1 /*
2 * Licensed to the Apache Software Foundation (ASF) under one or more
3 * contributor license agreements. See the NOTICE file distributed with
4 * this work for additional information regarding copyright ownership.
5 * The ASF licenses this file to You under the Apache License, Version 2.0
6 * (the "License"); you may not use this file except in compliance with
7 * the License. You may obtain a copy of the License at
8 *
9 * http://www.apache.org/licenses/LICENSE-2.0
10 *
11 * Unless required by applicable law or agreed to in writing, software
12 * distributed under the License is distributed on an "AS IS" BASIS,
13 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
14 * See the License for the specific language governing permissions and
15 * limitations under the License.
16 */
17 package org.apache.commons.fileupload.disk;
18
19 import java.io.ByteArrayInputStream;
20 import java.io.File;
21 import java.io.FileOutputStream;
22 import java.io.IOException;
23 import java.io.InputStream;
24 import java.io.OutputStream;
25 import java.io.UncheckedIOException;
26 import java.io.UnsupportedEncodingException;
27 import java.nio.file.Files;
28 import java.util.Map;
29 import java.util.UUID;
30 import java.util.concurrent.atomic.AtomicInteger;
31
32 import org.apache.commons.fileupload.FileItem;
33 import org.apache.commons.fileupload.FileItemHeaders;
34 import org.apache.commons.fileupload.FileUploadException;
35 import org.apache.commons.fileupload.ParameterParser;
36 import org.apache.commons.fileupload.util.Streams;
37 import org.apache.commons.io.FileUtils;
38 import org.apache.commons.io.function.Uncheck;
39 import org.apache.commons.io.output.DeferredFileOutputStream;
40
41 /**
42 * The default implementation of the {@link org.apache.commons.fileupload.FileItem FileItem} interface.
43 *
44 * <p>
45 * After retrieving an instance of this class from a {@link DiskFileItemFactory} instance (see {@link org.apache.commons.fileupload.servlet.ServletFileUpload
46 * #parseRequest(javax.servlet.http.HttpServletRequest)}), you may either request all contents of file at once using {@link #get()} or request an
47 * {@link java.io.InputStream InputStream} with {@link #getInputStream()} and process the file without attempting to load it into memory, which may come handy
48 * with large files.
49 * </p>
50 * <p>
51 * Temporary files, which are created for file items, should be deleted later on. The best way to do this is using a
52 * {@link org.apache.commons.io.FileCleaningTracker}, which you can set on the {@link DiskFileItemFactory}. However, if you do use such a tracker, then you must
53 * consider the following: Temporary files are automatically deleted as soon as they are no longer needed. (More precisely, when the corresponding instance of
54 * {@link java.io.File} is garbage collected.) This is done by the so-called reaper thread, which is started and stopped automatically by the
55 * {@link org.apache.commons.io.FileCleaningTracker} when there are files to be tracked. It might make sense to terminate that thread, for example, if your web
56 * application ends. See the section on "Resource cleanup" in the users guide of commons-fileupload.
57 * </p>
58 *
59 * @since FileUpload 1.1
60 */
61 public class DiskFileItem implements FileItem {
62
63 /**
64 * Counter used in unique identifier generation.
65 */
66 private static final AtomicInteger COUNTER = new AtomicInteger(0);
67
68 /**
69 * Default content charset to be used when no explicit charset
70 * parameter is provided by the sender. Media subtypes of the
71 * "text" type are defined to have a default charset value of
72 * "ISO-8859-1" when received via HTTP.
73 */
74 public static final String DEFAULT_CHARSET = "ISO-8859-1";
75
76 /**
77 * UID used in unique file name generation.
78 */
79 private static final String UID = UUID.randomUUID().toString().replace('-', '_');
80
81 /**
82 * Returns an identifier that is unique within the class loader used to
83 * load this class, but does not have random-like appearance.
84 *
85 * @return A String with the non-random looking instance identifier.
86 */
87 private static String getUniqueId() {
88 final int limit = 100000000;
89 final int current = COUNTER.getAndIncrement();
90 String id = Integer.toString(current);
91
92 // If you manage to get more than 100 million of ids, you'll
93 // start getting ids longer than 8 characters.
94 if (current < limit) {
95 id = ("00000000" + id).substring(id.length());
96 }
97 return id;
98 }
99
100 /**
101 * Cached contents of the file.
102 */
103 private byte[] cachedContent;
104
105 /**
106 * The content type passed by the browser, or {@code null} if not defined.
107 */
108 private final String contentType;
109
110 /**
111 * Default content charset to be used when no explicit charset parameter is provided by the sender.
