001/*
002 * Licensed to the Apache Software Foundation (ASF) under one or more
003 * contributor license agreements.  See the NOTICE file distributed with
004 * this work for additional information regarding copyright ownership.
005 * The ASF licenses this file to You under the Apache License, Version 2.0
006 * (the "License"); you may not use this file except in compliance with
007 * the License.  You may obtain a copy of the License at
008 *
009 *      http://www.apache.org/licenses/LICENSE-2.0
010 *
011 * Unless required by applicable law or agreed to in writing, software
012 * distributed under the License is distributed on an "AS IS" BASIS,
013 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
014 * See the License for the specific language governing permissions and
015 * limitations under the License.
016 */
017package org.apache.commons.io.output;
018
019import java.io.File;
020import java.io.IOException;
021import java.io.InputStream;
022import java.io.OutputStream;
023import java.nio.file.Files;
024import java.nio.file.Path;
025import java.util.Objects;
026import java.util.function.Supplier;
027
028import org.apache.commons.io.build.AbstractStreamBuilder;
029import org.apache.commons.io.file.PathUtils;
030
031/**
032 * An output stream which will retain data in memory until a specified threshold is reached, and only then commit it to disk. If the stream is closed before the
033 * threshold is reached, the data will not be written to disk at all.
034 * <p>
035 * To build an instance, use {@link Builder}.
036 * </p>
037 * <p>
038 * The caller is responsible for deleting the output file ({@link #getFile()}, {@link #getPath()}) created by a DeferredFileOutputStream when the caller only
039 * configured a prefix.
040 * </p>
041 * <p>
042 * The caller is responsible for deleting the output file passed to a constructor or builder through {@link Builder#setOutputFile(File)} or
043 * {@link Builder#setOutputFile(Path)}.
044 * </p>
045 * <p>
046 * This class originated in FileUpload processing. In this use case, you do not know in advance the size of the file being uploaded. If the file is small you
047 * want to store it in memory (for speed), but if the file is large you want to store it to file (to avoid memory issues).
048 * </p>
049 *
050 * @see Builder
051 */
052public class DeferredFileOutputStream extends ThresholdingOutputStream {
053
054    // @formatter:off
055    /**
056     * Builds a new {@link DeferredFileOutputStream}.
057     * <p>
058     * For example:
059     * </p>
060     * <pre>{@code
061     * DeferredFileOutputStream s = DeferredFileOutputStream.builder()
062     *   .setBufferSize(4096)
063     *   .setDirectory(dir)
064     *   .setOutputFile(outputFile)
065     *   .setPrefix(prefix)
066     *   .setSuffix(suffix)
067     *   .setThreshold(threshold)
068     *   .get();}
069     * </pre>
070     * <p>
071     * The only super's aspect used is buffer size.
072     * </p>
073     *
074     * @see #get()
075     * @since 2.12.0
076     */
077    // @formatter:on
078    public static class Builder extends AbstractStreamBuilder<DeferredFileOutputStream, Builder> {
079
080        private int threshold;
081        private Path outputFile;
082        private String prefix;
083        private String suffix;
084        private Path directory;
085
086        /**
087         * Constructs a new builder.
088         */
089        public Builder() {
090            setBufferSizeDefault(AbstractByteArrayOutputStream.DEFAULT_SIZE);
091            setBufferSize(AbstractByteArrayOutputStream.DEFAULT_SIZE);
092        }
093
094        /**
095         * Builds a new {@link DeferredFileOutputStream}.
096         * <p>
097         * This builder use the following aspects:
098         * </p>
099         * <ul>
100         * <li>{@link #getBufferSize()}</li>
101         * <li>threshold</li>
102         * <li>outputFile</li>
103         * <li>prefix</li>
104         * <li>suffix</li>
105         * <li>directory</li>
106         * </ul>
107         *
108         * @return a new instance.
109         */
110        @Override
111        public DeferredFileOutputStream get() {
112            return new DeferredFileOutputStream(threshold, outputFile, prefix, suffix, directory, getBufferSize());
113        }
114
115        /**
116         * Sets the temporary file directory.
117         *
118         * @param directory Temporary file directory.
119         * @return {@code this} instance.
120         */
121        public Builder setDirectory(final File directory) {
122            this.directory = toPath(directory, null);
123            return this;
124        }
125
126        /**
127         * Sets the temporary file directory.
128         *
129         * @param directory Temporary file directory.
130         * @return {@code this} instance.
131         * @since 2.14.0
132         */
133        public Builder setDirectory(final Path directory) {
134            this.directory = toPath(directory, null);
135            return this;
136        }
137
138        /**
139         * Sets the file to which data is saved beyond the threshold.
