TarReader.java
/* Copyright (c) 2001-2024, The HSQL Development Group
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions are met:
*
* Redistributions of source code must retain the above copyright notice, this
* list of conditions and the following disclaimer.
*
* Redistributions in binary form must reproduce the above copyright notice,
* this list of conditions and the following disclaimer in the documentation
* and/or other materials provided with the distribution.
*
* Neither the name of the HSQL Development Group nor the names of its
* contributors may be used to endorse or promote products derived from this
* software without specific prior written permission.
*
* THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
* AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
* ARE DISCLAIMED. IN NO EVENT SHALL HSQL DEVELOPMENT GROUP, HSQLDB.ORG,
* OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
* EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
* PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
* LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND
* ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
* (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
* SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
*/
package org.hsqldb.lib.tar;
import org.hsqldb.lib.StringUtil;
import java.io.File;
import java.io.FileOutputStream;
import java.io.IOException;
import java.io.PipedInputStream;
import java.io.PipedOutputStream;
import java.text.SimpleDateFormat;
import java.util.regex.Pattern;
/**
* Reads a Tar file for reporting or extraction.
* N.b. this is not a <I>Reader</I> in the {@code java.io.Reader} sense,
* but in the sense of differentiating {@code tar x} and
* {@code tar t} from {@code tar c}.
* <P>
* <B>SECURITY NOTE</B>
* Due to pitiful lack of support for file security in Java before version 1.6,
* this class will only explicitly set permissions if it is compiled for Java
* 1.6. If it was not, and if your tar entries contain private data in files
* with 0400 or similar, be aware that they will be extracted with privs such
* that they can be ready by anybody.
* </P>
*
* @author Blaine Simpson (blaine dot simpson at admc dot com)
*/
public class TarReader {
public static final int LIST_MODE = 0;
/**
* EXTRACT_MODE refuses to overwrite existing files.
*/
public static final int EXTRACT_MODE = 1;
/**
* OVERWRITE_MODE is just EXTRACT_MODE where we will silently overwrite
* existing files upon extraction.
*/
public static final int OVERWRITE_MODE = 2;
protected TarFileInputStream archive;
protected Pattern[] patterns = null;
protected int mode;
protected File extractBaseDir; // null means current directory
// Not used for Absolute path entries
// This path is always absolutized
/**
* Compression is determined directly by the suffix of the file name in
* the specified path.
*
* @param inFile Absolute or relative (from user.dir) path to
* tar file to be read. Suffix may indicate
* a compression method.
* @param mode Whether to list, extract-without-overwrite, or
* extract-with-overwrite.
* @param patternStrings
* List of regular expressions to match against tar entry
* names. If null, all entries will be listed or
* extracted. If non-null, then only entries with names
* which match will be extracted or listed.
* @param readBufferBlocks Null will use default tar value.
* @param inDir Directory that RELATIVE entries will be extracted
* relative to. Defaults to current directory (user.dir).
* Only used for extract modes and relative file entries.
* @throws IOException on access failure
* @throws IllegalArgumentException if any given pattern is an invalid
* regular expression. Don't have to worry about this if
* you call with null 'patterns' param.
* @see Pattern
*/
public TarReader(File inFile, int mode, String[] patternStrings,
Integer readBufferBlocks, File inDir) throws IOException {
this.mode = mode;
File archiveFile = inFile.getAbsoluteFile();
extractBaseDir = (inDir == null) ? null
: inDir.getAbsoluteFile();
int compression =
TarFileOutputStream.Compression.NO_COMPRESSION;
if (archiveFile.getName().endsWith(".tgz")
|| archiveFile.getName().endsWith(".gz")) {
compression = TarFileOutputStream.Compression.GZIP_COMPRESSION;
}
if (patternStrings != null) {
patterns = new Pattern[patternStrings.length];
for (int i = 0; i < patternStrings.length; i++) {
patterns[i] = Pattern.compile(patternStrings[i]);
}
}
// Don't check for archive file existence here. We can depend upon the
// TarFileInputStream to check that.
archive = (readBufferBlocks == null)
? new TarFileInputStream(archiveFile, compression)
: new TarFileInputStream(archiveFile, compression,
readBufferBlocks.intValue());
}
public void read() throws IOException, TarMalformatException {
TarEntryHeader header;
boolean anyUnsupporteds = false;
boolean matched;
Long paxSize = null;
String paxString = null;
try {
EACH_HEADER:
while (archive.readNextHeaderBlock()) {
header = new TarEntryHeader(archive.readBuffer);
char entryType = header.getEntryType();
if (entryType == 'x') {
/* Since we don't know the name of the target file yet,
* we must load the size from all pax headers.