112 */
113 private String defaultCharset = DEFAULT_CHARSET;
114
115 /**
116 * Output stream for this item.
117 */
118 private transient DeferredFileOutputStream dfos;
119
120 /**
121 * The name of the form field as provided by the browser.
122 */
123 private String fieldName;
124
125 /**
126 * The original file name in the user's file system.
127 */
128 private final String fileName;
129
130 /**
131 * The file items headers.
132 */
133 private FileItemHeaders headers;
134
135 /**
136 * Whether or not this item is a simple form field.
137 */
138 private boolean formField;
139
140 /**
141 * The directory in which uploaded files will be stored, if stored on disk.
142 */
143 private final File repository;
144
145 /**
146 * The size of the item, in bytes. This is used to cache the size when a
147 * file item is moved from its original location.
148 */
149 private long size = -1;
150
151 /**
152 * The threshold above which uploads will be stored on disk.
153 */
154 private final int sizeThreshold;
155
156 /**
157 * The temporary file to use.
158 */
159 private transient File tempFile;
160
161 /**
162 * Constructs a new {@code DiskFileItem} instance.
163 *
164 * @param fieldName The name of the form field.
165 * @param contentType The content type passed by the browser or {@code null} if not specified.
166 * @param isFormField Whether or not this item is a plain form field, as opposed to a file upload.
167 * @param fileName The original file name in the user's file system, or {@code null} if not specified.
168 * @param sizeThreshold The threshold, in bytes, below which items will be retained in memory and above which they
169 * will be stored as a file.
170 * @param repository The data repository, which is the directory in which files will be created, should the item
171 * size exceed the threshold.
172 */
173 public DiskFileItem(final String fieldName, final String contentType, final boolean isFormField,
174 final String fileName, final int sizeThreshold, final File repository) {
175 this.fieldName = fieldName;
176 this.contentType = contentType;
177 this.formField = isFormField;
178 this.fileName = fileName;
179 this.sizeThreshold = sizeThreshold;
180 this.repository = repository;
181 }
182
183 /**
184 * Clears the cache.
185 */
186 private void clear() {
187 cachedContent = null; // NOPMD
188 }
189
190 /**
191 * Deletes the underlying storage for a file item, including deleting any
192 * associated temporary disk file. Although this storage will be deleted
193 * automatically when the {@code FileItem} instance is garbage
194 * collected, this method can be used to ensure that this is done at an
195 * earlier time, thus preserving system resources.
196 */
197 @Override
198 public void delete() {
199 clear();
200 final File outputFile = getStoreLocation();
201 if (outputFile != null && !isInMemory() && outputFile.exists()) {
202 outputFile.delete();
203 }
204 }
205
206 /**
207 * Removes the file contents from the temporary storage.
208 *
209 * @throws Throwable Thrown by {@link Object#finalize()}.
210 */
211 @Override
212 protected void finalize() throws Throwable {
213 if (dfos == null || dfos.isInMemory()) {
214 return;
215 }
216 final File outputFile = dfos.getFile();
217 if (outputFile != null && outputFile.exists()) {
218 outputFile.delete();
219 }
220 super.finalize();
221 }
222
223 /**
224 * Gets the contents of the file as an array of bytes. If the contents of the file were not yet cached in memory, they will be loaded from the disk storage
225 * and cached.
226 *
227 * @return The contents of the file as an array of bytes or {@code null} if the data cannot be read.
228 * @throws UncheckedIOException if an I/O error occurs.
229 * @throws OutOfMemoryError if an array of the required size cannot be allocated, for example the file is larger that {@code 2GB}.
230 */
231 @Override
232 public byte[] get() {
233 if (isInMemory()) {
234 if (cachedContent == null && dfos != null) {
235 cachedContent = dfos.getData();
236 }
237 return cachedContent != null ? cachedContent.clone() : new byte[0];
238 }
239 return Uncheck.get(() -> Files.readAllBytes(dfos.getFile().toPath()));
240 }
241
242 /**
243 * Gets the content charset passed by the agent or {@code null} if not defined.
244 *
245 * @return The content charset passed by the agent or {@code null} if not defined.
246 */
247 public String getCharSet() {
248 final ParameterParser parser = new ParameterParser();
249 parser.setLowerCaseNames(true);
250 // Parameter parser can handle null input
251 final Map<String, String> params = parser.parse(getContentType(), ';');
252 return params.get("charset");
253 }
254
255 /**
256 * Gets the content type passed by the agent or {@code null} if not defined.
257 *
258 * @return The content type passed by the agent or {@code null} if not defined.
259 */
260 @Override
261 public String getContentType() {
262 return contentType;
263 }
264
265 /**
266 * Gets the default charset for use when no explicit charset parameter is provided by the sender.
