001/* 002 * Licensed to the Apache Software Foundation (ASF) under one 003 * or more contributor license agreements. See the NOTICE file 004 * distributed with this work for additional information 005 * regarding copyright ownership. The ASF licenses this file 006 * to you under the Apache License, Version 2.0 (the 007 * "License"); you may not use this file except in compliance 008 * with the License. You may obtain a copy of the License at 009 * 010 * http://www.apache.org/licenses/LICENSE-2.0 011 * 012 * Unless required by applicable law or agreed to in writing, 013 * software distributed under the License is distributed on an 014 * "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY 015 * KIND, either express or implied. See the License for the 016 * specific language governing permissions and limitations 017 * under the License. 018 */ 019package org.apache.commons.compress.archivers.ar; 020 021import java.io.EOFException; 022import java.io.IOException; 023import java.io.InputStream; 024import java.util.Arrays; 025 026import org.apache.commons.compress.archivers.ArchiveEntry; 027import org.apache.commons.compress.archivers.ArchiveInputStream; 028import org.apache.commons.compress.utils.ArchiveUtils; 029import org.apache.commons.compress.utils.IOUtils; 030 031/** 032 * Implements the "ar" archive format as an input stream. 033 * 034 * @NotThreadSafe 035 */ 036public class ArArchiveInputStream extends ArchiveInputStream { 037 038 // offsets and length of meta data parts 039 private static final int NAME_OFFSET = 0; 040 private static final int NAME_LEN = 16; 041 private static final int LAST_MODIFIED_OFFSET = NAME_LEN; 042 043 private static final int LAST_MODIFIED_LEN = 12; 044 045 private static final int USER_ID_OFFSET = LAST_MODIFIED_OFFSET + LAST_MODIFIED_LEN; 046 047 private static final int USER_ID_LEN = 6; 048 049 private static final int GROUP_ID_OFFSET = USER_ID_OFFSET + USER_ID_LEN; 050 private static final int GROUP_ID_LEN = 6; 051 private static final int FILE_MODE_OFFSET = GROUP_ID_OFFSET + GROUP_ID_LEN; 052 private static final int FILE_MODE_LEN = 8; 053 private static final int LENGTH_OFFSET = FILE_MODE_OFFSET + FILE_MODE_LEN; 054 private static final int LENGTH_LEN = 10; 055 static final String BSD_LONGNAME_PREFIX = "#1/"; 056 private static final int BSD_LONGNAME_PREFIX_LEN = 057 BSD_LONGNAME_PREFIX.length(); 058 private static final String BSD_LONGNAME_PATTERN = 059 "^" + BSD_LONGNAME_PREFIX + "\\d+"; 060 private static final String GNU_STRING_TABLE_NAME = "//"; 061 private static final String GNU_LONGNAME_PATTERN = "^/\\d+"; 062 /** 063 * Does the name look like it is a long name (or a name containing 064 * spaces) as encoded by BSD ar? 065 * 066 * <p>From the FreeBSD ar(5) man page:</p> 067 * <pre> 068 * BSD In the BSD variant, names that are shorter than 16 069 * characters and without embedded spaces are stored 070 * directly in this field. If a name has an embedded 071 * space, or if it is longer than 16 characters, then 072 * the string "#1/" followed by the decimal represen- 073 * tation of the length of the file name is placed in 074 * this field. The actual file name is stored immedi- 075 * ately after the archive header. The content of the 076 * archive member follows the file name. The ar_size 077 * field of the header (see below) will then hold the 078 * sum of the size of the file name and the size of 079 * the member. 080 * </pre> 081 * 082 * @since 1.3 083 */ 084 private static boolean isBSDLongName(final String name) { 085 return name != null && name.matches(BSD_LONGNAME_PATTERN); 086 } 087 088 /** 089 * Is this the name of the "Archive String Table" as used by 090 * SVR4/GNU to store long file names? 091 * 092 * <p>GNU ar stores multiple extended file names in the data section 093 * of a file with the name "//", this record is referred to by 094 * future headers.</p> 095 * 096 * <p>A header references an extended file name by storing a "/" 097 * followed by a decimal offset to the start of the file name in 098 * the extended file name data section.</p> 099 * 100 * <p>The format of the "//" file itself is simply a list of the 101 * long file names, each separated by one or more LF 102 * characters. Note that the decimal offsets are number of 103 * characters, not line or string number within the "//" file.</p> 104 */ 105 private static boolean isGNUStringTable(final String name) { 106 return GNU_STRING_TABLE_NAME.equals(name); 107 } 108 109 /** 110 * Checks if the signature matches ASCII "!