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.EnumSet;
29 import java.util.Set;
30 import java.util.function.BiConsumer;
31 import java.util.function.BinaryOperator;
32 import java.util.function.Function;
33 import java.util.function.Supplier;
34
35 /**
36 * A <a href="package-summary.html#Reduction">reduction operation</a> that
37 * folds input elements into a mutable result container, optionally transforming
38 * the accumulated result into a final representation after all input elements
39 * have been processed.
40 *
41 * <p>Examples of mutable reduction operations include:
42 * accumulating elements into a {@code Collection}; concatenating
43 * strings using a {@code StringBuilder}; computing summary information about
44 * elements such as sum, min, max, or average; computing "pivot table" summaries
45 * such as "maximum valued transaction by seller", etc. Reduction operations
46 * can be performed either sequentially or in parallel.
47 *
48 * <p>The following are examples of using the predefined {@code Collector}
49 * implementations in {@link Collectors} with the {@code Stream} API to perform
50 * mutable reduction tasks:
51 * <pre>{@code
52 * // Accumulate names into a List
53 * List<String> list = people.stream().map(Person::getName).collect(Collectors.toList());
54 *
55 * // Accumulate names into a TreeSet
56 * Set<String> list = people.stream().map(Person::getName).collect(Collectors.toCollection(TreeSet::new));
57 *
58 * // Convert elements to strings and concatenate them, separated by commas
59 * String joined = things.stream()
60 * .map(Object::toString)
61 * .collect(Collectors.joining(", "));
62 *
63 * // Find highest-paid employee
64 * Employee highestPaid = employees.stream()
65 * .collect(Collectors.maxBy(Comparators.comparing(Employee::getSalary)))
66 * .get();
67 *
68 * // Group employees by department
69 * Map<Department, List<Employee>> byDept
70 * = employees.stream()
71 * .collect(Collectors.groupingBy(Employee::getDepartment));
72 *
73 * // Find highest-paid employee by department
74 * Map<Department, Optional<Employee>> highestPaidByDept
75 * = employees.stream()
76 * .collect(Collectors.groupingBy(Employee::getDepartment,
77 * Collectors.maxBy(Comparators.comparing(Employee::getSalary))));
78 *
79 * // Partition students into passing and failing
80 * Map<Boolean, List<Student>> passingFailing =
81 * students.stream()
82 * .collect(Collectors.partitioningBy(s -> s.getGrade() >= PASS_THRESHOLD));
83 *
84 * }</pre>
85 *
86 * <p>A {@code Collector} is specified by four functions that work together to
87 * accumulate entries into a mutable result container, and optionally perform
88 * a final transform on the result. They are: creation of a new result container,
89 * incorporating a new data element into a result container, combining two
90 * result containers into one, and performing a final transform on the container.
91 * The combiner function is used during parallel operations, where
92 * subsets of the input are accumulated into separate result
93 * containers, and then the subresults merged into a combined result. The
94 * combiner function may merge one set of subresults into the other and return
95 * that, or it may return a new object to describe the combined results.
96 *
97 * <p>Collectors also have a set of characteristics, such as
98 * {@link Characteristics#CONCURRENT}. These characteristics provide
99 * hints that can be used by a reduction implementation to provide better
100 * performance.
101 *
102 * <p>Libraries that implement reduction based on {@code Collector}, such as
103 * {@link Stream#collect(Collector)}, must adhere to the following constraints:
104 * <ul>
105 * <li>The first argument passed to the accumulator function, both
106 * arguments passed to the combiner function, and the argument passed to the
107 * finisher function must be the result of a previous invocation of the
108 * result supplier, accumulator, or combiner functions.</li>
109 * <li>The implementation should not do anything with the result of any of
110 * the result supplier, accumulator, or combiner functions other than to
111 * pass them again to the accumulator, combiner, or finisher functions,
112 * or return them to the caller of the reduction operation.</li>
113 * <li>If a result is passed to the combiner or finisher
114 * function, and the same object is not returned from that function, it is
115 * never used again.</li>
116 * <li>Once a result is passed to the combiner or finisher function, it
117 * is never passed to the accumulator function again.</li>
118 * <li>For non-concurrent collectors, any result returned from the result
119 * supplier, accumulator, or combiner functions must be serially
120 * thread-confined. This enables collection to occur in parallel without
121 * the {@code Collector} needing to implement any additional synchronization.
122 * The reduction implementation must manage that the input is properly
123 * partitioned, that partitions are processed in isolation, and combining
124 * happens only after accumulation is complete.</li>
125 * <li>For concurrent collectors, an implementation is free to (but not
126 * required to) implement reduction concurrently. A concurrent reduction
127 * is one where the accumulator function is called concurrently from
128 * multiple threads, using the same concurrently-modifiable result container,
129 * rather than keeping the result isolated during accumulation.
130 * A concurrent reduction should only be applied if the collector has the
131 * {@link Characteristics#UNORDERED} characteristics or if the
132 * originating data is unordered.</li>
133 * </ul>
134 *
135 * @apiNote
136 * Performing a reduction operation with a {@code Collector} should produce a
137 * result equivalent to:
138 * <pre>{@code
139 * R container = collector.supplier().get();
140 * for (T t : data)
141 * collector.accumulator().accept(container, t);
142 * return collector.finisher().apply(container);
143 * }</pre>
144 *
145 * <p>However, the library is free to partition the input, perform the reduction
146 * on the partitions, and then use the combiner function to combine the partial
147 * results to achieve a parallel reduction. Depending on the specific reduction
148 * operation, this may perform better or worse, depending on the relative cost
149 * of the accumulator and combiner functions.
