.apache.lucene.document.Document;
25 import org.apache.lucene.document.Field;
26 import org.apache.lucene.index.IndexWriter;
27 import org.apache.lucene.util.Attribute;
28 import org.apache.lucene.util.AttributeImpl;
29 import org.apache.lucene.util.AttributeSource;
32 * A <code>TokenStream</code> enumerates the sequence of tokens, either from
33 * {@link Field}s of a {@link Document} or from query text.
35 * This is an abstract class; concrete subclasses are:
37 * <li>{@link Tokenizer}, a <code>TokenStream</code> whose input is a Reader; and
38 * <li>{@link TokenFilter}, a <code>TokenStream</code> whose input is another
39 * <code>TokenStream</code>.
41 * A new <code>TokenStream</code> API has been introduced with Lucene 2.9. This API
42 * has moved from being {@link Token}-based to {@link Attribute}-based. While
43 * {@link Token} still exists in 2.9 as a convenience class, the preferred way
44 * to store the information of a {@link Token} is to use {@link AttributeImpl}s.
46 * <code>TokenStream</code> now extends {@link AttributeSource}, which provides
47 * access to all of the token {@link Attribute}s for the <code>TokenStream</code>.
48 * Note that only one instance per {@link AttributeImpl} is created and reused
49 * for every token. This approach reduces object creation and allows local
50 * caching of references to the {@link AttributeImpl}s. See
51 * {@link #incrementToken()} for further details.
53 * <b>The workflow of the new <code>TokenStream</code> API is as follows:</b>
55 * <li>Instantiation of <code>TokenStream</code>/{@link TokenFilter}s which add/get
56 * attributes to/from the {@link AttributeSource}.
57 * <li>The consumer calls {@link TokenStream#reset()}.
58 * <li>The consumer retrieves attributes from the stream and stores local
59 * references to all attributes it wants to access.
60 * <li>The consumer calls {@link #incrementToken()} until it returns false
61 * consuming the attributes after each call.
62 * <li>The consumer calls {@link #end()} so that any end-of-stream operations
64 * <li>The consumer calls {@link #close()} to release any resource when finished
65 * using the <code>TokenStream</code>.
67 * To make sure that filters and consumers know which attributes are available,
68 * the attributes must be added during instantiation. Filters and consumers are
69 * not required to check for availability of attributes in
70 * {@link #incrementToken()}.
72 * You can find some example code for the new API in the analysis package level
75 * Sometimes it is desirable to capture a current state of a <code>TokenStream</code>,
76 * e.g., for buffering purposes (see {@link CachingTokenFilter},
77 * {@link TeeSinkTokenFilter}). For this usecase
78 * {@link AttributeSource#captureState} and {@link AttributeSource#restoreState}
80 * <p>The {@code TokenStream}-API in Lucene is based on the decorator pattern.
81 * Therefore all non-abstract subclasses must be final or have at least a final
82 * implementation of {@link #incrementToken}! This is checked when Java
83 * assertions are enabled.
85 public abstract class TokenStream extends AttributeSource implements Closeable {
88 * A TokenStream using the default attribute factory.
90 protected TokenStream() {
96 * A TokenStream that uses the same attributes as the supplied one.
98 protected TokenStream(AttributeSource input) {
100 assert assertFinal();
104 * A TokenStream using the supplied AttributeFactory for creating new {@link Attribute} instances.
106 protected TokenStream(AttributeFactory factory) {
108 assert assertFinal();
111 private boolean assertFinal() {
113 final Class<?> clazz = getClass();
114 if (!clazz.desiredAssertionStatus())
116 assert clazz.isAnonymousClass() ||
117 (clazz.getModifiers() & (Modifier.FINAL | Modifier.PRIVATE)) != 0 ||
118 Modifier.isFinal(clazz.getMethod("incrementToken").getModifiers()) :
119 "TokenStream implementation classes or at least their incrementToken() implementation must be final";
121 } catch (NoSuchMethodException nsme) {
127 * Consumers (i.e., {@link IndexWriter}) use this method to advance the stream to
128 * the next token. Implementing classes must implement this method and update
129 * the appropriate {@link AttributeImpl}s with the attributes of the next
132 * The producer must make no assumptions about the attributes after the method
133 * has been returned: the caller may arbitrarily change it. If the producer
134 * needs to preserve the state for subsequent calls, it can use
135 * {@link #captureState} to create a copy of the current attribute state.
137 * This method is called for every token of a document, so an efficient
138 * implementation is crucial for good performance. To avoid calls to
139 * {@link #addAttribute(Class)} and {@link #getAttribute(Class)},
140 * references to all {@link AttributeImpl}s that this stream uses should be
141 * retrieved during instantiation.
143 * To ensure that filters and consumers know which attributes are available,
144 * the attributes must be added during instantiation. Filters and consumers
145 * are not required to check for availability of attributes in
146 * {@link #incrementToken()}.
148 * @return false for end of stream; true otherwise
150 public abstract boolean incrementToken() throws IOException;
153 * This method is called by the consumer after the last token has been
154 * consumed, after {@link #incrementToken()} returned <code>false</code>
155 * (using the new <code>TokenStream</code> API). Streams implementing the old API
156 * should upgrade to use this feature.
158 * This method can be used to perform any end-of-stream operations, such as
159 * setting the final offset of a stream. The final offset of a stream might
160 * differ from the offset of the last token eg in case one or more whitespaces
161 * followed after the last token, but a {@link WhitespaceTokenizer} was used.
163 * @throws IOException
165 public void end() throws IOException {
166 // do nothing by default
170 * Resets this stream to the beginning. This is an optional operation, so
171 * subclasses may or may not implement this method. {@link #reset()} is not needed for
172 * the standard indexing process. However, if the tokens of a
173 * <code>TokenStream</code> are intended to be consumed more than once, it is
174 * necessary to implement {@link #reset()}. Note that if your TokenStream
175 * caches tokens and feeds them back again after a reset, it is imperative
176 * that you clone the tokens when you store them away (on the first pass) as
177 * well as when you return them (on future passes after {@link #reset()}).
179 public void reset() throws IOException {}
181 /** Releases resources associated with this stream. */
182 public void close() throws IOException {}