1 /*
2 * Copyright (c) 2012, 2013, Oracle and/or its affiliates. All rights reserved.
3 * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
4 *
5 * This code is free software; you can redistribute it and/or modify it
6 * under the terms of the GNU General Public License version 2 only, as
7 * published by the Free Software Foundation. Oracle designates this
8 * particular file as subject to the "Classpath" exception as provided
9 * by Oracle in the LICENSE file that accompanied this code.
10 *
11 * This code is distributed in the hope that it will be useful, but WITHOUT
12 * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
13 * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
14 * version 2 for more details (a copy is included in the LICENSE file that
15 * accompanied this code).
16 *
17 * You should have received a copy of the GNU General Public License version
18 * 2 along with this work; if not, write to the Free Software Foundation,
19 * Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA.
20 *
21 * Please contact Oracle, 500 Oracle Parkway, Redwood Shores, CA 94065 USA
22 * or visit www.oracle.com if you need additional information or have any
23 * questions.
24 */
25 package java.util.stream;
26
27 import java.util.Collections;
28 import java.util.Set;
29 import java.util.function.BiFunction;
30 import java.util.function.BinaryOperator;
31 import java.util.function.Supplier;
32
33 /**
34 * A <a href="package-summary.html#Reduction">reduction operation</a> that
35 * supports folding input elements into a cumulative result. The result may be
36 * a value or may be a mutable result container. Examples of operations
37 * accumulating results into a mutable result container include: accumulating
38 * input elements into a {@code Collection}; concatenating strings into a
39 * {@code StringBuilder}; computing summary information about elements such as
40 * sum, min, max, or average; computing "pivot table" summaries such as "maximum
41 * valued transaction by seller", etc. Reduction operations can be performed
42 * either sequentially or in parallel.
43 *
44 * <p>The following are examples of using the predefined {@code Collector}
45 * implementations in {@link Collectors} with the {@code Stream} API to perform
46 * mutable reduction tasks:
47 * <pre>{@code
48 * // Accumulate elements into a List
49 * List<String> list = stream.collect(Collectors.toList());
50 *
51 * // Accumulate elements into a TreeSet
52 * Set<String> list = stream.collect(Collectors.toCollection(TreeSet::new));
53 *
54 * // Convert elements to strings and concatenate them, separated by commas
55 * String joined = stream.map(Object::toString)
56 * .collect(Collectors.toStringJoiner(", "))
57 * .toString();
58 *
59 * // Find highest-paid employee
60 * Employee highestPaid = employees.stream()
61 * .collect(Collectors.maxBy(Comparators.comparing(Employee::getSalary)));
62 *
63 * // Group employees by department
64 * Map<Department, List<Employee>> byDept
65 * = employees.stream()
66 * .collect(Collectors.groupingBy(Employee::getDepartment));
67 *
68 * // Find highest-paid employee by department
69 * Map<Department, Employee> highestPaidByDept
70 * = employees.stream()
71 * .collect(Collectors.groupingBy(Employee::getDepartment,
72 * Collectors.maxBy(Comparators.comparing(Employee::getSalary))));
73 *
74 * // Partition students into passing and failing
75 * Map<Boolean, List<Student>> passingFailing =
76 * students.stream()
77 * .collect(Collectors.partitioningBy(s -> s.getGrade() >= PASS_THRESHOLD);
78 *
79 * }</pre>
80 *
81 * <p>A {@code Collector} is specified by three functions that work together to
82 * manage a result or result container. They are: creation of an initial
83 * result, incorporating a new data element into a result, and combining two
84 * results into one. The last function -- combining two results into one -- is
85 * used during parallel operations, where subsets of the input are accumulated
86 * in parallel, and then the subresults merged into a combined result. The
87 * result may be a mutable container or a value. If the result is mutable, the
88 * accumulation and combination functions may either mutate their left argument
89 * and return that (such as adding elements to a collection), or return a new
90 * result, in which case it should not perform any mutation.
91 *
92 * <p>Collectors also have a set of characteristics, including
93 * {@link Characteristics#CONCURRENT} and
94 * {@link Characteristics#STRICTLY_MUTATIVE}. These characteristics provide
95 * hints that can be used by a reduction implementation to provide better
96 * performance.
97 *
98 * <p>Libraries that implement reduction based on {@code Collector}, such as
99 * {@link Stream#collect(Collector)}, must adhere to the following constraints:
100 * <ul>
101 * <li>The first argument passed to the accumulator function, and both
102 * arguments passed to the combiner function, must be the result of a
103 * previous invocation of {@link #resultSupplier()}, {@link #accumulator()},
104 * or {@link #combiner()}.</li>
105 * <li>The implementation should not do anything with the result of any of
106 * the result supplier, accumulator, or combiner functions other than to
107 * pass them again to the accumulator or combiner functions, or return them
108 * to the caller of the reduction operation.</li>
109 * <li>If a result is passed to the accumulator or combiner function, and
110 * the same object is not returned from that function, it is never used
111 * again.</li>
112 * <li>Once a result is passed to the combiner function, it is never passed
113 * to the accumulator function again.</li>
114 * <li>For non-concurrent collectors, any result returned from the result
115 * supplier, accumulator, or combiner functions must be serially
116 * thread-confined. This enables collection to occur in parallel without
117 * the {@code Collector} needing to implement any additional synchronization.
