View Javadoc
1   /*
2    * Licensed to the Apache Software Foundation (ASF) under one
3    * or more contributor license agreements.  See the NOTICE file
4    * distributed with this work for additional information
5    * regarding copyright ownership.  The ASF licenses this file
6    * to you under the Apache License, Version 2.0 (the
7    * "License"); you may not use this file except in compliance
8    * with the License.  You may obtain a copy of the License at
9    *
10   * http://www.apache.org/licenses/LICENSE-2.0
11   *
12   * Unless required by applicable law or agreed to in writing,
13   * software distributed under the License is distributed on an
14   * "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
15   * KIND, either express or implied.  See the License for the
16   * specific language governing permissions and limitations
17   * under the License.
18   */
19  package org.apache.commons.compress.archivers.ar;
20  
21  import java.io.EOFException;
22  import java.io.IOException;
23  import java.io.InputStream;
24  import java.nio.charset.StandardCharsets;
25  import java.util.Arrays;
26  import java.util.regex.Pattern;
27  
28  import org.apache.commons.compress.archivers.ArchiveInputStream;
29  import org.apache.commons.compress.utils.ArchiveUtils;
30  import org.apache.commons.compress.utils.IOUtils;
31  import org.apache.commons.compress.utils.ParsingUtils;
32  
33  /**
34   * Implements the "ar" archive format as an input stream.
35   *
36   * @NotThreadSafe
37   */
38  public class ArArchiveInputStream extends ArchiveInputStream<ArArchiveEntry> {
39  
40      // offsets and length of meta data parts
41      private static final int NAME_OFFSET = 0;
42      private static final int NAME_LEN = 16;
43      private static final int LAST_MODIFIED_OFFSET = NAME_LEN;
44  
45      private static final int LAST_MODIFIED_LEN = 12;
46  
47      private static final int USER_ID_OFFSET = LAST_MODIFIED_OFFSET + LAST_MODIFIED_LEN;
48  
49      private static final int USER_ID_LEN = 6;
50  
51      private static final int GROUP_ID_OFFSET = USER_ID_OFFSET + USER_ID_LEN;
52      private static final int GROUP_ID_LEN = 6;
53      private static final int FILE_MODE_OFFSET = GROUP_ID_OFFSET + GROUP_ID_LEN;
54      private static final int FILE_MODE_LEN = 8;
55      private static final int LENGTH_OFFSET = FILE_MODE_OFFSET + FILE_MODE_LEN;
56      private static final int LENGTH_LEN = 10;
57      static final String BSD_LONGNAME_PREFIX = "#1/";
58      private static final int BSD_LONGNAME_PREFIX_LEN = BSD_LONGNAME_PREFIX.length();
59      private static final Pattern BSD_LONGNAME_PATTERN = Pattern.compile("^" + BSD_LONGNAME_PREFIX + "\\d+");
60      private static final String GNU_STRING_TABLE_NAME = "//";
61      private static final Pattern GNU_LONGNAME_PATTERN = Pattern.compile("^/\\d+");
62  
63      /**
64       * Does the name look like it is a long name (or a name containing spaces) as encoded by BSD ar?
65       * <p>
66       * From the FreeBSD ar(5) man page:
67       * </p>
68       * <pre>
69       * BSD   In the BSD variant, names that are shorter than 16
70       *       characters and without embedded spaces are stored
71       *       directly in this field.  If a name has an embedded
72       *       space, or if it is longer than 16 characters, then
73       *       the string "#1/" followed by the decimal represen-
74       *       tation of the length of the file name is placed in
75       *       this field. The actual file name is stored immedi-
76       *       ately after the archive header.  The content of the
77       *       archive member follows the file name.  The ar_size
78       *       field of the header (see below) will then hold the
79       *       sum of the size of the file name and the size of
80       *       the member.
81       * </pre>
82       *
83       * @since 1.3
84       */
85      private static boolean isBSDLongName(final String name) {
86          return name != null && BSD_LONGNAME_PATTERN.matcher(name).matches();
87      }
88  
89      /**
90       * Is this the name of the "Archive String Table" as used by SVR4/GNU to store long file names?