140         *
141         * @param outputFile The file to which data is saved beyond the threshold.
142         * @return {@code this} instance.
143         */
144        public Builder setOutputFile(final File outputFile) {
145            this.outputFile = toPath(outputFile, null);
146            return this;
147        }
148
149        /**
150         * Sets the file to which data is saved beyond the threshold.
151         *
152         * @param outputFile The file to which data is saved beyond the threshold.
153         * @return {@code this} instance.
154         * @since 2.14.0
155         */
156        public Builder setOutputFile(final Path outputFile) {
157            this.outputFile = toPath(outputFile, null);
158            return this;
159        }
160
161        /**
162         * Sets the prefix to use for the temporary file.
163         *
164         * @param prefix Prefix to use for the temporary file.
165         * @return {@code this} instance.
166         */
167        public Builder setPrefix(final String prefix) {
168            this.prefix = prefix;
169            return this;
170        }
171
172        /**
173         * Sets the suffix to use for the temporary file.
174         *
175         * @param suffix Suffix to use for the temporary file.
176         * @return {@code this} instance.
177         */
178        public Builder setSuffix(final String suffix) {
179            this.suffix = suffix;
180            return this;
181        }
182
183        /**
184         * Sets the number of bytes at which to trigger an event.
185         *
186         * @param threshold The number of bytes at which to trigger an event.
187         * @return {@code this} instance.
188         */
189        public Builder setThreshold(final int threshold) {
190            this.threshold = threshold;
191            return this;
192        }
193
194    }
195
196    /**
197     * Constructs a new {@link Builder}.
198     *
199     * @return a new {@link Builder}.
200     * @since 2.12.0
201     */
202    public static Builder builder() {
203        return new Builder();
204    }
205
206    private static int checkBufferSize(final int initialBufferSize) {
207        if (initialBufferSize < 0) {
208            throw new IllegalArgumentException("Initial buffer size must be at least 0.");
209        }
210        return initialBufferSize;
211    }
212
213    private static Path toPath(final File file, final Supplier<Path> defaultPathSupplier) {
214        return file != null ? file.toPath() : defaultPathSupplier == null ? null : defaultPathSupplier.get();
215    }
216
217    private static Path toPath(final Path file, final Supplier<Path> defaultPathSupplier) {
218        return file != null ? file : defaultPathSupplier == null ? null : defaultPathSupplier.get();
219    }
220
221    /**
222     * The output stream to which data will be written prior to the threshold being reached.
223     */
224    private ByteArrayOutputStream memoryOutputStream;
225
226    /**
227     * The output stream to which data will be written at any given time. This will always be one of {@code memoryOutputStream} or {@code diskOutputStream}.
228     */
229    private OutputStream currentOutputStream;
230
231    /**
232     * The file to which output will be directed if the threshold is exceeded.
233     */
234    private Path outputPath;
235
236    /**
237     * The temporary file prefix.
238     */
239    private final String prefix;
240
241    /**
242     * The temporary file suffix.
243     */
244    private final String suffix;
245
246    /**
247     * The directory to use for temporary files.
248     */
249    private final Path directory;
250
251    /**
252     * True when close() has been called successfully.
253     */
254    private boolean closed;
255
256    /**
257     * Constructs an instance of this class which will trigger an event at the specified threshold, and save data to a file beyond that point. The initial
258     * buffer size will default to {@value AbstractByteArrayOutputStream#DEFAULT_SIZE} bytes which is ByteArrayOutputStream's default buffer size.
259     *
260     * @param threshold  The number of bytes at which to trigger an event.
261     * @param outputFile The file to which data is saved beyond the threshold.
262     * @deprecated Use {@link #builder()}, {@link Builder}, and {@link Builder#get()}
263     */
264    @Deprecated
265    public DeferredFileOutputStream(final int threshold, final File outputFile) {
266        this(threshold, outputFile, null, null, null, AbstractByteArrayOutputStream.DEFAULT_SIZE);
267    }
268
269    /**
270     * Constructs an instance of this class which will trigger an event at the specified threshold, and save data either to a file beyond that point.
271     *
272     * @param threshold         The number of bytes at which to trigger an event.
273     * @param outputFile        The file to which data is saved beyond the threshold.
274     * @param prefix            Prefix to use for the temporary file.
275     * @param suffix            Suffix to use for the temporary file.
276     * @param directory         Temporary file directory.
277     * @param initialBufferSize The initial size of the in memory buffer.