118 * The reduction implementation must manage that the input is properly
119 * partitioned, that partitions are processed in isolation, and combining
120 * happens only after accumulation is complete.</li>
121 * <li>For concurrent collectors, an implementation is free to (but not
122 * required to) implement reduction concurrently. A concurrent reduction
123 * is one where the accumulator function is called concurrently from
124 * multiple threads, using the same concurrently-modifiable result container,
125 * rather than keeping the result isolated during accumulation.
126 * A concurrent reduction should only be applied if the collector has the
127 * {@link Characteristics#UNORDERED} characteristics or if the
128 * originating data is unordered.</li>
129 * </ul>
130 *
131 * @apiNote
132 * Performing a reduction operation with a {@code Collector} should produce a
133 * result equivalent to:
134 * <pre>{@code
135 * BiFunction<R,T,R> accumulator = collector.accumulator();
136 * R result = collector.resultSupplier().get();
137 * for (T t : data)
138 * result = accumulator.apply(result, t);
139 * return result;
140 * }</pre>
141 *
142 * <p>However, the library is free to partition the input, perform the reduction
143 * on the partitions, and then use the combiner function to combine the partial
144 * results to achieve a parallel reduction. Depending on the specific reduction
145 * operation, this may perform better or worse, depending on the relative cost
146 * of the accumulator and combiner functions.
147 *
148 * <p>An example of an operation that can be easily modeled by {@code Collector}
149 * is accumulating elements into a {@code TreeSet}. In this case, the {@code
150 * resultSupplier()} function is {@code () -> new Treeset<T>()}, the
151 * {@code accumulator} function is
152 * {@code (set, element) -> { set.add(element); return set; }}, and the combiner
153 * function is {@code (left, right) -> { left.addAll(right); return left; }}.
154 * (This behavior is implemented by
155 * {@code Collectors.toCollection(TreeSet::new)}).
156 *
157 * TODO Associativity and commutativity
158 *
159 * @see Stream#collect(Collector)
160 * @see Collectors
161 *
162 * @param <T> the type of input element to the collect operation
163 * @param <R> the result type of the collect operation
164 * @since 1.8
165 */
166 public interface Collector<T, R> {
167 /**
168 * A function that creates and returns a new result that represents
169 * "no values". If the accumulator or combiner functions may mutate their
170 * arguments, this must be a new, empty result container.
171 *
172 * @return a function which, when invoked, returns a result representing
173 * "no values"
174 */
175 Supplier<R> resultSupplier();
176
177 /**
178 * A function that folds a new value into a cumulative result. The result
179 * may be a mutable result container or a value. The accumulator function
180 * may modify a mutable container and return it, or create a new result and
181 * return that, but if it returns a new result object, it must not modify
182 * any of its arguments.
183 *
184 * <p>If the collector has the {@link Characteristics#STRICTLY_MUTATIVE}
185 * characteristic, then the accumulator function <em>must</em> always return
186 * its first argument, after possibly mutating its state.
187 *
188 * @return a function which folds a new value into a cumulative result
189 */
190 BiFunction<R, T, R> accumulator();
191
192 /**
193 * A function that accepts two partial results and merges them. The
194 * combiner function may fold state from one argument into the other and
195 * return that, or may return a new result object, but if it returns
196 * a new result object, it must not modify the state of either of its
197 * arguments.
198 *
199 * <p>If the collector has the {@link Characteristics#STRICTLY_MUTATIVE}
200 * characteristic, then the combiner function <em>must</em> always return
201 * its first argument, after possibly mutating its state.
202 *
203 * @return a function which combines two partial results into a cumulative
204 * result
205 */
206 BinaryOperator<R> combiner();
207
208 /**
209 * Returns a {@code Set} of {@code Collector.Characteristics} indicating
210 * the characteristics of this Collector. This set should be immutable.
211 *
212 * @return an immutable set of collector characteristics
213 */
214 Set<Characteristics> characteristics();
215
216 /**
217 * Characteristics indicating properties of a {@code Collector}, which can
218 * be used to optimize reduction implementations.
219 */
220 enum Characteristics {
221 /**
222 * Indicates that this collector is <em>concurrent</em>, meaning that
223 * the result container can support the accumulator function being
224 * called concurrently with the same result container from multiple
225 * threads. Concurrent collectors must also always have the
226 * {@code STRICTLY_MUTATIVE} characteristic.
227 *
228 * <p>If a {@code CONCURRENT} collector is not also {@code UNORDERED},
229 * then it should only be evaluated concurrently if applied to an
230 * unordered data source.
231 */
232 CONCURRENT,
233
234 /**
235 * Indicates that the result container has no intrinsic order, such as
236 * a {@link Set}.
237 */
238 UNORDERED,
239
240 /**
241 * Indicates that this collector operates by strict mutation of its
242 * result container. This means that the {@link #accumulator()} and
243 * {@link #combiner()} functions will always modify the state of and
244 * return their first argument, rather than returning a different result
245 * container.
246 */
247 STRICTLY_MUTATIVE
248 }
249 }