Improved error message when attempting to read BIFF2 file (bugzilla 46269)

git-svn-id: https://svn.apache.org/repos/asf/poi/trunk@721620 13f79535-47bb-0310-9956-ffa450edef68

src/documentation/content/xdocs/changes.xml

@@ -37,6 +37,7 @@
 
 		<!-- Don't forget to update status.xml too! -->
         <release version="3.5-beta5" date="2008-??-??">
+           <action dev="POI-DEVELOPERS" type="add">46269 - Improved error message when attempting to read BIFF2 file</action>
            <action dev="POI-DEVELOPERS" type="fix">46206 - Fixed Sheet to tolerate missing DIMENSION records</action>
            <action dev="POI-DEVELOPERS" type="add">46301 - added pivot table records: SXDI, SXVDEX, SXPI, SXIDSTM, SXVIEW, SXVD, SXVS, et al</action>
            <action dev="POI-DEVELOPERS" type="fix">46280 - Fixed RowRecordsAggregate etc to properly skip PivotTable records</action>

src/documentation/content/xdocs/status.xml

@@ -34,6 +34,7 @@
 	<!-- Don't forget to update changes.xml too! -->
     <changes>
         <release version="3.5-beta5" date="2008-??-??">
+           <action dev="POI-DEVELOPERS" type="add">46269 - Improved error message when attempting to read BIFF2 file</action>
            <action dev="POI-DEVELOPERS" type="fix">46206 - Fixed Sheet to tolerate missing DIMENSION records</action>
            <action dev="POI-DEVELOPERS" type="add">46301 - added pivot table records: SXDI, SXVDEX, SXPI, SXIDSTM, SXVIEW, SXVD, SXVS, et al</action>
            <action dev="POI-DEVELOPERS" type="fix">46280 - Fixed RowRecordsAggregate etc to properly skip PivotTable records</action>

src/java/org/apache/poi/poifs/storage/HeaderBlockReader.java

@@ -1,4 +1,3 @@
-
 /* ====================================================================
    Licensed to the Apache Software Foundation (ASF) under one or more
    contributor license agreements.  See the NOTICE file distributed with
@@ -15,191 +14,199 @@
    See the License for the specific language governing permissions and
    limitations under the License.
 ==================================================================== */
-        
 
 package org.apache.poi.poifs.storage;
 
-import java.io.*;
+import static org.apache.poi.poifs.storage.HeaderBlockConstants._bat_array_offset;
+import static org.apache.poi.poifs.storage.HeaderBlockConstants._bat_count_offset;
+import static org.apache.poi.poifs.storage.HeaderBlockConstants._max_bats_in_header;
+import static org.apache.poi.poifs.storage.HeaderBlockConstants._property_start_offset;
+import static org.apache.poi.poifs.storage.HeaderBlockConstants._sbat_start_offset;
+import static org.apache.poi.poifs.storage.HeaderBlockConstants._signature;
+import static org.apache.poi.poifs.storage.HeaderBlockConstants._signature_offset;
+import static org.apache.poi.poifs.storage.HeaderBlockConstants._xbat_count_offset;
+import static org.apache.poi.poifs.storage.HeaderBlockConstants._xbat_start_offset;
+
+import java.io.IOException;
+import java.io.InputStream;
 
 import org.apache.poi.poifs.common.POIFSConstants;
 import org.apache.poi.poifs.filesystem.OfficeXmlFileException;
+import org.apache.poi.util.HexDump;
 import org.apache.poi.util.IOUtils;
-import org.apache.poi.util.IntegerField;
 import org.apache.poi.util.LittleEndian;
 import org.apache.poi.util.LittleEndianConsts;
-import org.apache.poi.util.LongField;
 
 /**
  * The block containing the archive header
  *
  * @author Marc Johnson (mjohnson at apache dot org)
  */
+public final class HeaderBlockReader {
+	/**
+	 * What big block size the file uses. Most files
+	 *  use 512 bytes, but a few use 4096
+	 */
+	private int bigBlockSize = POIFSConstants.BIG_BLOCK_SIZE;
 
-public class HeaderBlockReader
+	/** number of big block allocation table blocks (int) */
-    implements HeaderBlockConstants
+	private int _bat_count;
-{
-    /**
-     * What big block size the file uses. Most files
-     *  use 512 bytes, but a few use 4096
-     */
-    private int bigBlockSize = POIFSConstants.BIG_BLOCK_SIZE;
 
-    // number of big block allocation table blocks (int)
+	/** start of the property set block (int index of the property set
-    private IntegerField _bat_count;
+	 * chain's first big block)
+	 */
+	private int _property_start;
 