<arch>" followed by a single LF 111 * control character 112 * 113 * @param signature 114 * the bytes to check 115 * @param length 116 * the number of bytes to check 117 * @return true, if this stream is an Ar archive stream, false otherwise 118 */ 119 public static boolean matches(final byte[] signature, final int length) { 120 // 3c21 7261 6863 0a3e 121 122 return length >= 8 && signature[0] == 0x21 && 123 signature[1] == 0x3c && signature[2] == 0x61 && 124 signature[3] == 0x72 && signature[4] == 0x63 && 125 signature[5] == 0x68 && signature[6] == 0x3e && 126 signature[7] == 0x0a; 127 } 128 129 private final InputStream input; 130 131 private long offset; 132 133 private boolean closed; 134 135 /* 136 * If getNextEntry has been called, the entry metadata is stored in 137 * currentEntry. 138 */ 139 private ArArchiveEntry currentEntry; 140 141 // Storage area for extra long names (GNU ar) 142 private byte[] namebuffer; 143 144 /* 145 * The offset where the current entry started. -1 if no entry has been 146 * called 147 */ 148 private long entryOffset = -1; 149 150 // cached buffer for meta data - must only be used locally in the class (COMPRESS-172 - reduce garbage collection) 151 private final byte[] metaData = 152 new byte[NAME_LEN + LAST_MODIFIED_LEN + USER_ID_LEN + GROUP_ID_LEN + FILE_MODE_LEN + LENGTH_LEN]; 153 154 /** 155 * Constructs an Ar input stream with the referenced stream 156 * 157 * @param inputStream 158 * the ar input stream 159 */ 160 public ArArchiveInputStream(final InputStream inputStream) { 161 this.input = inputStream; 162 } 163 164 private int asInt(final byte[] byteArray, final int offset, final int len) { 165 return asInt(byteArray, offset, len, 10, false); 166 } 167 168 private int asInt(final byte[] byteArray, final int offset, final int len, final boolean treatBlankAsZero) { 169 return asInt(byteArray, offset, len, 10, treatBlankAsZero); 170 } 171 172 private int asInt(final byte[] byteArray, final int offset, final int len, final int base) { 173 return asInt(byteArray, offset, len, base, false); 174 } 175 176 private int asInt(final byte[] byteArray, final int offset, final int len, final int base, final boolean treatBlankAsZero) { 177 final String string = ArchiveUtils.toAsciiString(byteArray, offset, len).trim(); 178 if (string.isEmpty() && treatBlankAsZero) { 179 return 0; 180 } 181 return Integer.parseInt(string, base); 182 } 183 private long asLong(final byte[] byteArray, final int offset, final int len) { 184 return Long.parseLong(ArchiveUtils.toAsciiString(byteArray, offset, len).trim()); 185 } 186 /* 187 * (non-Javadoc) 188 * 189 * @see java.io.InputStream#close() 190 */ 191 @Override 192 public void close() throws IOException { 193 if (!closed) { 194 closed = true; 195 input.close(); 196 } 197 currentEntry = null; 198 } 199 200 /** 201 * Reads the real name from the current stream assuming the very 202 * first bytes to be read are the real file name. 203 * 204 * @see #isBSDLongName 205 * 206 * @since 1.3 207 */ 208 private String getBSDLongName(final String bsdLongName) throws IOException { 209 final int nameLen = 210 Integer.parseInt(bsdLongName.substring(BSD_LONGNAME_PREFIX_LEN)); 211 final byte[] name = IOUtils.readRange(input, nameLen); 212 final int read = name.length; 213 trackReadBytes(read); 214 if (read != nameLen) { 215 throw new EOFException(); 216 } 217 return ArchiveUtils.toAsciiString(name); 218 } 219 220 /** 221 * Get an extended name from the GNU extended name buffer. 222 * 223 * @param offset pointer to entry within the buffer 224 * @return the extended file name; without trailing "/" if present. 225 * @throws IOException if name not found or buffer not set up 226 */ 227 private String getExtendedName(final int offset) throws IOException { 228 if (namebuffer == null) { 229 throw new IOException("Cannot process GNU long filename as no // record was found"); 230 } 231 for (int i = offset; i < namebuffer.length; i++) { 232 if (namebuffer[i] == '\012' || namebuffer[i] == 0) { 233 if (namebuffer[i - 1] == '/') { 234 i--; // drop trailing / 235 } 236 return ArchiveUtils.toAsciiString(namebuffer, offset, i - offset); 237 } 238 } 239 throw new IOException("Failed to read entry: " + offset); 240 } 241 242 /** 243 * Returns the next AR entry in this stream. 244 * 245 * @return the next AR entry. 