91       * <p>
92       * GNU ar stores multiple extended file names in the data section of a file with the name "//", this record is referred to by future headers.
93       * </p>
94       * <p>
95       * A header references an extended file name by storing a "/" followed by a decimal offset to the start of the file name in the extended file name data
96       * section.
97       * </p>
98       * <p>
99       * The format of the "//" file itself is simply a list of the long file names, each separated by one or more LF characters. Note that the decimal offsets
100      * are number of characters, not line or string number within the "//" file.
101      * </p>
102      */
103     private static boolean isGNUStringTable(final String name) {
104         return GNU_STRING_TABLE_NAME.equals(name);
105     }
106 
107     /**
108      * Checks if the signature matches ASCII "!&lt;arch&gt;" followed by a single LF control character
109      *
110      * @param signature the bytes to check
111      * @param length    the number of bytes to check
112      * @return true, if this stream is an Ar archive stream, false otherwise
113      */
114     public static boolean matches(final byte[] signature, final int length) {
115         // 3c21 7261 6863 0a3e
116 
117         return length >= 8 && signature[0] == 0x21 && signature[1] == 0x3c && signature[2] == 0x61 && signature[3] == 0x72 && signature[4] == 0x63
118                 && signature[5] == 0x68 && signature[6] == 0x3e && signature[7] == 0x0a;
119     }
120 
121     private long offset;
122 
123     private boolean closed;
124 
125     /*
126      * If getNextEntry has been called, the entry metadata is stored in currentEntry.
127      */
128     private ArArchiveEntry currentEntry;
129 
130     /** Storage area for extra long names (GNU ar). */
131     private byte[] namebuffer;
132 
133     /**
134      * The offset where the current entry started. -1 if no entry has been called
135      */
136     private long entryOffset = -1;
137 
138     /** Cached buffer for meta data - must only be used locally in the class (COMPRESS-172 - reduce garbage collection). */
139     private final byte[] metaData = new byte[NAME_LEN + LAST_MODIFIED_LEN + USER_ID_LEN + GROUP_ID_LEN + FILE_MODE_LEN + LENGTH_LEN];
140 
141     /**
142      * Constructs an Ar input stream with the referenced stream
143      *
144      * @param inputStream the ar input stream
145      */
146     public ArArchiveInputStream(final InputStream inputStream) {
147         super(inputStream, StandardCharsets.US_ASCII.name());
148     }
149 
150     private int asInt(final byte[] byteArray, final int offset, final int len) throws IOException {
151         return asInt(byteArray, offset, len, 10, false);
152     }
153 
154     private int asInt(final byte[] byteArray, final int offset, final int len, final boolean treatBlankAsZero) throws IOException {
155         return asInt(byteArray, offset, len, 10, treatBlankAsZero);
156     }
157 
158     private int asInt(final byte[] byteArray, final int offset, final int len, final int base) throws IOException {
159         return asInt(byteArray, offset, len, base, false);
160     }
161 
162     private int asInt(final byte[] byteArray, final int offset, final int len, final int base, final boolean treatBlankAsZero) throws IOException {
163         final String string = ArchiveUtils.toAsciiString(byteArray, offset, len).trim();
164         if (string.isEmpty() && treatBlankAsZero) {
165             return 0;
166         }
167         return ParsingUtils.parseIntValue(string, base);
168     }
169 
170     private long asLong(final byte[] byteArray, final int offset, final int len) throws IOException {
171         return ParsingUtils.parseLongValue(ArchiveUtils.toAsciiString(byteArray, offset, len).trim());
172     }
173 
174     /*
175      * (non-Javadoc)
176      *
177      * @see java.io.InputStream#close()
178      */
179     @Override
180     public void close() throws IOException {
181         if (!closed) {
182             closed = true;
183             in.close();
184         }
185         currentEntry = null;
186     }
187 
188     /**
189      * Reads the real name from the current stream assuming the very first bytes to be read are the real file name.