278     * @throws IllegalArgumentException if initialBufferSize &lt; 0.
279     */
280    private DeferredFileOutputStream(final int threshold, final File outputFile, final String prefix, final String suffix, final File directory,
281            final int initialBufferSize) {
282        super(threshold);
283        this.outputPath = toPath(outputFile, null);
284        this.prefix = prefix;
285        this.suffix = suffix;
286        this.directory = toPath(directory, PathUtils::getTempDirectory);
287        this.memoryOutputStream = new ByteArrayOutputStream(checkBufferSize(initialBufferSize));
288        this.currentOutputStream = memoryOutputStream;
289    }
290
291    /**
292     * Constructs an instance of this class which will trigger an event at the specified threshold, and save data to a file beyond that point.
293     *
294     * @param threshold         The number of bytes at which to trigger an event.
295     * @param initialBufferSize The initial size of the in memory buffer.
296     * @param outputFile        The file to which data is saved beyond the threshold.
297     * @since 2.5
298     * @deprecated Use {@link #builder()}, {@link Builder}, and {@link Builder#get()}
299     */
300    @Deprecated
301    public DeferredFileOutputStream(final int threshold, final int initialBufferSize, final File outputFile) {
302        this(threshold, outputFile, null, null, null, initialBufferSize);
303    }
304
305    /**
306     * Constructs an instance of this class which will trigger an event at the specified threshold, and save data to a temporary file beyond that point.
307     *
308     * @param threshold         The number of bytes at which to trigger an event.
309     * @param initialBufferSize The initial size of the in memory buffer.
310     * @param prefix            Prefix to use for the temporary file.
311     * @param suffix            Suffix to use for the temporary file.
312     * @param directory         Temporary file directory.
313     * @since 2.5
314     * @deprecated Use {@link #builder()}, {@link Builder}, and {@link Builder#get()}
315     */
316    @Deprecated
317    public DeferredFileOutputStream(final int threshold, final int initialBufferSize, final String prefix, final String suffix, final File directory) {
318        this(threshold, null, Objects.requireNonNull(prefix, "prefix"), suffix, directory, initialBufferSize);
319    }
320
321    /**
322     * Constructs an instance of this class which will trigger an event at the specified threshold, and save data either to a file beyond that point.
323     *
324     * @param threshold         The number of bytes at which to trigger an event.
325     * @param outputFile        The file to which data is saved beyond the threshold.
326     * @param prefix            Prefix to use for the temporary file.
327     * @param suffix            Suffix to use for the temporary file.
328     * @param directory         Temporary file directory.
329     * @param initialBufferSize The initial size of the in memory buffer.
330     * @throws IllegalArgumentException if initialBufferSize &lt; 0.
331     */
332    private DeferredFileOutputStream(final int threshold, final Path outputFile, final String prefix, final String suffix, final Path directory,
333            final int initialBufferSize) {
334        super(threshold);
335        this.outputPath = toPath(outputFile, null);
336        this.prefix = prefix;
337        this.suffix = suffix;
338        this.directory = toPath(directory, PathUtils::getTempDirectory);
339        this.memoryOutputStream = new ByteArrayOutputStream(checkBufferSize(initialBufferSize));
340        this.currentOutputStream = memoryOutputStream;
341    }
342
343    /**
344     * Constructs an instance of this class which will trigger an event at the specified threshold, and save data to a temporary file beyond that point. The
345     * initial buffer size will default to 32 bytes which is ByteArrayOutputStream's default buffer size.
346     *
347     * @param threshold The number of bytes at which to trigger an event.
348     * @param prefix    Prefix to use for the temporary file.
349     * @param suffix    Suffix to use for the temporary file.
350     * @param directory Temporary file directory.
351     * @since 1.4
352     * @deprecated Use {@link #builder()}, {@link Builder}, and {@link Builder#get()}
353     */
354    @Deprecated
355    public DeferredFileOutputStream(final int threshold, final String prefix, final String suffix, final File directory) {
356        this(threshold, null, Objects.requireNonNull(prefix, "prefix"), suffix, directory, AbstractByteArrayOutputStream.DEFAULT_SIZE);
357    }
358
359    /**
360     * Closes underlying output stream, and mark this as closed
361     *
362     * @throws IOException if an error occurs.
363     */
364    @Override
365    public void close() throws IOException {
366        super.close();
367        closed = true;
368    }
369
370    /**
371     * Gets the data for this output stream as an array of bytes, assuming that the data has been retained in memory. If the data was written to disk, this
372     * method returns {@code null}.
373     *
374     * @return The data for this output stream, or {@code null} if no such data is available.
