You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@poi.apache.org by ce...@apache.org on 2015/03/22 14:33:43 UTC
svn commit: r1668367 - in /poi/trunk:
src/integrationtest/org/apache/poi/stress/
src/scratchpad/src/org/apache/poi/hdf/extractor/
src/scratchpad/testcases/org/apache/poi/hdf/extractor/ test-data/document/
Author: centic
Date: Sun Mar 22 13:33:43 2015
New Revision: 1668367
URL: http://svn.apache.org/r1668367
Log:
Bug 47304: use fixed encoding when extracting text in WordDocument
Added:
poi/trunk/test-data/document/47304.doc
Modified:
poi/trunk/src/integrationtest/org/apache/poi/stress/HWPFFileHandler.java
poi/trunk/src/scratchpad/src/org/apache/poi/hdf/extractor/WordDocument.java
poi/trunk/src/scratchpad/testcases/org/apache/poi/hdf/extractor/TestWordDocument.java
Modified: poi/trunk/src/integrationtest/org/apache/poi/stress/HWPFFileHandler.java
URL: http://svn.apache.org/viewvc/poi/trunk/src/integrationtest/org/apache/poi/stress/HWPFFileHandler.java?rev=1668367&r1=1668366&r2=1668367&view=diff
==============================================================================
--- poi/trunk/src/integrationtest/org/apache/poi/stress/HWPFFileHandler.java (original)
+++ poi/trunk/src/integrationtest/org/apache/poi/stress/HWPFFileHandler.java Sun Mar 22 13:33:43 2015
@@ -18,12 +18,21 @@ package org.apache.poi.stress;
import static org.junit.Assert.assertNotNull;
+import java.io.ByteArrayInputStream;
+import java.io.ByteArrayOutputStream;
+import java.io.File;
import java.io.FileInputStream;
+import java.io.IOException;
import java.io.InputStream;
+import java.io.PrintWriter;
+import java.io.StringWriter;
+import org.apache.poi.hdf.extractor.WordDocument;
import org.apache.poi.hwpf.HWPFDocument;
+import org.apache.poi.hwpf.extractor.WordExtractor;
import org.junit.Test;
+@SuppressWarnings("deprecation")
public class HWPFFileHandler extends POIFSFileHandler {
@Override
public void handleFile(InputStream stream) throws Exception {
@@ -33,16 +42,53 @@ public class HWPFFileHandler extends POI
assertNotNull(doc.getEndnotes());
handlePOIDocument(doc);
+
+ // fails for many documents, but is deprecated anyway...
+ // handleWordDocument(doc);
+ }
+
+ protected void handleWordDocument(HWPFDocument doc) throws IOException {
+ ByteArrayOutputStream outStream = new ByteArrayOutputStream();
+ doc.write(outStream);
+
+ WordDocument wordDoc = new WordDocument(new ByteArrayInputStream(outStream.toByteArray()));
+
+ StringWriter docTextWriter = new StringWriter();
+ PrintWriter out = new PrintWriter(docTextWriter);
+ try {
+ wordDoc.writeAllText(out);
+ } finally {
+ out.close();
+ }
+ docTextWriter.close();
}
+
+
// a test-case to test this locally without executing the full TestAllFiles
@Test
public void test() throws Exception {
- InputStream stream = new FileInputStream("test-data/document/HeaderFooterUnicode.doc");
+ File file = new File("test-data/document/47304.doc");
+
+ InputStream stream = new FileInputStream(file);
try {
handleFile(stream);
} finally {
stream.close();
}
+
+ handleExtracting(file);
+
+ stream = new FileInputStream(file);
+ try {
+ WordExtractor extractor = new WordExtractor(stream);
+ try {
+ assertNotNull(extractor.getText());
+ } finally {
+ extractor.close();
+ }
+ } finally {
+ stream.close();
+ }
}
}
\ No newline at end of file
Modified: poi/trunk/src/scratchpad/src/org/apache/poi/hdf/extractor/WordDocument.java
URL: http://svn.apache.org/viewvc/poi/trunk/src/scratchpad/src/org/apache/poi/hdf/extractor/WordDocument.java?rev=1668367&r1=1668366&r2=1668367&view=diff
==============================================================================
--- poi/trunk/src/scratchpad/src/org/apache/poi/hdf/extractor/WordDocument.java (original)
+++ poi/trunk/src/scratchpad/src/org/apache/poi/hdf/extractor/WordDocument.java Sun Mar 22 13:33:43 2015
@@ -177,7 +177,7 @@ public final class WordDocument {
}
else
{
- String sText = new String(_header, start, end-start);
+ String sText = new String(_header, start, end-start, "windows-1252");
out.write(sText);
}
}
Modified: poi/trunk/src/scratchpad/testcases/org/apache/poi/hdf/extractor/TestWordDocument.java
URL: http://svn.apache.org/viewvc/poi/trunk/src/scratchpad/testcases/org/apache/poi/hdf/extractor/TestWordDocument.java?rev=1668367&r1=1668366&r2=1668367&view=diff
==============================================================================
--- poi/trunk/src/scratchpad/testcases/org/apache/poi/hdf/extractor/TestWordDocument.java (original)
+++ poi/trunk/src/scratchpad/testcases/org/apache/poi/hdf/extractor/TestWordDocument.java Sun Mar 22 13:33:43 2015
@@ -17,6 +17,15 @@
package org.apache.poi.hdf.extractor;
+import static org.junit.Assert.*;
+
+import java.io.IOException;
+import java.io.PrintWriter;
+import java.io.StringWriter;
+
+import org.apache.poi.hwpf.HWPFDocument;
+import org.apache.poi.hwpf.HWPFTestDataSamples;
+import org.apache.poi.hwpf.extractor.WordExtractor;
import org.junit.Test;
@@ -31,4 +40,31 @@ public class TestWordDocument {
//WordDocument.main(new String[] {"test-data/document/Word6.doc", "/tmp/test.doc"});
WordDocument.main(new String[] {"test-data/document/53446.doc", "/tmp/test.doc"});
}
+
+ @SuppressWarnings("deprecation")
+ @Test
+ public void test47304() throws IOException {
+ HWPFDocument doc = HWPFTestDataSamples.openSampleFile("47304.doc");
+ assertNotNull(doc);
+
+ WordExtractor extractor = new WordExtractor(doc);
+ String text = extractor.getText();
+ //System.out.println(text);
+ assertTrue("Had: " + text, text.contains("Just a \u201Ctest\u201D"));
+ extractor.close();
+
+ WordDocument wordDoc = new WordDocument("test-data/document/47304.doc");
+
+ StringWriter docTextWriter = new StringWriter();
+ PrintWriter out = new PrintWriter(docTextWriter);
+ try {
+ wordDoc.writeAllText(out);
+ } finally {
+ out.close();
+ }
+ docTextWriter.close();
+
+ //System.out.println(docTextWriter.toString());
+ assertTrue("Had: " + docTextWriter.toString(), docTextWriter.toString().contains("Just a \u201Ctest\u201D"));
+ }
}
Added: poi/trunk/test-data/document/47304.doc
URL: http://svn.apache.org/viewvc/poi/trunk/test-data/document/47304.doc?rev=1668367&view=auto
==============================================================================
Binary files poi/trunk/test-data/document/47304.doc (added) and poi/trunk/test-data/document/47304.doc Sun Mar 22 13:33:43 2015 differ
---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscribe@poi.apache.org
For additional commands, e-mail: commits-help@poi.apache.org