-    // start of the property set block (int index of the property set
+	/** start of the small block allocation table (int index of small
-    // chain's first big block)
+	 * block allocation table's first big block)
-    private IntegerField _property_start;
+	 */
+	private int _sbat_start;
 
-    // start of the small block allocation table (int index of small
+	/** big block index for extension to the big block allocation table */
-    // block allocation table's first big block)
+	private int _xbat_start;
-    private IntegerField _sbat_start;
+	private int _xbat_count;
+	private byte[] _data;
 
-    // big block index for extension to the big block allocation table
+	/**
-    private IntegerField _xbat_start;
+	 * create a new HeaderBlockReader from an InputStream
-    private IntegerField _xbat_count;
+	 *
-    private byte[]       _data;
+	 * @param stream the source InputStream
+	 *
+	 * @exception IOException on errors or bad data
+	 */
+	public HeaderBlockReader(InputStream stream) throws IOException {
+		// At this point, we don't know how big our
+		//  block sizes are
+		// So, read the first 32 bytes to check, then
+		//  read the rest of the block
+		byte[] blockStart = new byte[32];
+		int bsCount = IOUtils.readFully(stream, blockStart);
+		if(bsCount != 32) {
+			alertShortRead(bsCount);
+		}
+		
+		// Figure out our block size
+		if(blockStart[30] == 12) {
+			bigBlockSize = POIFSConstants.LARGER_BIG_BLOCK_SIZE;
+		}
+		_data = new byte[ bigBlockSize ];
+		System.arraycopy(blockStart, 0, _data, 0, blockStart.length);
+		
+		// Now we can read the rest of our header
+		int byte_count = IOUtils.readFully(stream, _data, blockStart.length, _data.length - blockStart.length);
+		if (byte_count+bsCount != bigBlockSize) {
+			alertShortRead(byte_count);
+		}
 
-    /**
+		// verify signature
-     * create a new HeaderBlockReader from an InputStream
+		long signature = LittleEndian.getLong(_data, _signature_offset);
-     *
-     * @param stream the source InputStream
-     *
-     * @exception IOException on errors or bad data
-     */
 
-    public HeaderBlockReader(final InputStream stream)
+		if (signature != _signature) {
-        throws IOException
-    {
-    	// At this point, we don't know how big our
-    	//  block sizes are
-    	// So, read the first 32 bytes to check, then
-    	//  read the rest of the block
-    	byte[] blockStart = new byte[32];
-    	int bsCount = IOUtils.readFully(stream, blockStart);
-    	if(bsCount != 32) {
-    		alertShortRead(bsCount);
-    	}
-    	
-    	// Figure out our block size
-    	if(blockStart[30] == 12) {
-    		bigBlockSize = POIFSConstants.LARGER_BIG_BLOCK_SIZE;
-    	}
-        _data = new byte[ bigBlockSize ];
-        System.arraycopy(blockStart, 0, _data, 0, blockStart.length);
-    	
-    	// Now we can read the rest of our header
-        int byte_count = IOUtils.readFully(stream, _data, blockStart.length, _data.length - blockStart.length);
-        if (byte_count+bsCount != bigBlockSize) {
-    		alertShortRead(byte_count);
-        }
-
-        // verify signature
-        LongField signature = new LongField(_signature_offset, _data);
-
-        if (signature.get() != _signature)
-        {
 			// Is it one of the usual suspects?
-        	byte[] OOXML_FILE_HEADER = POIFSConstants.OOXML_FILE_HEADER;
+			byte[] OOXML_FILE_HEADER = POIFSConstants.OOXML_FILE_HEADER;
 			if(_data[0] == OOXML_FILE_HEADER[0] && 
 					_data[1] == OOXML_FILE_HEADER[1] && 
 					_data[2] == OOXML_FILE_HEADER[2] &&
 					_data[3] == OOXML_FILE_HEADER[3]) {
 				throw new OfficeXmlFileException("The supplied data appears to be in the Office 2007+ XML. You are calling the part of POI that deals with OLE2 Office Documents. You need to call a different part of POI to process this data (eg XSSF instead of HSSF)");
 			}
+			if ((signature & 0xFF8FFFFFFFFFFFFFL) == 0x0010000200040009L) {
+				// BIFF2 raw stream starts with BOF (sid=0x0009, size=0x0004, data=0x00t0)
+				throw new IllegalArgumentException("The supplied data appears to be in BIFF2 format.  "
+						+ "POI only supports BIFF8 format");
+			}
 