375     */
376    public byte[] getData() {
377        return memoryOutputStream != null ? memoryOutputStream.toByteArray() : null;
378    }
379
380    /**
381     * Gets either the output File specified in the constructor or the temporary File created or null.
382     * <p>
383     * If the constructor specifying the File is used then it returns that same output File, even when threshold has not been reached.
384     * </p>
385     * <p>
386     * If constructor specifying a temporary File prefix/suffix is used then the temporary File created once the threshold is reached is returned if the
387     * threshold was not reached then {@code null} is returned.
388     * </p>
389     *
390     * @return The File for this output stream, or {@code null} if no such File exists.
391     */
392    public File getFile() {
393        return outputPath != null ? outputPath.toFile() : null;
394    }
395
396    /**
397     * Gets either the output Path specified in the constructor or the temporary Path created or null.
398     * <p>
399     * If the constructor specifying the file is used then it returns that same output Path, even when threshold has not been reached.
400     * </p>
401     * <p>
402     * If constructor specifying a temporary Path prefix/suffix is used then the temporary Path created once the threshold is reached is returned if the
403     * threshold was not reached then {@code null} is returned.
404     * </p>
405     *
406     * @return The Path for this output stream, or {@code null} if no such Path exists.
407     * @since 2.14.0
408     */
409    public Path getPath() {
410        return outputPath;
411    }
412
413    /**
414     * Gets the current output stream. This may be memory based or disk based, depending on the current state with respect to the threshold.
415     *
416     * @return The underlying output stream.
417     * @throws IOException if an error occurs.
418     */
419    @Override
420    protected OutputStream getStream() throws IOException {
421        return currentOutputStream;
422    }
423
424    /**
425     * Tests whether or not the data for this output stream has been retained in memory.
426     *
427     * @return {@code true} if the data is available in memory; {@code false} otherwise.
428     */
429    public boolean isInMemory() {
430        return !isThresholdExceeded();
431    }
432
433    /**
434     * Switches the underlying output stream from a memory based stream to one that is backed by disk. This is the point at which we realize that too much data
435     * is being written to keep in memory, so we elect to switch to disk-based storage.
436     *
437     * @throws IOException if an error occurs.
438     */
439    @Override
440    protected void thresholdReached() throws IOException {
441        if (prefix != null) {
442            outputPath = Files.createTempFile(directory, prefix, suffix);
443        }
444        PathUtils.createParentDirectories(outputPath, null, PathUtils.EMPTY_FILE_ATTRIBUTE_ARRAY);
445        final OutputStream fos = Files.newOutputStream(outputPath);
446        try {
447            memoryOutputStream.writeTo(fos);
448        } catch (final IOException e) {
449            fos.close();
450            throw e;
451        }
452        currentOutputStream = fos;
453        memoryOutputStream = null;
454    }
455
456    /**
457     * Converts the current contents of this byte stream to an {@link InputStream}. If the data for this output stream has been retained in memory, the returned
458     * stream is backed by buffers of {@code this} stream, avoiding memory allocation and copy, thus saving space and time.<br>
459     * Otherwise, the returned stream will be one that is created from the data that has been committed to disk.
460     *
461     * @return the current contents of this output stream.
462     * @throws IOException if this stream is not yet closed or an error occurs.
463     * @see org.apache.commons.io.output.ByteArrayOutputStream#toInputStream()
464     *
465     * @since 2.9.0
466     */
467    public InputStream toInputStream() throws IOException {
468        // we may only need to check if this is closed if we are working with a file
469        // but we should force the habit of closing whether we are working with
470        // a file or memory.
471        if (!closed) {
472            throw new IOException("Stream not closed");
473        }
474        if (isInMemory()) {
475            return memoryOutputStream.toInputStream();
476        }
477        return Files.newInputStream(outputPath);
478    }
479
480    /**
481     * Writes the data from this output stream to the specified output stream, after it has been closed.
482     *
483     * @param outputStream output stream to write to.
484     * @throws NullPointerException if the OutputStream is {@code null}.
485     * @throws IOException          if this stream is not yet closed or an error occurs.
486     */
487    public void writeTo(final OutputStream outputStream) throws IOException {
488        // we may only need to check if this is closed if we are working with a file
489        // but we should force the habit of closing whether we are working with
490        // a file or memory.
491        if (!closed) {
492            throw new IOException("Stream not closed");
493        }
494        if (isInMemory()) {
495            memoryOutputStream.writeTo(outputStream);
496        } else {
497            Files.copy(outputPath, outputStream);
498        }
499    }
500}