190      *
191      * @see #isBSDLongName
192      *
193      * @since 1.3
194      */
195     private String getBSDLongName(final String bsdLongName) throws IOException {
196         final int nameLen = ParsingUtils.parseIntValue(bsdLongName.substring(BSD_LONGNAME_PREFIX_LEN));
197         final byte[] name = IOUtils.readRange(in, nameLen);
198         final int read = name.length;
199         trackReadBytes(read);
200         if (read != nameLen) {
201             throw new EOFException();
202         }
203         return ArchiveUtils.toAsciiString(name);
204     }
205 
206     /**
207      * Gets an extended name from the GNU extended name buffer.
208      *
209      * @param offset pointer to entry within the buffer
210      * @return the extended file name; without trailing "/" if present.
211      * @throws IOException if name not found or buffer not set up
212      */
213     private String getExtendedName(final int offset) throws IOException {
214         if (namebuffer == null) {
215             throw new IOException("Cannot process GNU long file name as no // record was found");
216         }
217         for (int i = offset; i < namebuffer.length; i++) {
218             if (namebuffer[i] == '\012' || namebuffer[i] == 0) {
219                 // Avoid array errors
220                 if (i == 0) {
221                     break;
222                 }
223                 if (namebuffer[i - 1] == '/') {
224                     i--; // drop trailing /
225                 }
226 
227                 // Check there is a something to return, otherwise break out of the loop
228                 if (i - offset > 0) {
229                     return ArchiveUtils.toAsciiString(namebuffer, offset, i - offset);
230                 }
231                 break;
232             }
233         }
234         throw new IOException("Failed to read entry: " + offset);
235     }
236 
237     /**
238      * Returns the next AR entry in this stream.
239      *
240      * @return the next AR entry.
241      * @throws IOException if the entry could not be read
242      * @deprecated Use {@link #getNextEntry()}.
243      */
244     @Deprecated
245     public ArArchiveEntry getNextArEntry() throws IOException {
246         if (currentEntry != null) {
247             final long entryEnd = entryOffset + currentEntry.getLength();
248             final long skipped = org.apache.commons.io.IOUtils.skip(in, entryEnd - offset);
249             trackReadBytes(skipped);
250             currentEntry = null;
251         }
252 
253         if (offset == 0) {
254             final byte[] expected = ArchiveUtils.toAsciiBytes(ArArchiveEntry.HEADER);
255             final byte[] realized = IOUtils.readRange(in, expected.length);
256             final int read = realized.length;
257             trackReadBytes(read);
258             if (read != expected.length) {
259                 throw new IOException("Failed to read header. Occurred at byte: " + getBytesRead());
260             }
261             if (!Arrays.equals(expected, realized)) {
262                 throw new IOException("Invalid header " + ArchiveUtils.toAsciiString(realized));
263             }
264         }
265 
266         if (offset % 2 != 0) {
267             if (in.read() < 0) {
268                 // hit eof
269                 return null;
270             }
271             trackReadBytes(1);
272         }
273 
274         {
275             final int read = IOUtils.readFully(in, metaData);
276             trackReadBytes(read);
277             if (read == 0) {
278                 return null;
279             }
280             if (read < metaData.length) {
281                 throw new IOException("Truncated ar archive");
282             }
283         }
284 
285         {
286             final byte[] expected = ArchiveUtils.toAsciiBytes(ArArchiveEntry.TRAILER);
287             final byte[] realized = IOUtils.readRange(in, expected.length);
288             final int read = realized.length;
289             trackReadBytes(read);
290             if (read != expected.length) {
291                 throw new IOException("Failed to read entry trailer. Occurred at byte: " + getBytesRead());
292             }
293             if (!Arrays.equals(expected, realized)) {
294                 throw new IOException("Invalid entry trailer. not read the content? Occurred at byte: " + getBytesRead());
295             }
296         }
297 
298         entryOffset = offset;
299 
300 //        GNU ar uses a '/' to mark the end of the file name; this allows for the use of spaces without the use of an extended file name.