 			// Give a generic error
-            throw new IOException("Invalid header signature; read "
+			throw new IOException("Invalid header signature; read "
-                                  + signature.get() + ", expected "
+				                  + longToHex(signature) + ", expected "
-                                  + _signature);
+				                  + longToHex(_signature));
-        }
+		}
-        _bat_count      = new IntegerField(_bat_count_offset, _data);
+		_bat_count      = getInt(_bat_count_offset, _data);
-        _property_start = new IntegerField(_property_start_offset, _data);
+		_property_start = getInt(_property_start_offset, _data);
-        _sbat_start     = new IntegerField(_sbat_start_offset, _data);
+		_sbat_start     = getInt(_sbat_start_offset, _data);
-        _xbat_start     = new IntegerField(_xbat_start_offset, _data);
+		_xbat_start     = getInt(_xbat_start_offset, _data);
-        _xbat_count     = new IntegerField(_xbat_count_offset, _data);
+		_xbat_count     = getInt(_xbat_count_offset, _data);
-    }
+	}
-    
+	
-    private void alertShortRead(int read) throws IOException {
+	private static int getInt(int offset, byte[] data) {
-    	if (read == -1)
+		return LittleEndian.getInt(data, offset);
-    		//Cant have -1 bytes read in the error message!
+	}
-    		read = 0;
-        String type = " byte" + ((read == 1) ? ("")
-                                                   : ("s"));
 
-        throw new IOException("Unable to read entire header; "
+	private static String longToHex(long value) {
-                              + read + type + " read; expected "
+		return new String(HexDump.longToHex(value));
-                              + bigBlockSize + " bytes");
+	}
-    }
 
-    /**
+	private void alertShortRead(int pRead) throws IOException {
-     * get start of Property Table
+		int read;
-     *
+		if (pRead < 0) {
-     * @return the index of the first block of the Property Table
+			//Can't have -1 bytes read in the error message!
-     */
+			read = 0;
-    public int getPropertyStart()
+		} else {
-    {
+			read = pRead;
-        return _property_start.get();
+		}
-    }
+		String type = " byte" + (read == 1 ? (""): ("s"));
 
-    /**
+		throw new IOException("Unable to read entire header; "
-     * @return start of small block allocation table
+				+ read + type + " read; expected "
-     */
+				+ bigBlockSize + " bytes");
+	}
 
-    public int getSBATStart()
+	/**
-    {
+	 * get start of Property Table
-        return _sbat_start.get();
+	 *
-    }
+	 * @return the index of the first block of the Property Table
+	 */
+	public int getPropertyStart() {
+		return _property_start;
+	}
 
-    /**
+	/**
-     * @return number of BAT blocks
+	 * @return start of small block allocation table
-     */
+	 */
+	public int getSBATStart() {
+		return _sbat_start;
+	}
 
-    public int getBATCount()
+	/**
-    {
+	 * @return number of BAT blocks
-        return _bat_count.get();
+	 */
-    }
+	public int getBATCount() {
+		return _bat_count;
+	}
 
-    /**
+	/**
-     * @return BAT array
+	 * @return BAT array
-     */
+	 */
+	public int[] getBATArray() {
+		int[] result = new int[ _max_bats_in_header ];
+		int   offset = _bat_array_offset;
 
-    public int [] getBATArray()
+		for (int j = 0; j < _max_bats_in_header; j++) {
-    {
+			result[ j ] = LittleEndian.getInt(_data, offset);
-        int[] result = new int[ _max_bats_in_header ];
+			offset      += LittleEndianConsts.INT_SIZE;
-        int   offset = _bat_array_offset;
+		}
+		return result;
+	}
 
-        for (int j = 0; j < _max_bats_in_header; j++)
+	/**
-        {
+	 * @return XBAT count
-            result[ j ] = LittleEndian.getInt(_data, offset);
+	 */
-            offset      += LittleEndianConsts.INT_SIZE;
+	public int getXBATCount() {
-        }
+		return _xbat_count;
-        return result;
+	}
-    }
 
-    /**
+	/**
-     * @return XBAT count
+	 * @return XBAT index
-     */
+	 */
-
+	public int getXBATIndex() {
-    public int getXBATCount()
+		return _xbat_start;
-    {
+	}
-        return _xbat_count.get();
+	
-    }
+	/**
-
+	 * @return The Big Block size, normally 512 bytes, sometimes 4096 bytes
-    /**
+	 */
-     * @return XBAT index
+	public int getBigBlockSize() {
-     */
+		return bigBlockSize;
-
+	}
-    public int getXBATIndex()
+}
-    {
-        return _xbat_start.get();
-    }
-    
-    /**
-     * @return The Big Block size, normally 512 bytes, sometimes 4096 bytes
-     */
-    public int getBigBlockSize() {
-    	return bigBlockSize;
-    }
-}   // end public class HeaderBlockReader