1 /*
2
3 dsh-compress Compression utility classes.
4 Copyright (c) 2014-2017 held jointly by the individual authors.
5
6 This library is free software; you can redistribute it and/or modify it
7 under the terms of the GNU Lesser General Public License as published
8 by the Free Software Foundation; either version 3 of the License, or (at
9 your option) any later version.
10
11 This library is distributed in the hope that it will be useful, but WITHOUT
12 ANY WARRANTY; with out even the implied warranty of MERCHANTABILITY or
13 FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public
14 License for more details.
15
16 You should have received a copy of the GNU Lesser General Public License
17 along with this library; if not, write to the Free Software Foundation,
18 Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA.
19
20 > http://www.fsf.org/licensing/licenses/lgpl.html
21 > http://www.opensource.org/licenses/lgpl-license.php
22
23 */
24 package org.dishevelled.compress;
25
26 import static com.google.common.base.Preconditions.checkNotNull;
27
28 import java.io.BufferedInputStream;
29 import java.io.File;
30 import java.io.FileInputStream;
31 import java.io.InputStream;
32 import java.io.IOException;
33
34 import javax.annotation.Nullable;
35
36 import com.google.common.io.Files;
37
38 import htsjdk.samtools.util.BlockCompressedInputStream;
39
40 import org.apache.commons.compress.compressors.bzip2.BZip2Utils;
41
42 import org.apache.commons.compress.compressors.gzip.GzipUtils;
43
44 /**
45 * Compression utility methods.
46 *
47 * @since 1.3
48 * @author Michael Heuer
49 */
50 public final class Compress
51 {
52
53 /**
54 * Private no-arg constructor.
55 */
56 private Compress()
57 {
58 // empty
59 }
60
61
62 /**
63 * Return true if the specified file is a block compressed gzip (BGZF) file name.
64 *
65 * @since 1.3
66 * @param fileName file name, must not be null
67 * @return true if the specified file is a block compressed gzip (BGZF) file name
68 */
69 public static boolean isBgzfFilename(final String fileName)
70 {
71 checkNotNull(fileName);
72 return BgzfUtils.isCompressedFilename(fileName);
73 }
74
75 /**
76 * Return true if the specified file is a block compressed gzip (BGZF) file.
77 *
78 * @param file file, if any
79 * @return true if the specified file is a block compressed gzip (BGZF) file
80 */
81 public static boolean isBgzfFile(@Nullable final File file)
82 {
83 if (file == null)
84 {
85 return false;
86 }
87 if (isBgzfFilename(file.getName()))
88 {
89 return true;
90 }
91 // block compressed gzip files may also have .gz file extension, check the stream
92 try (BufferedInputStream inputStream = new BufferedInputStream(new FileInputStream(file)))
93 {
94 return isBgzfInputStream(inputStream);
95 }
96 catch (IOException e)
97 {
98 return false;
99 }
100 }
101
102 /**
103 * Return true if the specified file is a block compressed gzip (BGZF) input stream.
104 *
105 * @param inputStream input stream, must not be null
106 * @return true if the specified file is a block compressed gzip (BGZF) input stream
107 */
108 public static boolean isBgzfInputStream(final InputStream inputStream)
109 {
110 checkNotNull(inputStream);
111 BufferedInputStream bufferedInputStream = inputStream instanceof BufferedInputStream ? (BufferedInputStream) inputStream : new BufferedInputStream(inputStream);
112 try
113 {
114 return BlockCompressedInputStream.isValidFile(bufferedInputStream);
115 }
116 catch (IOException e)
117 {
118 return false;
119 }
120 }
121
122 /**
123 * Return true if the specified file is a gzip file name.
124 *
125 * @since 1.3
126 * @param fileName file name, must not be null
127 * @return true if the specified file is a gzip file name
128 */
129 public static boolean isGzipFilename(final String fileName)
130 {
131 checkNotNull(fileName);
132 return GzipUtils.isCompressedFilename(fileName);
133 }
134
135 /**
136 * Return true if the specified file is a gzip file. Block compressed gzip (BGZF)
137 * files are also gzip files, so <code>isBgzfFile(File)</code> should be called before
138 * this method.
139 *
140 * @param file file, if any
141 * @return true if the specified file is a gzip file
142 */
143 public static boolean isGzipFile(@Nullable final File file)
144 {
145 if (file == null)
146 {
147 return false;
148 }
149 return isGzipFilename(file.getName());
150 }
151
152 /**
153 * Return true if the specified file is a gzip input stream.
154 *
155 * @since 1.3
156 * @param inputStream input stream, must not be null
157 * @return true if the specified file is a gzip input stream
158 */
159 public static boolean isGzipInputStream(final InputStream inputStream)
160 {
161 checkNotNull(inputStream);
162 InputStream in = inputStream.markSupported() ? inputStream : new BufferedInputStream(inputStream);
163 try
164 {
165 in.mark(2);
166 return in.read() == 31 && in.read() == 139;
167 }
168 catch (IOException e)
169 {
170 return false;
171 }
172 finally
173 {
174 try
175 {
176 in.reset();
177 }
178 catch (IOException e)
179 {
180 // ignore
181 }
182 }
183 }
184
185 /**
186 * Return true if the specified file is a bzip2 file name.
187 *
188 * @since 1.3
189 * @param fileName file name, must not be null
190 * @return true if the specified file is a bzip2 file name
191 */
192 public static boolean isBzip2Filename(final String fileName)
193 {
194 return BZip2Utils.isCompressedFilename(fileName);
195 }
196
197 /**
198 * Return true if the specified file is a bzip2 file.
199 *
200 * @param file file, if any
201 * @return true if the specified file is a bzip2 file
202 */
203 public static boolean isBzip2File(@Nullable final File file)
204 {
205 if (file == null)
206 {
207 return false;
208 }
209 return isBzip2Filename(file.getName());
210 }
211
212 /**
213 * Return true if the specified file is a bzip2 input stream.
214 *
215 * @since 1.3
216 * @param inputStream input stream, must not be null
217 * @return true if the specified file is a bzip2 input stream
218 */
219 public static boolean isBzip2InputStream(final InputStream inputStream)
220 {
221 checkNotNull(inputStream);
222 InputStream in = inputStream.markSupported() ? inputStream : new BufferedInputStream(inputStream);
223 try
224 {
225 in.mark(3);
226 return in.read() == 'B' && in.read() == 'Z' && in.read() == 'h';
227 }
228 catch (IOException e)
229 {
230 return false;
231 }
232 finally
233 {
234 try
235 {
236 in.reset();
237 }
238 catch (IOException e)
239 {
240 // ignore
241 }
242 }
243 }
244 }