246 * @throws IOException 247 * if the entry could not be read 248 */ 249 public ArArchiveEntry getNextArEntry() throws IOException { 250 if (currentEntry != null) { 251 final long entryEnd = entryOffset + currentEntry.getLength(); 252 final long skipped = IOUtils.skip(input, entryEnd - offset); 253 trackReadBytes(skipped); 254 currentEntry = null; 255 } 256 257 if (offset == 0) { 258 final byte[] expected = ArchiveUtils.toAsciiBytes(ArArchiveEntry.HEADER); 259 final byte[] realized = IOUtils.readRange(input, expected.length); 260 final int read = realized.length; 261 trackReadBytes(read); 262 if (read != expected.length) { 263 throw new IOException("Failed to read header. Occurred at byte: " + getBytesRead()); 264 } 265 if (!Arrays.equals(expected, realized)) { 266 throw new IOException("Invalid header " + ArchiveUtils.toAsciiString(realized)); 267 } 268 } 269 270 if (offset % 2 != 0) { 271 if (input.read() < 0) { 272 // hit eof 273 return null; 274 } 275 trackReadBytes(1); 276 } 277 278 { 279 final int read = IOUtils.readFully(input, metaData); 280 trackReadBytes(read); 281 if (read == 0) { 282 return null; 283 } 284 if (read < metaData.length) { 285 throw new IOException("Truncated ar archive"); 286 } 287 } 288 289 { 290 final byte[] expected = ArchiveUtils.toAsciiBytes(ArArchiveEntry.TRAILER); 291 final byte[] realized = IOUtils.readRange(input, expected.length); 292 final int read = realized.length; 293 trackReadBytes(read); 294 if (read != expected.length) { 295 throw new IOException("Failed to read entry trailer. Occurred at byte: " + getBytesRead()); 296 } 297 if (!Arrays.equals(expected, realized)) { 298 throw new IOException("Invalid entry trailer. not read the content? Occurred at byte: " + getBytesRead()); 299 } 300 } 301 302 entryOffset = offset; 303 304// GNU ar uses a '/' to mark the end of the filename; this allows for the use of spaces without the use of an extended filename. 305 306 // entry name is stored as ASCII string 307 String temp = ArchiveUtils.toAsciiString(metaData, NAME_OFFSET, NAME_LEN).trim(); 308 if (isGNUStringTable(temp)) { // GNU extended filenames entry 309 currentEntry = readGNUStringTable(metaData, LENGTH_OFFSET, LENGTH_LEN); 310 return getNextArEntry(); 311 } 312 313 long len = asLong(metaData, LENGTH_OFFSET, LENGTH_LEN); 314 if (temp.endsWith("/")) { // GNU terminator 315 temp = temp.substring(0, temp.length() - 1); 316 } else if (isGNULongName(temp)) { 317 final int off = Integer.parseInt(temp.substring(1));// get the offset 318 temp = getExtendedName(off); // convert to the long name 319 } else if (isBSDLongName(temp)) { 320 temp = getBSDLongName(temp); 321 // entry length contained the length of the file name in 322 // addition to the real length of the entry. 323 // assume file name was ASCII, there is no "standard" otherwise 324 final int nameLen = temp.length(); 325 len -= nameLen; 326 entryOffset += nameLen; 327 } 328 329 if (len < 0) { 330 throw new IOException("broken archive, entry with negative size"); 331 } 332 333 currentEntry = new ArArchiveEntry(temp, len, 334 asInt(metaData, USER_ID_OFFSET, USER_ID_LEN, true), 335 asInt(metaData, GROUP_ID_OFFSET, GROUP_ID_LEN, true), 336 asInt(metaData, FILE_MODE_OFFSET, FILE_MODE_LEN, 8), 337 asLong(metaData, LAST_MODIFIED_OFFSET, LAST_MODIFIED_LEN)); 338 return currentEntry; 339 } 340 341 /* 342 * (non-Javadoc) 343 * 344 * @see 345 * org.apache.commons.compress.archivers.ArchiveInputStream#getNextEntry() 346 */ 347 @Override 348 public ArchiveEntry getNextEntry() throws IOException { 349 return getNextArEntry(); 350 } 351 352 /** 353 * Does the name look like it is a long name (or a name containing 354 * spaces) as encoded by SVR4/GNU ar? 355 * 356 * @see #isGNUStringTable 357 */ 358 private boolean isGNULongName(final String name) { 359 return name != null && name.matches(GNU_LONGNAME_PATTERN); 360 } 361 362 /* 363 * (non-Javadoc) 364 * 365 * @see java.io.InputStream#read(byte[], int, int) 366 */ 367 @Override 368 public int read(final byte[] b, final int off, final int len) throws IOException { 369 if (len == 0) { 370 return 0; 371 } 372 if (currentEntry == null) { 373 throw new IllegalStateException("No current ar entry"); 374 } 375 final long entryEnd = entryOffset + currentEntry.getLength(); 376 if (len < 0 || offset >= entryEnd) { 377 return -1; 378 } 379 final int toRead = (int) Math.min(len, entryEnd - offset); 380 final int ret = this.input.read(b, off, toRead); 381 trackReadBytes(ret); 382 return ret; 383 } 384 385 /** 386 * Reads the GNU archive String Table. 387 * 388 * @see #isGNUStringTable 389 */ 390 private ArArchiveEntry readGNUStringTable(final byte[] length, final int offset, final int len) throws IOException { 391 final int bufflen = asInt(length, offset, len); // Assume length will fit in an int 392 namebuffer = IOUtils.readRange(input, bufflen); 393 final int read = namebuffer.length; 394 trackReadBytes(read); 395 if (read != bufflen){ 396 throw new IOException("Failed to read complete // record: expected=" 397 + bufflen + " read=" + read); 398 } 399 return new ArArchiveEntry(GNU_STRING_TABLE_NAME, bufflen); 400 } 401 402 private void trackReadBytes(final long read) { 403 count(read); 404 if (read > 0) { 405 offset += read; 406 } 407 } 408}