267 *
268 * @return the default charset
269 */
270 public String getDefaultCharset() {
271 return defaultCharset;
272 }
273
274 /**
275 * Gets the name of the field in the multipart form corresponding to this file item.
276 *
277 * @return The name of the form field.
278 * @see #setFieldName(String)
279 */
280 @Override
281 public String getFieldName() {
282 return fieldName;
283 }
284
285 /**
286 * Gets the file item headers.
287 *
288 * @return The file items headers.
289 */
290 @Override
291 public FileItemHeaders getHeaders() {
292 return headers;
293 }
294
295 /**
296 * Gets an {@link java.io.InputStream InputStream} that can be used to retrieve the contents of the file.
297 *
298 * @return An {@link java.io.InputStream InputStream} that can be used to retrieve the contents of the file.
299 *
300 * @throws IOException if an error occurs.
301 */
302 @Override
303 public InputStream getInputStream() throws IOException {
304 if (!isInMemory()) {
305 return Files.newInputStream(dfos.getPath());
306 }
307 if (cachedContent == null) {
308 cachedContent = dfos.getData();
309 }
310 return new ByteArrayInputStream(cachedContent);
311 }
312
313 /**
314 * Gets the original file name in the client's file system.
315 *
316 * @return The original file name in the client's file system.
317 * @throws org.apache.commons.fileupload.InvalidFileNameException The file name contains a NUL character,
318 * which might be an indicator of a security attack. If you intend to
319 * use the file name anyways, catch the exception and use
320 * {@link org.apache.commons.fileupload.InvalidFileNameException#getName()}.
321 */
322 @Override
323 public String getName() {
324 return Streams.checkFileName(fileName);
325 }
326
327 /**
328 * Gets an {@link java.io.OutputStream OutputStream} that can be used for storing the contents of the file.
329 *
330 * @return An {@link java.io.OutputStream OutputStream} that can be used for storing the contents of the file.
331 *
332 * @throws IOException if an error occurs (never happens).
333 */
334 @Override
335 public OutputStream getOutputStream() throws IOException {
336 if (dfos == null) {
337 final File outputFile = getTempFile();
338 // @formatter:off
339 dfos = DeferredFileOutputStream.builder()
340 .setThreshold(sizeThreshold)
341 .setOutputFile(outputFile)
342 .get();
343 // @formatter:on
344 }
345 return dfos;
346 }
347
348 /**
349 * Gets the size of the file.
350 *
351 * @return The size of the file, in bytes.
352 */
353 @Override
354 public long getSize() {
355 if (size >= 0) {
356 return size;
357 }
358 if (cachedContent != null) {
359 return cachedContent.length;
360 }
361 if (dfos.isInMemory()) {
362 return dfos.getData().length;
363 }
364 return dfos.getFile().length();
365 }
366
367 /**
368 * Gets the {@link java.io.File} object for the {@code FileItem}'s
369 * data's temporary location on the disk. Note that for
370 * {@code FileItem}s that have their data stored in memory,
371 * this method will return {@code null}. When handling large
372 * files, you can use {@link java.io.File#renameTo(java.io.File)} to
373 * move the file to new location without copying the data, if the
374 * source and destination locations reside within the same logical
375 * volume.
376 *
377 * @return The data file, or {@code null} if the data is stored in memory.
378 */
379 public File getStoreLocation() {
380 if (dfos == null) {
381 return null;
382 }
383 if (isInMemory()) {
384 return null;
385 }
386 return dfos.getFile();
387 }
388
389 /**
390 * Gets the contents of the file as a String, using the default character encoding. This method uses
391 * {@link #get()} to retrieve the contents of the file.
392 * <p>
393 * <strong>TODO</strong> Consider making this method throw UnsupportedEncodingException.
394 *
395 * @return The contents of the file, as a string.
396 */
397 @Override
398 public String getString() {
399 final byte[] rawData = get();
400 String charset = getCharSet();
401 if (charset == null) {
402 charset = defaultCharset;
403 }
404 try {
405 return new String(rawData, charset);
406 } catch (final UnsupportedEncodingException e) {
407 return "";
408 }
409 }
410
411 /**
412 * Gets the contents of the file as a String, using the specified encoding. This method uses {@link #get()} to
413 * retrieve the contents of the file.
414 *
415 * @param charset The charset to use.
416 * @return The contents of the file, as a string.
417 * @throws UnsupportedEncodingException if the requested character encoding is not available.
418 */
419 @Override
420 public String getString(final String charset) throws UnsupportedEncodingException {
421 return new String(get(), charset);
422 }
423
424 /**
425 * Creates and returns a {@link java.io.File File} representing a uniquely named temporary file in the configured
426 * repository path. The lifetime of the file is tied to the lifetime of the {@code FileItem} instance; the file will
427 * be deleted when the instance is garbage collected.
