Repository: commons-compress Updated Branches: refs/heads/master 7a3889d47 -> fd69d8ca2
COMPRESS-469 add lenient flag that allows accepting some broken tars Project: http://git-wip-us.apache.org/repos/asf/commons-compress/repo Commit: http://git-wip-us.apache.org/repos/asf/commons-compress/commit/fd69d8ca Tree: http://git-wip-us.apache.org/repos/asf/commons-compress/tree/fd69d8ca Diff: http://git-wip-us.apache.org/repos/asf/commons-compress/diff/fd69d8ca Branch: refs/heads/master Commit: fd69d8ca2212d373a5a8f220530e8c0fa17ddf3b Parents: 7a3889d Author: Stefan Bodewig <bode...@apache.org> Authored: Sat Nov 10 21:28:54 2018 +0100 Committer: Stefan Bodewig <bode...@apache.org> Committed: Sat Nov 10 21:28:54 2018 +0100 ---------------------------------------------------------------------- src/changes/changes.xml | 4 ++ .../compress/archivers/tar/TarArchiveEntry.java | 55 ++++++++++++++++---- .../archivers/tar/TarArchiveInputStream.java | 31 ++++++++++- .../commons/compress/archivers/TarTestCase.java | 9 ++++ 4 files changed, 88 insertions(+), 11 deletions(-) ---------------------------------------------------------------------- http://git-wip-us.apache.org/repos/asf/commons-compress/blob/fd69d8ca/src/changes/changes.xml ---------------------------------------------------------------------- diff --git a/src/changes/changes.xml b/src/changes/changes.xml index c220dc0..5b9619a 100644 --- a/src/changes/changes.xml +++ b/src/changes/changes.xml @@ -54,6 +54,10 @@ The <action> type attribute can be add,update,fix,remove. cost of potentially missing important information. See the javadocs of the ZipFile class for details. </action> + <action issue="COMPRESS-469" type="add" date="2018-11-10"> + TarArchiveInputStream has a new constructor-arg lenient that + can be used to accept certain broken archives. + </action> </release> <release version="1.18" date="2018-08-16" description="Release 1.18"> http://git-wip-us.apache.org/repos/asf/commons-compress/blob/fd69d8ca/src/main/java/org/apache/commons/compress/archivers/tar/TarArchiveEntry.java ---------------------------------------------------------------------- diff --git a/src/main/java/org/apache/commons/compress/archivers/tar/TarArchiveEntry.java b/src/main/java/org/apache/commons/compress/archivers/tar/TarArchiveEntry.java index b6b735a..e5c3a06 100644 --- a/src/main/java/org/apache/commons/compress/archivers/tar/TarArchiveEntry.java +++ b/src/main/java/org/apache/commons/compress/archivers/tar/TarArchiveEntry.java @@ -146,6 +146,13 @@ import org.apache.commons.compress.utils.ArchiveUtils; public class TarArchiveEntry implements ArchiveEntry, TarConstants { private static final TarArchiveEntry[] EMPTY_TAR_ARCHIVE_ENTRIES = new TarArchiveEntry[0]; + /** + * Value used to indicate unknown mode, user/groupids, device numbers and modTime when parsing a file in lenient + * mode an the archive contains illegal fields. + * @since 1.19 + */ + private static final long UNKNOWN = -1l; + /** The entry's name. */ private String name = ""; @@ -400,8 +407,25 @@ public class TarArchiveEntry implements ArchiveEntry, TarConstants { */ public TarArchiveEntry(final byte[] headerBuf, final ZipEncoding encoding) throws IOException { + this(headerBuf, encoding, false); + } + + /** + * Construct an entry from an archive's header bytes. File is set + * to null. + * + * @param headerBuf The header bytes from a tar archive entry. + * @param encoding encoding to use for file names + * @param lenient when set to true illegal values for group/userid, mode, device numbers and timestamp will be + * ignored and the fields set to {@link #UNKNOWN}. When set to false such