* If the target file is not thereafter excluded via
* patterns, we will need this size for the listing or to
* extract the data.
*/
paxSize = getPifData(header).getSize();
paxString = header.toString();
continue;
}
if (paxSize != null) {
// Ignore "size" field in the entry header because PIF
// setting overrides.
header.setDataSize(paxSize.longValue());
paxSize = null;
}
if (patterns != null) {
matched = false;
for (int i = 0; i < patterns.length; i++) {
if (patterns[i].matcher(header.getPath()).matches()) {
matched = true;
break;
}
}
if (!matched) {
paxString = null;
skipFileData(header);
continue EACH_HEADER;
}
}
if (entryType != '\0' && entryType != '0'
&& entryType != 'x') {
anyUnsupporteds = true;
}
switch (mode) {
case LIST_MODE :
if (paxString != null) {
System.out.println(paxString);
}
System.out.println(header.toString());
skipFileData(header);
break;
case EXTRACT_MODE :
case OVERWRITE_MODE :
if (paxString != null) {
System.out.println(paxString);
}
/* Display entry summary before successful extraction.
* Both "tar" and "rsync" display the name of the
* currently extracting file, and we do the same.
* Therefore the currently "shown" name is still being
* extracted.
*/
System.out.println(header.toString());
// Instance variable mode will be used to differentiate
// behavior inside of extractFile().
if (entryType == '\0' || entryType == '0'
|| entryType == 'x') {
extractFile(header);
} else {
skipFileData(header);
}
break;
default :
throw new IllegalArgumentException(
RB.unsupported_mode.getString(mode));
}
paxString = null;
}
if (anyUnsupporteds) {
System.out.println(RB.unsupported_entry_present.getString());
}
} catch (IOException ioe) {
archive.close();
throw ioe;
}
}
protected PIFData getPifData(TarEntryHeader header)
throws IOException, TarMalformatException {
/*
* If you modify this, make sure to not intermix reading/writing of
* the PipedInputStream and the PipedOutputStream, or you could
* cause dead-lock. Everything is safe if you close the
* PipedOutputStream before reading the PipedInputStream.
*/
long dataSize = header.getDataSize();
if (dataSize < 1) {
throw new TarMalformatException(
RB.pif_unknown_datasize.getString());
}
if (dataSize > Integer.MAX_VALUE) {
throw new TarMalformatException(RB.pif_data_toobig.getString(
Long.toString(dataSize), Integer.MAX_VALUE));
}
int readNow;
int readBlocks = (int) (dataSize / 512L);
int modulus = (int) (dataSize % 512L);
// Couldn't care less about the entry "name" field.
PipedInputStream inPipe = null;
PipedOutputStream outPipe = new PipedOutputStream();
/* This constructor not available until Java 1.6:
new PipedInputStream(outPipe, (int) dataSize);
*/
try {
inPipe = new PipedInputStream(outPipe);
while (readBlocks > 0) {
readNow = (readBlocks > archive.getReadBufferBlocks())
? archive.getReadBufferBlocks()
: readBlocks;
archive.readBlocks(readNow);
readBlocks -= readNow;
outPipe.write(archive.readBuffer, 0, readNow * 512);
}
if (modulus != 0) { archive.readBlock();
outPipe.write(archive.readBuffer, 0, modulus);
}
outPipe.flush(); // Do any good on a pipe?
} catch (IOException ioe) {
if (inPipe != null) {
inPipe.close();
}
throw ioe;
} finally {
try {
outPipe.close();
} finally {
outPipe = null; // Encourage buffer GC
}
}
return new PIFData(inPipe);
}
protected void extractFile(TarEntryHeader header)
throws IOException, TarMalformatException {
if (header.getDataSize() < 1) {
throw new TarMalformatException(RB.data_size_unknown.getString());
}
int readNow;
int readBlocks = (int) (header.getDataSize() / 512L);
int modulus = (int) (header.getDataSize() % 512L);
File newFile = header.generateFile();
if (!newFile.isAbsolute()) {
newFile = (extractBaseDir == null) ? newFile.getAbsoluteFile()
: new File(extractBaseDir,
newFile.getPath());
}
// newFile is definitively Absolutized at this point
File parentDir = newFile.getParentFile();
if (newFile.exists()) {
if (mode != TarReader.OVERWRITE_MODE) {
throw new IOException(
RB.extraction_exists.getString(newFile.getAbsolutePath()));
}
if (!newFile.isFile()) {
throw new IOException(
RB.extraction_exists_notfile.getString(
newFile.getAbsolutePath()));
}
// Better to let FileOutputStream creation zero it than to
// to newFile.delete().
