1 package org.apache.lucene.facet.index;
3 import java.io.IOException;
4 import java.util.ArrayList;
5 import java.util.HashMap;
8 import java.util.Map.Entry;
10 import org.apache.lucene.analysis.TokenStream;
11 import org.apache.lucene.document.Document;
12 import org.apache.lucene.document.Field;
14 import org.apache.lucene.facet.index.attributes.CategoryAttribute;
15 import org.apache.lucene.facet.index.attributes.CategoryAttributesIterable;
16 import org.apache.lucene.facet.index.categorypolicy.OrdinalPolicy;
17 import org.apache.lucene.facet.index.categorypolicy.PathPolicy;
18 import org.apache.lucene.facet.index.params.DefaultFacetIndexingParams;
19 import org.apache.lucene.facet.index.params.FacetIndexingParams;
20 import org.apache.lucene.facet.index.streaming.CategoryAttributesStream;
21 import org.apache.lucene.facet.index.streaming.CategoryListTokenizer;
22 import org.apache.lucene.facet.index.streaming.CategoryParentsStream;
23 import org.apache.lucene.facet.index.streaming.CategoryTokenizer;
24 import org.apache.lucene.facet.index.streaming.CountingListTokenizer;
25 import org.apache.lucene.facet.taxonomy.CategoryPath;
26 import org.apache.lucene.facet.taxonomy.TaxonomyWriter;
29 * Licensed to the Apache Software Foundation (ASF) under one or more
30 * contributor license agreements. See the NOTICE file distributed with
31 * this work for additional information regarding copyright ownership.
32 * The ASF licenses this file to You under the Apache License, Version 2.0
33 * (the "License"); you may not use this file except in compliance with
34 * the License. You may obtain a copy of the License at
36 * http://www.apache.org/licenses/LICENSE-2.0
38 * Unless required by applicable law or agreed to in writing, software
39 * distributed under the License is distributed on an "AS IS" BASIS,
40 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
41 * See the License for the specific language governing permissions and
42 * limitations under the License.
46 * A utility class which allows attachment of {@link CategoryPath}s or
47 * {@link CategoryAttribute}s to a given document using a taxonomy.<br>
48 * Construction could be done with either a given {@link FacetIndexingParams} or
49 * the default implementation {@link DefaultFacetIndexingParams}.<br>
50 * A CategoryDocumentBuilder can be reused by repeatedly setting the categories
51 * and building the document. Categories are provided either as
52 * {@link CategoryAttribute} elements through {@link #setCategories(Iterable)},
53 * or as {@link CategoryPath} elements through
54 * {@link #setCategoryPaths(Iterable)}.
56 * Note that both {@link #setCategories(Iterable)} and
57 * {@link #setCategoryPaths(Iterable)} return this
58 * {@link CategoryDocumentBuilder}, allowing the following pattern: {@code new
59 * CategoryDocumentBuilder(taxonomy,
60 * params).setCategories(categories).build(doc)}.
62 * @lucene.experimental
64 public class CategoryDocumentBuilder {
67 * A {@link TaxonomyWriter} for adding categories and retrieving their
70 protected final TaxonomyWriter taxonomyWriter;
73 * Parameters to be used when indexing categories.
75 protected final FacetIndexingParams indexingParams;
78 * A list of fields which is filled at ancestors' construction and used
79 * during {@link CategoryDocumentBuilder#build(Document)}.
81 protected final ArrayList<Field> fieldList = new ArrayList<Field>();
83 protected Map<String, List<CategoryAttribute>> categoriesMap;
86 * Creating a facets document builder with default facet indexing
89 * {@link #CategoryDocumentBuilder(TaxonomyWriter, FacetIndexingParams)}
91 * @param taxonomyWriter
92 * to which new categories will be added, as well as translating
93 * known categories to ordinals
97 public CategoryDocumentBuilder(TaxonomyWriter taxonomyWriter)
99 this(taxonomyWriter, new DefaultFacetIndexingParams());
103 * Creating a facets document builder with a given facet indexing parameters
106 * @param taxonomyWriter
107 * to which new categories will be added, as well as translating
108 * known categories to ordinals
110 * holds all parameters the indexing process should use such as
111 * category-list parameters
112 * @throws IOException
114 public CategoryDocumentBuilder(TaxonomyWriter taxonomyWriter,
115 FacetIndexingParams params) throws IOException {
116 this.taxonomyWriter = taxonomyWriter;
117 this.indexingParams = params;
118 this.categoriesMap = new HashMap<String, List<CategoryAttribute>>();
122 * Set the categories of the document builder from an {@link Iterable} of
123 * {@link CategoryPath} objects.
125 * @param categoryPaths
126 * An iterable of CategoryPath objects which holds the categories
127 * (facets) which will be added to the document at
128 * {@link #build(Document)}
129 * @return This CategoryDocumentBuilder, to enable this one line call:
130 * {@code new} {@link #CategoryDocumentBuilder(TaxonomyWriter)}.