illegal fields cause an exception instead. + * @since 1.19 + * @throws IllegalArgumentException if any of the numeric fields have an invalid format + * @throws IOException on error + */ + public TarArchiveEntry(final byte[] headerBuf, final ZipEncoding encoding, boolean lenient) + throws IOException { this(false); - parseTarHeader(headerBuf, encoding); + parseTarHeader(headerBuf, encoding, false, lenient); } /** @@ -1243,7 +1267,7 @@ public class TarArchiveEntry implements ArchiveEntry, TarConstants { parseTarHeader(header, TarUtils.DEFAULT_ENCODING); } catch (final IOException ex) { // NOSONAR try { - parseTarHeader(header, TarUtils.DEFAULT_ENCODING, true); + parseTarHeader(header, TarUtils.DEFAULT_ENCODING, true, false); } catch (final IOException ex2) { // not really possible throw new RuntimeException(ex2); //NOSONAR @@ -1263,26 +1287,26 @@ public class TarArchiveEntry implements ArchiveEntry, TarConstants { */ public void parseTarHeader(final byte[] header, final ZipEncoding encoding) throws IOException { - parseTarHeader(header, encoding, false); + parseTarHeader(header, encoding, false, false); } private void parseTarHeader(final byte[] header, final ZipEncoding encoding, - final boolean oldStyle) + final boolean oldStyle, final boolean lenient) throws IOException { int offset = 0; name = oldStyle ? TarUtils.parseName(header, offset, NAMELEN) : TarUtils.parseName(header, offset, NAMELEN, encoding); offset += NAMELEN; - mode = (int) TarUtils.parseOctalOrBinary(header, offset, MODELEN); + mode = (int) parseOctalOrBinary(header, offset, MODELEN, lenient); offset += MODELEN; - userId = (int) TarUtils.parseOctalOrBinary(header, offset, UIDLEN); + userId = (int) parseOctalOrBinary(header, offset, UIDLEN, lenient); offset += UIDLEN; - groupId = (int) TarUtils.parseOctalOrBinary(header, offset, GIDLEN); + groupId = (int) parseOctalOrBinary(header, offset, GIDLEN, lenient); offset += GIDLEN; size = TarUtils.parseOctalOrBinary(header, offset, SIZELEN); offset += SIZELEN; - modTime = TarUtils.parseOctalOrBinary(header, offset, MODTIMELEN); + modTime = parseOctalOrBinary(header, offset, MODTIMELEN, lenient); offset += MODTIMELEN; checkSumOK = TarUtils.verifyCheckSum(header); offset += CHKSUMLEN; @@ -1301,9 +1325,9 @@ public class TarArchiveEntry implements ArchiveEntry, TarConstants { : TarUtils.parseName(header, offset, GNAMELEN, encoding); offset += GNAMELEN; if (linkFlag == LF_CHR || linkFlag == LF_BLK) { - devMajor = (int) TarUtils.parseOctalOrBinary(header, offset, DEVLEN); + devMajor = (int) parseOctalOrBinary(header, offset, DEVLEN, lenient); offset += DEVLEN; - devMinor = (int) TarUtils.parseOctalOrBinary(header, offset, DEVLEN); + devMinor = (int) parseOctalOrBinary(header, offset, DEVLEN, lenient); offset += DEVLEN; } else { offset += 2 * DEVLEN; @@ -1350,6 +1374,17 @@ public class TarArchiveEntry implements ArchiveEntry, TarConstants { } } + private long parseOctalOrBinary(byte[] header, int offset, int length, boolean lenient) { + if (lenient) { + try { + return TarUtils.parseOctalOrBinary(header, offset, length); + } catch (IllegalArgumentException ex) { + return UNKNOWN; + } + } + return TarUtils.parseOctalOrBinary(header, offset, length); + } + /** * Strips Windows' drive letter as well as any leading slashes, * turns path separators into forward slahes. http://git-wip-us.apache.org/repos/asf/commons-compress/blob/fd69d8ca/src/main/java/org/apache/commons/compress/archivers/tar/TarArchiveInputStream.java ---------------------------------------------------------------------- diff --git a/src/main/java/org/apache/commons/compress/archivers/tar/TarArchiveInputStream.java