301 
302         // entry name is stored as ASCII string
303         String temp = ArchiveUtils.toAsciiString(metaData, NAME_OFFSET, NAME_LEN).trim();
304         if (isGNUStringTable(temp)) { // GNU extended file names entry
305             currentEntry = readGNUStringTable(metaData, LENGTH_OFFSET, LENGTH_LEN);
306             return getNextArEntry();
307         }
308 
309         long len;
310         try {
311             len = asLong(metaData, LENGTH_OFFSET, LENGTH_LEN);
312         } catch (final NumberFormatException ex) {
313             throw new IOException("Broken archive, unable to parse ar_size field as a number", ex);
314         }
315         if (temp.endsWith("/")) { // GNU terminator
316             temp = temp.substring(0, temp.length() - 1);
317         } else if (isGNULongName(temp)) {
318             final int off = ParsingUtils.parseIntValue(temp.substring(1));// get the offset
319             temp = getExtendedName(off); // convert to the long name
320         } else if (isBSDLongName(temp)) {
321             temp = getBSDLongName(temp);
322             // entry length contained the length of the file name in
323             // addition to the real length of the entry.
324             // assume file name was ASCII, there is no "standard" otherwise
325             final int nameLen = temp.length();
326             len -= nameLen;
327             entryOffset += nameLen;
328         }
329 
330         if (len < 0) {
331             throw new IOException("broken archive, entry with negative size");
332         }
333 
334         try {
335             currentEntry = new ArArchiveEntry(temp, len, asInt(metaData, USER_ID_OFFSET, USER_ID_LEN, true),
336                     asInt(metaData, GROUP_ID_OFFSET, GROUP_ID_LEN, true), asInt(metaData, FILE_MODE_OFFSET, FILE_MODE_LEN, 8),
337                     asLong(metaData, LAST_MODIFIED_OFFSET, LAST_MODIFIED_LEN));
338             return currentEntry;
339         } catch (final NumberFormatException ex) {
340             throw new IOException("Broken archive, unable to parse entry metadata fields as numbers", ex);
341         }
342     }
343 
344     /*
345      * (non-Javadoc)
346      *
347      * @see org.apache.commons.compress.archivers.ArchiveInputStream#getNextEntry()
348      */
349     @Override
350     public ArArchiveEntry getNextEntry() throws IOException {
351         return getNextArEntry();
352     }
353 
354     /**
355      * Does the name look like it is a long name (or a name containing spaces) as encoded by SVR4/GNU ar?
356      *
357      * @see #isGNUStringTable
358      */
359     private boolean isGNULongName(final String name) {
360         return name != null && GNU_LONGNAME_PATTERN.matcher(name).matches();
361     }
362 
363     /*
364      * (non-Javadoc)
365      *
366      * @see java.io.InputStream#read(byte[], int, int)
367      */
368     @Override
369     public int read(final byte[] b, final int off, final int len) throws IOException {
370         if (len == 0) {
371             return 0;
372         }
373         if (currentEntry == null) {
374             throw new IllegalStateException("No current ar entry");
375         }
376         final long entryEnd = entryOffset + currentEntry.getLength();
377         if (len < 0 || offset >= entryEnd) {
378             return -1;
379         }
380         final int toRead = (int) Math.min(len, entryEnd - offset);
381         final int ret = this.in.read(b, off, toRead);
382         trackReadBytes(ret);
383         return ret;
384     }
385 
386     /**
387      * Reads the GNU archive String Table.
388      *
389      * @see #isGNUStringTable
390      */
391     private ArArchiveEntry readGNUStringTable(final byte[] length, final int offset, final int len) throws IOException {
392         int bufflen;
393         try {
394             bufflen = asInt(length, offset, len); // Assume length will fit in an int
395         } catch (final NumberFormatException ex) {
396             throw new IOException("Broken archive, unable to parse GNU string table length field as a number", ex);
397         }
398 
399         namebuffer = IOUtils.readRange(in, bufflen);
400         final int read = namebuffer.length;
401         trackReadBytes(read);
402         if (read != bufflen) {
403             throw new IOException("Failed to read complete // record: expected=" + bufflen + " read=" + read);
404         }
405         return new ArArchiveEntry(GNU_STRING_TABLE_NAME, bufflen);
406     }
407 
408     private void trackReadBytes(final long read) {
409         count(read);
410         if (read > 0) {
411             offset += read;
412         }
413     }
414 }