Author: bdelacretaz
Date: Thu Oct 11 22:46:55 2007
New Revision: 584054
URL: http://svn.apache.org/viewvc?rev=584054&view=rev
Log:
TIKA-52 - RereadableInputStream needs to support not closing the input stream
it wraps
Modified:
incubator/tika/trunk/CHANGES.txt
incubator/tika/trunk/src/main/java/org/apache/tika/ms/MSParser.java
incubator/tika/trunk/src/main/java/org/apache/tika/utils/RereadableInputStream.java
incubator/tika/trunk/src/test/java/org/apache/tika/RereadableInputStreamTest.java
Modified: incubator/tika/trunk/CHANGES.txt
URL:
http://svn.apache.org/viewvc/incubator/tika/trunk/CHANGES.txt?rev=584054&r1=584053&r2=584054&view=diff
==============================================================================
--- incubator/tika/trunk/CHANGES.txt (original)
+++ incubator/tika/trunk/CHANGES.txt Thu Oct 11 22:46:55 2007
@@ -88,3 +88,6 @@
40. TIKA-55 - ParseUtils.getParser() method variants should have consistent
parameter orders
(K. Bennett)
+41. TIKA-52 - RereadableInputStream needs to support not closing the input
stream it wraps.
+ (K. Bennett via bdelacretaz)
+
Modified: incubator/tika/trunk/src/main/java/org/apache/tika/ms/MSParser.java
URL:
http://svn.apache.org/viewvc/incubator/tika/trunk/src/main/java/org/apache/tika/ms/MSParser.java?rev=584054&r1=584053&r2=584054&view=diff
==============================================================================
--- incubator/tika/trunk/src/main/java/org/apache/tika/ms/MSParser.java
(original)
+++ incubator/tika/trunk/src/main/java/org/apache/tika/ms/MSParser.java Thu Oct
11 22:46:55 2007
@@ -40,7 +40,7 @@
public String parse(InputStream input, Metadata metadata)
throws IOException, TikaException {
RereadableInputStream ris =
- new RereadableInputStream(input, MEMORY_THRESHOLD);
+ new RereadableInputStream(input, MEMORY_THRESHOLD, true, false);
try {
// First, extract properties
POIFSReader reader = new POIFSReader();
Modified:
incubator/tika/trunk/src/main/java/org/apache/tika/utils/RereadableInputStream.java
URL:
http://svn.apache.org/viewvc/incubator/tika/trunk/src/main/java/org/apache/tika/utils/RereadableInputStream.java?rev=584054&r1=584053&r2=584054&view=diff
==============================================================================
---
incubator/tika/trunk/src/main/java/org/apache/tika/utils/RereadableInputStream.java
(original)
+++
incubator/tika/trunk/src/main/java/org/apache/tika/utils/RereadableInputStream.java
Thu Oct 11 22:46:55 2007
@@ -35,6 +35,12 @@
*/
public class RereadableInputStream extends InputStream {
+
+ /**
+ * Input stream originally passed to the constructor.
+ */
+ private InputStream originalInputStream;
+
/**
* The inputStream currently being used by this object to read contents;
* may be the original stream passed in, or a stream that reads
@@ -92,6 +98,14 @@
*/
private boolean readToEndOfStreamOnFirstRewind = true;
+
+ /**
+ * Specifies whether or not to close the original input stream
+ * when close() is called. Defaults to true.
+ */
+ private boolean closeOriginalStreamOnClose = true;
+
+
// TODO: At some point it would be better to replace the current approach
// (specifying the above) with more automated behavior. The stream could
// keep the original stream open until EOF was reached. For example, if:
@@ -122,33 +136,20 @@
* garbage collection (i.e. its reference set to null) when the
* content size exceeds the array's size, when close() is called, or
* when there are no more references to the instance.
- */
- public RereadableInputStream(InputStream inputStream, int
maxBytesInMemory) {
- this(inputStream, maxBytesInMemory, true);
- }
-
- /**
- * Creates a rereadable input stream.
- *
- * @param inputStream stream containing the source of data
- * @param maxBytesInMemory maximum number of bytes to use to store
- * the stream's contents in memory before switching to disk; note that
- * the instance will preallocate a byte array whose size is
- * maxBytesInMemory. This byte array will be made available for
- * garbage collection (i.e. its reference set to null) when the
- * content size exceeds the array's size, when close() is called, or
- * when there are no more references to the instance.