131 * {@link #setCategoryPaths(Iterable)}.{@link #build(Document)}.
132 * @throws IOException
134 public CategoryDocumentBuilder setCategoryPaths(
135 Iterable<CategoryPath> categoryPaths) throws IOException {
136 if (categoryPaths == null) {
140 return setCategories(new CategoryAttributesIterable(categoryPaths));
144 * Set the categories of the document builder from an {@link Iterable} of
145 * {@link CategoryAttribute} objects.
148 * An iterable of {@link CategoryAttribute} objects which holds
149 * the categories (facets) which will be added to the document at
150 * {@link #build(Document)}
151 * @return This CategoryDocumentBuilder, to enable this one line call:
152 * {@code new} {@link #CategoryDocumentBuilder(TaxonomyWriter)}.
153 * {@link #setCategories(Iterable)}.{@link #build(Document)}.
154 * @throws IOException
156 public CategoryDocumentBuilder setCategories(
157 Iterable<CategoryAttribute> categories) throws IOException {
159 if (categories == null) {
163 // get field-name to a list of facets mapping as different facets could
164 // be added to different category-lists on different fields
165 fillCategoriesMap(categories);
167 // creates a different stream for each different field
168 for (Entry<String, List<CategoryAttribute>> e : categoriesMap
170 // create a category attributes stream for the array of facets
171 CategoryAttributesStream categoryAttributesStream = new CategoryAttributesStream(
174 // Set a suitable {@link TokenStream} using
175 // CategoryParentsStream, followed by CategoryListTokenizer and
176 // CategoryTokenizer composition (the ordering of the last two is
178 CategoryParentsStream parentsStream = (CategoryParentsStream) getParentsStream(categoryAttributesStream);
179 CategoryListTokenizer categoryListTokenizer = getCategoryListTokenizer(parentsStream);
180 CategoryTokenizer stream = getCategoryTokenizer(categoryListTokenizer);
182 // Finally creating a suitable field with stream and adding it to a
183 // master field-list, used during the build process (see
185 fieldList.add(new Field(e.getKey(), stream));
192 * Get a stream of categories which includes the parents, according to
193 * policies defined in indexing parameters.
195 * @param categoryAttributesStream
197 * @return The parents stream.
198 * @see OrdinalPolicy OrdinalPolicy (for policy of adding category tokens for parents)
199 * @see PathPolicy PathPolicy (for policy of adding category <b>list</b> tokens for parents)
201 protected TokenStream getParentsStream(
202 CategoryAttributesStream categoryAttributesStream) {
203 return new CategoryParentsStream(categoryAttributesStream,
204 taxonomyWriter, indexingParams);
208 * Fills the categories mapping between a field name and a list of
209 * categories that belongs to it according to this builder's
210 * {@link FacetIndexingParams} object
213 * Iterable over the category attributes
215 protected void fillCategoriesMap(Iterable<CategoryAttribute> categories)
217 categoriesMap.clear();
220 for (CategoryAttribute category : categories) {
221 // extracting the field-name to which this category belongs
222 String fieldName = indexingParams.getCategoryListParams(
223 category.getCategoryPath()).getTerm().field();
225 // getting the list of categories which belongs to that field
226 List<CategoryAttribute> list = categoriesMap.get(fieldName);
228 // if no such list exists
230 // adding a new one to the map
231 list = new ArrayList<CategoryAttribute>();
232 categoriesMap.put(fieldName, list);
235 // adding the new category to the list
236 list.add(category.clone());
241 * Get a category list tokenizer (or a series of such tokenizers) to create
242 * the <b>category list tokens</b>.
244 * @param categoryStream
245 * A stream containing {@link CategoryAttribute} with the
247 * @return The category list tokenizer (or series of tokenizers) to be used
248 * in creating category list tokens.
250 protected CategoryListTokenizer getCategoryListTokenizer(
251 TokenStream categoryStream) {
252 return getCountingListTokenizer(categoryStream);
256 * Get a {@link CountingListTokenizer} for creating counting list token.
258 * @param categoryStream
259 * A stream containing {@link CategoryAttribute}s with the
261 * @return A counting list tokenizer to be used in creating counting list
264 protected CountingListTokenizer getCountingListTokenizer(
265 TokenStream categoryStream) {
266 return new CountingListTokenizer(categoryStream, indexingParams);
270 * Get a {@link CategoryTokenizer} to create the <b>category tokens</b>.
271 * This method can be overridden for adding more attributes to the category
274 * @param categoryStream
275 * A stream containing {@link CategoryAttribute} with the
277 * @return The {@link CategoryTokenizer} to be used in creating category
279 * @throws IOException
281 protected CategoryTokenizer getCategoryTokenizer(TokenStream categoryStream)
283 return new CategoryTokenizer(categoryStream, indexingParams);
286 /** Adds the fields created in one of the "set" methods to the document */
287 public Document build(Document doc) {
288 for (Field f : fieldList) {
289 f.setOmitNorms(true);