b/src/main/java/org/apache/commons/compress/archivers/tar/TarArchiveInputStream.java index daaf729..7aeb6dd 100644 --- a/src/main/java/org/apache/commons/compress/archivers/tar/TarArchiveInputStream.java +++ b/src/main/java/org/apache/commons/compress/archivers/tar/TarArchiveInputStream.java @@ -80,6 +80,8 @@ public class TarArchiveInputStream extends ArchiveInputStream { // the global PAX header private Map<String, String> globalPaxHeaders = new HashMap<>(); + private final boolean lenient; + /** * Constructor for TarInputStream. * @param is the input stream to use @@ -91,6 +93,17 @@ public class TarArchiveInputStream extends ArchiveInputStream { /** * Constructor for TarInputStream. * @param is the input stream to use + * @param lenient when set to true illegal values for group/userid, mode, device numbers and timestamp will be + * ignored and the fields set to {@link #UNKNOWN}. When set to false such illegal fields cause an exception instead. + * @since 1.19 + */ + public TarArchiveInputStream(final InputStream is, boolean lenient) { + this(is, TarConstants.DEFAULT_BLKSIZE, TarConstants.DEFAULT_RCDSIZE, null, lenient); + } + + /** + * Constructor for TarInputStream. + * @param is the input stream to use * @param encoding name of the encoding to use for file names * @since 1.4 */ @@ -140,12 +153,28 @@ public class TarArchiveInputStream extends ArchiveInputStream { */ public TarArchiveInputStream(final InputStream is, final int blockSize, final int recordSize, final String encoding) { + this(is, blockSize, recordSize, encoding, false); + } + + /** + * Constructor for TarInputStream. + * @param is the input stream to use + * @param blockSize the block size to use + * @param recordSize the record size to use + * @param encoding name of the encoding to use for file names + * @param lenient when set to true illegal values for group/userid, mode, device numbers and timestamp will be + * ignored and the fields set to {@link #UNKNOWN}. When set to false such illegal fields cause an exception instead. + * @since 1.19 + */ + public TarArchiveInputStream(final InputStream is, final int blockSize, final int recordSize, + final String encoding, boolean lenient) { this.is = is; this.hasHitEOF = false; this.encoding = encoding; this.zipEncoding = ZipEncodingHelper.getZipEncoding(encoding); this.recordSize = recordSize; this.blockSize = blockSize; + this.lenient = lenient; } /** @@ -280,7 +309,7 @@ public class TarArchiveInputStream extends ArchiveInputStream { } try { - currEntry = new TarArchiveEntry(headerBuf, zipEncoding); + currEntry = new TarArchiveEntry(headerBuf, zipEncoding, lenient); } catch (final IllegalArgumentException e) { throw new IOException("Error detected parsing the header", e); } http://git-wip-us.apache.org/repos/asf/commons-compress/blob/fd69d8ca/src/test/java/org/apache/commons/compress/archivers/TarTestCase.java ---------------------------------------------------------------------- diff --git a/src/test/java/org/apache/commons/compress/archivers/TarTestCase.java b/src/test/java/org/apache/commons/compress/archivers/TarTestCase.java index 4792583..e694927 100644 --- a/src/test/java/org/apache/commons/compress/archivers/TarTestCase.java +++ b/src/test/java/org/apache/commons/compress/archivers/TarTestCase.java @@ -330,4 +330,13 @@ public final class TarTestCase extends AbstractTestCase { in.close(); } + @Test + public void testCOMPRESS178Lenient() throws Exception { + final File input = getFile("COMPRESS-178.tar"); + final InputStream is = new FileInputStream(input); + try (final ArchiveInputStream in = new TarArchiveInputStream(is, true)) { + in.getNextEntry(); + } + } + }