}
if (parentDir.exists()) {
if (!parentDir.isDirectory()) {
throw new IOException(
RB.extraction_parent_not_dir.getString(
parentDir.getAbsolutePath()));
}
if (!parentDir.canWrite()) {
throw new IOException(
RB.extraction_parent_not_writable.getString(
parentDir.getAbsolutePath()));
}
} else {
if (!parentDir.mkdirs()) {
throw new IOException(
RB.extraction_parent_mkfail.getString(
parentDir.getAbsolutePath()));
}
}
int fileMode = header.getFileMode();
FileOutputStream outStream = new FileOutputStream(newFile);
try {
// Don't know exactly why I am still able to write to the file
// after removing read and write privs from myself, but it does
// work.
newFile.setExecutable(false, false);
newFile.setReadable(false, false);
newFile.setWritable(false, false);
newFile.setExecutable(((fileMode & 0100) != 0), true);
newFile.setReadable((fileMode & 0400) != 0, true);
newFile.setWritable((fileMode & 0200) != 0, true);
while (readBlocks > 0) {
readNow = (readBlocks > archive.getReadBufferBlocks())
? archive.getReadBufferBlocks()
: readBlocks;
archive.readBlocks(readNow);
readBlocks -= readNow;
outStream.write(archive.readBuffer, 0, readNow * 512);
}
if (modulus != 0) {
archive.readBlock();
outStream.write(archive.readBuffer, 0, modulus);
}
outStream.flush();
} finally {
try {
outStream.close();
} finally {
outStream = null; // Encourage buffer GC
}
}
newFile.setLastModified(header.getModTime() * 1000);
if (newFile.length() != header.getDataSize()) {
throw new IOException(RB.write_count_mismatch.getString(
Long.toString(header.getDataSize()),
newFile.getAbsolutePath(),
Long.toString(newFile.length())));
}
}
protected void skipFileData(TarEntryHeader header)
throws IOException, TarMalformatException {
/*
* Some entry types which we don't support have 0 data size.
* If we just return here, the entry will just be skipped./
*/
if (header.getDataSize() == 0) {
return;
}
if (header.getDataSize() < 0) {
throw new TarMalformatException(RB.data_size_unknown.getString());
}
int skipNow;
int oddBlocks = (header.getDataSize() % 512L == 0L) ? 0
: 1;
int skipBlocks = (int) (header.getDataSize() / 512L) + oddBlocks;
while (skipBlocks > 0) {
skipNow = (skipBlocks > archive.getReadBufferBlocks())
? archive.getReadBufferBlocks()
: skipBlocks;
archive.readBlocks(skipNow);
skipBlocks -= skipNow;
}
}
/**
* A Tar entry header constituted from a header block in a tar file.
*
* @author Blaine Simpson (blaine dot simpson at admc dot com)
*/
@SuppressWarnings("serial")
static protected class TarEntryHeader {
static protected class MissingField extends Exception {
private final TarHeaderField field;
public MissingField(TarHeaderField field) {
this.field = field;
}
public String getMessage() {
return RB.header_field_missing.getString(field.toString());
}
}
protected SimpleDateFormat sdf =
new SimpleDateFormat("yyyy-MM-dd'T'HH:mm:ss");
/**
* @param rawHeader May be longer than 512 bytes, but the first 512
* bytes MUST COMPRISE a raw tar entry header.
*
* @throws TarMalformatException if malformed
*/
public TarEntryHeader(byte[] rawHeader) throws TarMalformatException {
this.rawHeader = rawHeader;
Long expectedCheckSum = readInteger(TarHeaderField.checksum);
try {
if (expectedCheckSum == null) {
throw new MissingField(TarHeaderField.checksum);
}
long calculatedCheckSum = headerChecksum();
if (expectedCheckSum.longValue() != calculatedCheckSum) {
throw new TarMalformatException(
RB.checksum_mismatch.getString(
expectedCheckSum.toString(),
Long.toString(calculatedCheckSum)));
}
path = readString(TarHeaderField.name);
if (path == null) {
throw new MissingField(TarHeaderField.name);
}
Long longObject = readInteger(TarHeaderField.mode);
if (longObject == null) {
throw new MissingField(TarHeaderField.mode);
}
fileMode = (int) longObject.longValue();
longObject = readInteger(TarHeaderField.size);
if (longObject != null) {
dataSize = longObject.longValue();
}
longObject = readInteger(TarHeaderField.mtime);
if (longObject == null) {
throw new MissingField(TarHeaderField.mtime);
}
modTime = longObject.longValue();
} catch (MissingField mf) {
throw new TarMalformatException(mf.getMessage());
}
entryType = readChar(TarHeaderField.typeflag);
ownerName = readString(TarHeaderField.uname);
String pathPrefix = readString(TarHeaderField.prefix);
if (pathPrefix != null) {
path = pathPrefix + '/' + path;
}
// We're not loading the "gname" field, since there is nothing at
// all that Java can do with it.
ustar = isUstar();
}
protected byte[] rawHeader;
/* CRITICALLY IMPORTANT: TO NOT USE rawHeader.length OR DEPEND ON
* THE LENGTH OF the rawHeader ARRAY! Use only the first 512 bytes!