* @param readToEndOfStreamOnFirstRewind Specifies whether or not to
* read to the end of stream on first rewind. If this is set to false,
* then when rewind() is first called, only those bytes already read
* from the original stream will be available from then on.
*/
public RereadableInputStream(InputStream inputStream, int maxBytesInMemory,
- boolean readToEndOfStreamOnFirstRewind) {
+ boolean readToEndOfStreamOnFirstRewind,
+ boolean closeOriginalStreamOnClose) {
this.inputStream = inputStream;
+ this.originalInputStream = inputStream;
this.maxBytesInMemory = maxBytesInMemory;
byteBuffer = new byte[maxBytesInMemory];
this.readToEndOfStreamOnFirstRewind = readToEndOfStreamOnFirstRewind;
+ this.closeOriginalStreamOnClose = closeOriginalStreamOnClose;
}
/**
@@ -201,7 +202,10 @@
*/
// Does anyone need/want for this to be public?
private void closeStream() throws IOException {
- if (inputStream != null) {
+ if (inputStream != null
+ &&
+ (inputStream != originalInputStream
+ || closeOriginalStreamOnClose)) {
inputStream.close();
inputStream = null;
}
Modified:
incubator/tika/trunk/src/test/java/org/apache/tika/RereadableInputStreamTest.java
URL:
http://svn.apache.org/viewvc/incubator/tika/trunk/src/test/java/org/apache/tika/RereadableInputStreamTest.java?rev=584054&r1=584053&r2=584054&view=diff
==============================================================================
---
incubator/tika/trunk/src/test/java/org/apache/tika/RereadableInputStreamTest.java
(original)
+++
incubator/tika/trunk/src/test/java/org/apache/tika/RereadableInputStreamTest.java
Thu Oct 11 22:46:55 2007
@@ -39,7 +39,7 @@
InputStream is = createTestInputStream();
RereadableInputStream ris = new RereadableInputStream(is,
- MEMORY_THRESHOLD);
+ MEMORY_THRESHOLD, true, true);
try {
for (int pass = 0; pass < NUM_PASSES; pass++) {
for (int byteNum = 0; byteNum < TEST_SIZE; byteNum++) {
@@ -74,16 +74,12 @@
try {
InputStream s1 = createTestInputStream();
- ris = new RereadableInputStream(s1, 5, readToEndOnRewind);
+ ris = new RereadableInputStream(s1, 5, readToEndOnRewind, true);
ris.read();
assertEquals(1, ris.getSize());
ris.rewind();
boolean moreBytesWereRead = (ris.getSize() > 1);
- if (readToEndOnRewind) {
- assertTrue(moreBytesWereRead);
- } else {
- assertFalse(moreBytesWereRead);
- }
+ assertEquals(readToEndOnRewind, moreBytesWereRead);
} finally {
if (ris != null) {
ris.close();
@@ -92,8 +88,10 @@
}
- private InputStream createTestInputStream() throws IOException {
- return new BufferedInputStream(new FileInputStream(createTestFile()));
+ private TestInputStream createTestInputStream() throws IOException {
+ return new TestInputStream(
+ new BufferedInputStream(
+ new FileInputStream(createTestFile())));
}
private File createTestFile() throws IOException {
@@ -106,4 +104,46 @@
fos.close();
return testfile;
}
+
+
+ public void testCloseBehavior() throws IOException {
+ doACloseBehaviorTest(true);
+ doACloseBehaviorTest(false);
+ }
+
+ private void doACloseBehaviorTest(boolean wantToClose) throws IOException {
+
+ TestInputStream tis = createTestInputStream();
+ RereadableInputStream ris =
+ new RereadableInputStream(tis, 5, true, wantToClose);
+ ris.close();
+ assertEquals(wantToClose, tis.isClosed());
+
+ if (! tis.isClosed()) {
+ tis.close();
+ }
+ }
+
+
+ /**
+ * Adds isClosed() to a BufferedInputStream.
+ */
+ class TestInputStream extends BufferedInputStream {
+
+ private boolean closed;
+
+ public TestInputStream(InputStream inputStream) {
+ super(inputStream);
+ }
+
+ public void close() throws IOException {
+ super.close();
+ closed = true;
+ }
+
+ public boolean isClosed() {
+ return closed;
+ }
+ }
+
}