150 *
151 * <p>An example of an operation that can be easily modeled by {@code Collector}
152 * is accumulating elements into a {@code TreeSet}. In this case, the {@code
153 * resultSupplier()} function is {@code () -> new Treeset<T>()}, the
154 * {@code accumulator} function is
155 * {@code (set, element) -> set.add(element) }, and the combiner
156 * function is {@code (left, right) -> { left.addAll(right); return left; }}.
157 * (This behavior is implemented by
158 * {@code Collectors.toCollection(TreeSet::new)}).
159 *
160 * TODO Associativity and commutativity
161 *
162 * @see Stream#collect(Collector)
163 * @see Collectors
164 *
165 * @param <T> the type of input elements to the reduction operation
166 * @param <A> the mutable accumulation type of the reduction operation (often
167 * hidden as an implementation detail)
168 * @param <R> the result type of the reduction operation
169 * @since 1.8
170 */
171 public interface Collector<T, A, R> {
172 /**
173 * A function that creates and returns a new mutable result container.
174 *
175 * @return a function which returns a new, mutable result container
176 */
177 Supplier<A> supplier();
178
179 /**
180 * A function that folds a new value into a mutable result container.
181 *
182 * @return a function which folds a new value into a mutable result container
183 */
184 BiConsumer<A, T> accumulator();
185
186 /**
187 * A function that accepts two partial results and merges them. The
188 * combiner function may fold state from one argument into the other and
189 * return that, or may return a new result object.
190 *
191 * @return a function which combines two partial results into a cumulative
192 * result
193 */
194 BinaryOperator<A> combiner();
195
196 /**
197 * Perform the final transformation from the intermediate accumulation type
198 * {@code A} to the final result representation {@code R}.
199 *
200 * <p>If the characteristic {@code IDENTITY_TRANSFORM} is
201 * set, this function may be presumed to be an identity transform with an
202 * unchecked cast from {@code A} to {@code R}.
203 *
204 * @return a function which transforms the intermediate result to the final
205 * result
206 */
207 Function<A, R> finisher();
208
209 /**
210 * Returns a {@code Set} of {@code Collector.Characteristics} indicating
211 * the characteristics of this Collector. This set should be immutable.
212 *
213 * @return an immutable set of collector characteristics
214 */
215 Set<Characteristics> characteristics();
216
217 /**
218 * Returns a new {@code Collector} described by the given {@code supplier},
219 * {@code accumulator}, and {@code combiner} functions. The resulting
220 * {@code Collector} has the {@code Collector.Characteristics.IDENTITY_FINISH}
221 * characteristic.
222 *
223 * @param supplier The supplier function for the new collector
224 * @param accumulator The accumulator function for the new collector
225 * @param combiner The combiner function for the new collector
226 * @param characteristics The collector characteristics for the new
227 * collector
228 * @param <T> The type of input elements for the new collector
229 * @param <R> The type of intermediate accumulation result, and final result,
230 * for the new collector
231 * @return the new {@code Collector}
232 */
233 public static<T, R> Collector<T, R, R> of(Supplier<R> supplier,
234 BiConsumer<R, T> accumulator,
235 BinaryOperator<R> combiner,
236 Characteristics... characteristics) {
237 Set<Characteristics> cs = (characteristics.length == 0)
238 ? Collectors.CH_ID
239 : Collections.unmodifiableSet(EnumSet.of(Collector.Characteristics.IDENTITY_FINISH,
240 characteristics));
241 return new Collectors.CollectorImpl<>(supplier, accumulator, combiner, cs);
242 }
243
244 /**
245 * Returns a new {@code Collector} described by the given {@code supplier},
246 * {@code accumulator}, {@code combiner}, and {@code finisher} functions.
247 *
248 * @param supplier The supplier function for the new collector
249 * @param accumulator The accumulator function for the new collector
250 * @param combiner The combiner function for the new collector
251 * @param finisher The finisher function for the new collector
252 * @param characteristics The collector characteristics for the new
253 * collector
254 * @param <T> The type of input elements for the new collector
255 * @param <A> The intermediate accumulation type of the new collector
256 * @param <R> The final result type of the new collector
257 * @return the new {@code Collector}
258 */
259 public static<T, A, R> Collector<T, A, R> of(Supplier<A> supplier,
260 BiConsumer<A, T> accumulator,
261 BinaryOperator<A> combiner,
262 Function<A, R> finisher,
263 Characteristics... characteristics) {
264 Set<Characteristics> cs = Collectors.CH_NOID;
265 if (characteristics.length > 0) {
266 cs = EnumSet.noneOf(Characteristics.class);
267 Collections.addAll(cs, characteristics);
268 cs = Collections.unmodifiableSet(cs);
269 }
270 return new Collectors.CollectorImpl<>(supplier, accumulator, combiner, finisher, cs);
271 }
272
273 /**
274 * Characteristics indicating properties of a {@code Collector}, which can
275 * be used to optimize reduction implementations.
276 */
277 enum Characteristics {
278 /**
279 * Indicates that this collector is <em>concurrent</em>, meaning that
280 * the result container can support the accumulator function being
281 * called concurrently with the same result container from multiple
282 * threads.
283 *
284 * <p>If a {@code CONCURRENT} collector is not also {@code UNORDERED},
285 * then it should only be evaluated concurrently if applied to an
286 * unordered data source.
287 */
288 CONCURRENT,
289
290 /**
291 * Indicates that the result container has no intrinsic order, such as
292 * a {@link Set}.
293 */
294 UNORDERED,
295
296 /**
297 * Indicates that the finisher function is the identity function and
298 * can be elided. If set, it must be the case that an unchecked cast
299 * from A to R will succeed.
300 */
301 IDENTITY_FINISH
302 }
303 }