*/
protected String path;
protected int fileMode;
protected long dataSize = -1; // In bytes
protected long modTime;
protected char entryType;
protected String ownerName;
protected boolean ustar;
/**
* @return a new Absolutized File object generated from this
* TarEntryHeader.
*/
public File generateFile() {
if (entryType != '\0' && entryType != '0') {
throw new IllegalStateException(
RB.create_only_normal.getString());
}
// Unfortunately, it does no good to set modification times or
// privileges here, since those settings have no effect on our
// new file until after is created by the FileOutputStream
// constructor.
return new File(path);
}
public char getEntryType() {
return entryType;
}
public String getPath() {
return path;
}
/**
* Setter is needed in order to override header size setting for Pax.
*
* @param dataSize long
*/
public void setDataSize(long dataSize) {
this.dataSize = dataSize;
}
public long getDataSize() {
return dataSize;
}
public long getModTime() {
return modTime;
}
public int getFileMode() {
return fileMode;
}
/**
* Choosing not to report fields that we don't write (e.g. "gname"),
* but which would certainly be useful for a general Java tar client
* implementation.
*
* <P> This design decision is subject to change.
*
* @return String
*/
public String toString() {
StringBuilder sb = new StringBuilder(
sdf.format(Long.valueOf(modTime * 1000L)) + ' ');
sb.append((entryType == '\0') ? ' '
: entryType);
sb.append(ustar ? '*'
: ' ');
sb.append(
" "
+ StringUtil.toPaddedString(
Integer.toOctalString(fileMode), 4, ' ', false) + ' '
+ StringUtil.toPaddedString(
Long.toString(dataSize), 11, ' ', false) + " ");
sb.append(StringUtil.toPaddedString(((ownerName == null) ? "-"
: ownerName), 8,
' ',
true));
sb.append(" " + path);
return sb.toString();
}
/**
* Is this any UStar variant
*
* @throws TarMalformatException
* @return boolean
*/
public boolean isUstar() throws TarMalformatException {
String magicString = readString(TarHeaderField.magic);
return magicString != null && magicString.startsWith("ustar");
}
/**
*
* @return index based at 0 == from
* @param ba byte[]
* @param val byte
* @param from int
* @param to int
*/
public static int indexOf(byte[] ba, byte val, int from, int to) {
for (int i = from; i < to; i++) {
if (ba[i] == val) {
return i - from;
}
}
return -1;
}
protected char readChar(TarHeaderField field) throws TarMalformatException {
/* Depends on readString(int) contract that it will never return
* a 0-length String */
String s = readString(field);
return (s == null) ? '\0'
: s.charAt(0);
}
/**
*
* @return null or String with {@code length() > 0}.
* @param field TarHeaderField
* @throws TarMalformatException
*/
protected String readString(TarHeaderField field) throws TarMalformatException {
int start = field.getStart();
int stop = field.getStop();
int termIndex = TarEntryHeader.indexOf(rawHeader, (byte) 0, start,
stop);
switch (termIndex) {
case 0 :
return null;
case -1 :
termIndex = stop - start;
break;
default:
break;
}
try {
return new String(rawHeader, start, termIndex);
} catch (Throwable t) {
// Java API does not specify behavior if decoding fails.
throw new TarMalformatException(
RB.bad_header_value.getString(field.toString()));
}
}
/**
* Integer as in positive whole number, which does not imply Java types
* of {@code int} or {@code Integer}.
*
* @param field TarHeaderField
* @throws TarMalformatException if malformed
* @return Long
*/
protected Long readInteger(TarHeaderField field)
throws TarMalformatException {
String s = readString(field);
if (s == null) {
return null;
}
try {
return Long.valueOf(s, 8);
} catch (NumberFormatException nfe) {
throw new TarMalformatException(
RB.bad_numeric_header_value.getString(
field.toString(), nfe.toString()));
}
}
protected long headerChecksum() {
long sum = 0;
for (int i = 0; i < 512; i++) {
boolean isInRange =
(i >= TarHeaderField.checksum.getStart()
&& i < TarHeaderField.checksum.getStop());
// We ignore current contents of the checksum field so that
// this method will continue to work right, even if we later
// recycle the header or RE-calculate a header.
sum += isInRange ? 32
: (255 & rawHeader[i]);
}
return sum;
}
}
}