Author: maxcom
Date: Mon Aug 22 07:56:43 2011
New Revision: 1160137
URL: http://svn.apache.org/viewvc?rev=1160137&view=rev
Log:
bug#51686 - ConcurrentModificationException in Tika's OfficeParser
Added:
poi/trunk/test-data/document/testWORD.doc (with props)
Modified:
poi/trunk/src/scratchpad/src/org/apache/poi/hwpf/HWPFDocumentCore.java
poi/trunk/src/scratchpad/testcases/org/apache/poi/hwpf/extractor/TestWordExtractor.java
Modified: poi/trunk/src/scratchpad/src/org/apache/poi/hwpf/HWPFDocumentCore.java
URL:
http://svn.apache.org/viewvc/poi/trunk/src/scratchpad/src/org/apache/poi/hwpf/HWPFDocumentCore.java?rev=1160137&r1=1160136&r2=1160137&view=diff
==============================================================================
--- poi/trunk/src/scratchpad/src/org/apache/poi/hwpf/HWPFDocumentCore.java
(original)
+++ poi/trunk/src/scratchpad/src/org/apache/poi/hwpf/HWPFDocumentCore.java Mon
Aug 22 07:56:43 2011
@@ -141,41 +141,32 @@ public abstract class HWPFDocumentCore e
* @throws IOException If there is an unexpected IOException from the passed
* in POIFSFileSystem.
*/
- public HWPFDocumentCore(DirectoryNode directory) throws IOException
- {
+ public HWPFDocumentCore(DirectoryNode directory) throws IOException {
// Sort out the hpsf properties
- super(directory);
+ super(directory);
// read in the main stream.
DocumentEntry documentProps = (DocumentEntry)
- directory.getEntry("WordDocument");
+ directory.getEntry("WordDocument");
_mainStream = new byte[documentProps.getSize()];
directory.createDocumentInputStream(STREAM_WORD_DOCUMENT).read(_mainStream);
// Create our FIB, and check for the doc being encrypted
_fib = new FileInformationBlock(_mainStream);
- if(_fib.isFEncrypted()) {
- throw new EncryptedDocumentException("Cannot process encrypted word
files!");
+ if (_fib.isFEncrypted()) {
+ throw new EncryptedDocumentException("Cannot process encrypted word
files!");
}
- {
- DirectoryEntry objectPoolEntry;
- try
- {
- objectPoolEntry = (DirectoryEntry) directory
- .getEntry( STREAM_OBJECT_POOL );
- }
- catch ( FileNotFoundException exc )
- {
- objectPoolEntry = directory
- .createDirectory( STREAM_OBJECT_POOL );
- }
- _objectPool = new ObjectPoolImpl( objectPoolEntry );
- }
+ try {
+ DirectoryEntry objectPoolEntry = (DirectoryEntry) directory
+ .getEntry(STREAM_OBJECT_POOL);
+ _objectPool = new ObjectPoolImpl(objectPoolEntry);
+ } catch (FileNotFoundException exc) {
}
+ }
- /**
+ /**
* Returns the range which covers the whole of the document, but excludes
* any headers and footers.
*/
Modified:
poi/trunk/src/scratchpad/testcases/org/apache/poi/hwpf/extractor/TestWordExtractor.java
URL:
http://svn.apache.org/viewvc/poi/trunk/src/scratchpad/testcases/org/apache/poi/hwpf/extractor/TestWordExtractor.java?rev=1160137&r1=1160136&r2=1160137&view=diff
==============================================================================
---
poi/trunk/src/scratchpad/testcases/org/apache/poi/hwpf/extractor/TestWordExtractor.java
(original)
+++
poi/trunk/src/scratchpad/testcases/org/apache/poi/hwpf/extractor/TestWordExtractor.java
Mon Aug 22 07:56:43 2011
@@ -24,9 +24,13 @@ import org.apache.poi.hwpf.HWPFDocument;
import org.apache.poi.hwpf.HWPFTestDataSamples;
import org.apache.poi.hwpf.OldWordFileFormatException;
import org.apache.poi.poifs.filesystem.DirectoryNode;
+import org.apache.poi.poifs.filesystem.Entry;
import org.apache.poi.poifs.filesystem.NPOIFSFileSystem;
import org.apache.poi.poifs.filesystem.POIFSFileSystem;
+import java.io.IOException;
+import java.io.InputStream;
+
/**
* Test the different routes to extracting text
*
@@ -353,4 +357,21 @@ public final class TestWordExtractor ext
assertEquals(p_text1_block, extractor.getText());
}
}
+
+ public void testRootEntiesNavigation() throws IOException {
+ InputStream is =
POIDataSamples.getDocumentInstance().openResourceAsStream("testWORD.doc");
+
+ POIFSFileSystem fs = new POIFSFileSystem(is);
+
+ String text = null;
+
+ for (Entry entry : fs.getRoot()) {
+ if ("WordDocument".equals(entry.getName())) {
+ WordExtractor ex = new WordExtractor(fs);
+ text = ex.getText();
+ }
+ }
+
+ assertNotNull(text);
+ }
}
Added: poi/trunk/test-data/document/testWORD.doc
URL:
http://svn.apache.org/viewvc/poi/trunk/test-data/document/testWORD.doc?rev=1160137&view=auto
==============================================================================
Binary file - no diff available.
Propchange: poi/trunk/test-data/document/testWORD.doc
------------------------------------------------------------------------------
svn:mime-type = application/octet-stream
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