428 * <p>
429 * <b>Note: Subclasses that override this method must ensure that they return the same File each time.</b>
430 *
431 * @return The {@link java.io.File File} to be used for temporary storage.
432 */
433 protected File getTempFile() {
434 if (tempFile == null) {
435 File tempDir = repository;
436 if (tempDir == null) {
437 tempDir = new File(System.getProperty("java.io.tmpdir"));
438 }
439
440 final String tempFileName = String.format("upload_%s_%s.tmp", UID, getUniqueId());
441
442 tempFile = new File(tempDir, tempFileName);
443 }
444 return tempFile;
445 }
446
447 /**
448 * Tests whether or not a {@code FileItem} instance represents a simple form field.
449 *
450 * @return {@code true} if the instance represents a simple form field; {@code false} if it represents an uploaded
451 * file.
452 *
453 * @see #setFormField(boolean)
454 */
455 @Override
456 public boolean isFormField() {
457 return formField;
458 }
459
460 /**
461 * Provides a hint as to whether or not the file contents will be read from memory.
462 *
463 * @return {@code true} if the file contents will be read from memory; {@code false} otherwise.
464 */
465 @Override
466 public boolean isInMemory() {
467 if (cachedContent != null) {
468 return true;
469 }
470 return dfos.isInMemory();
471 }
472
473 /**
474 * Sets the default charset for use when no explicit charset parameter is provided by the sender.
475 *
476 * @param charset the default charset
477 */
478 public void setDefaultCharset(final String charset) {
479 defaultCharset = charset;
480 }
481
482 /**
483 * Sets the field name used to reference this file item.
484 *
485 * @param fieldName The name of the form field.
486 * @see #getFieldName()
487 */
488 @Override
489 public void setFieldName(final String fieldName) {
490 this.fieldName = fieldName;
491 }
492
493 /**
494 * Sets whether or not a {@code FileItem} instance represents a simple form field.
495 *
496 * @param formField {@code true} if the instance represents a simple form
497 * field; {@code false} if it represents an uploaded file.
498 *
499 * @see #isFormField()
500 */
501 @Override
502 public void setFormField(final boolean formField) {
503 this.formField = formField;
504 }
505
506 /**
507 * Sets the file item headers.
508 *
509 * @param headers The file items headers.
510 */
511 @Override
512 public void setHeaders(final FileItemHeaders headers) {
513 this.headers = headers;
514 }
515
516 /**
517 * Returns a string representation of this object.
518 *
519 * @return a string representation of this object.
520 */
521 @Override
522 public String toString() {
523 return String.format("name=%s, StoreLocation=%s, size=%s bytes, isFormField=%s, FieldName=%s",
524 getName(), getStoreLocation(), Long.valueOf(getSize()), Boolean.valueOf(isFormField()), getFieldName());
525 }
526
527 /**
528 * A convenience method to write an uploaded item to disk. The client code
529 * is not concerned with whether or not the item is stored in memory, or on
530 * disk in a temporary location. They just want to write the uploaded item
531 * to a file.
532 * <p>
533 * This implementation first attempts to rename the uploaded item to the
534 * specified destination file, if the item was originally written to disk.
535 * Otherwise, the data will be copied to the specified file.
536 * </p>
537 * <p>
538 * This method is only guaranteed to work <em>once</em>, the first time it
539 * is invoked for a particular item. This is because, in the event that the
540 * method renames a temporary file, that file will no longer be available
541 * to copy or rename again at a later time.
542 * </p>
543 *
544 * @param file The {@code File} into which the uploaded item should
545 * be stored.
546 * @throws Exception if an error occurs.
547 */
548 @Override
549 public void write(final File file) throws Exception {
550 if (isInMemory()) {
551 try (FileOutputStream fout = new FileOutputStream(file);) {
552 fout.write(get());
553 } catch (final IOException ignore) {
554 // ignore
555 }
556 } else {
557 final File outputFile = getStoreLocation();
558 if (outputFile == null) {
559 /*
560 * For whatever reason we cannot write the file to disk.
561 */
562 throw new FileUploadException("Cannot write uploaded file to disk!");
563 }
564 // Save the length of the file
565 size = outputFile.length();
566 /*
567 * The uploaded file is being stored on disk in a temporary location so move it to the desired file.
568 */
569 if (file.exists() && !file.delete()) {
570 throw new FileUploadException("Cannot write uploaded file to disk!");
571 }
572 FileUtils.moveFile(outputFile, file);
573 }
574 }
575 }