Author: jukka
Date: Fri Sep 19 15:06:53 2008
New Revision: 697265
URL: http://svn.apache.org/viewvc?rev=697265&view=rev
Log:
TIKA-126: Add Parser.parse(InputStream, Metadata) for metadata extraction
Reverted this feature.
The potential gains from this extra method are limited as most parser
implementations will in any case need to parse the entire document even if they
are just extracting metadata.
We can consider restoring this feature when we have more compelling use cases.
For now I think we should rather keep the Parser interface as clean and simple
as possible.
Removed:
incubator/tika/trunk/src/main/java/org/apache/tika/parser/AbstractParser.java
Modified:
incubator/tika/trunk/CHANGES.txt
incubator/tika/trunk/src/main/java/org/apache/tika/parser/CompositeParser.java
incubator/tika/trunk/src/main/java/org/apache/tika/parser/EmptyParser.java
incubator/tika/trunk/src/main/java/org/apache/tika/parser/ErrorParser.java
incubator/tika/trunk/src/main/java/org/apache/tika/parser/ExternalParser.java
incubator/tika/trunk/src/main/java/org/apache/tika/parser/Parser.java
incubator/tika/trunk/src/main/java/org/apache/tika/parser/ParserDecorator.java
incubator/tika/trunk/src/main/java/org/apache/tika/parser/asm/ClassParser.java
incubator/tika/trunk/src/main/java/org/apache/tika/parser/html/HtmlParser.java
incubator/tika/trunk/src/main/java/org/apache/tika/parser/image/ImageParser.java
incubator/tika/trunk/src/main/java/org/apache/tika/parser/microsoft/OfficeParser.java
incubator/tika/trunk/src/main/java/org/apache/tika/parser/mp3/Mp3Parser.java
incubator/tika/trunk/src/main/java/org/apache/tika/parser/opendocument/OpenOfficeParser.java
incubator/tika/trunk/src/main/java/org/apache/tika/parser/pdf/PDFParser.java
incubator/tika/trunk/src/main/java/org/apache/tika/parser/pkg/PackageParser.java
incubator/tika/trunk/src/main/java/org/apache/tika/parser/rtf/RTFParser.java
incubator/tika/trunk/src/main/java/org/apache/tika/parser/txt/TXTParser.java
incubator/tika/trunk/src/main/java/org/apache/tika/parser/xml/XMLParser.java
Modified: incubator/tika/trunk/CHANGES.txt
URL:
http://svn.apache.org/viewvc/incubator/tika/trunk/CHANGES.txt?rev=697265&r1=697264&r2=697265&view=diff
==============================================================================
--- incubator/tika/trunk/CHANGES.txt (original)
+++ incubator/tika/trunk/CHANGES.txt Fri Sep 19 15:06:53 2008
@@ -19,73 +19,70 @@
8. TIKA-112 - Use Commons IO 1.4 (Jukka Zitting)
-9. TIKA-126 - Add Parser.parse(InputStream, Metadata) for metadata extraction
- (Jukka Zitting)
+9. TIKA-127 - Add support for Visio files (Jukka Zitting)
-10. TIKA-127 - Add support for Visio files (Jukka Zitting)
+10. TIKA-129 - node() support for the streaming XPath utility (Jukka Zitting)
-11. TIKA-129 - node() support for the streaming XPath utility (Jukka Zitting)
-
-12. TIKA-130 - self-or-descendant axis does not match self in streaming XPath
+11. TIKA-130 - self-or-descendant axis does not match self in streaming XPath
(Jukka Zitting)
-13. TIKA-131 - Lazy XHTML prefix generation (Jukka Zitting)
+12. TIKA-131 - Lazy XHTML prefix generation (Jukka Zitting)
-14. TIKA-128 - HTML parser should produce XHTML SAX events (Jukka Zitting)
+13. TIKA-128 - HTML parser should produce XHTML SAX events (Jukka Zitting)
-15. TIKA-133 - TeeContentHandler constructor should use varargs (Jukka Zitting)
+14. TIKA-133 - TeeContentHandler constructor should use varargs (Jukka Zitting)
-16. TIKA-132 - Refactor Excel extractor to parse per sheet and add
+15. TIKA-132 - Refactor Excel extractor to parse per sheet and add
hyperlink support (Niall Pemberton)
-17. TIKA-134 - mvn package does not produce packages for bin/src
+16. TIKA-134 - mvn package does not produce packages for bin/src
(Karl Heinz Marbaise)
-18. TIKA-138 - Ignore HTML style and script content (Jukka Zitting)
+17. TIKA-138 - Ignore HTML style and script content (Jukka Zitting)
-19. TIKA-113 - Metadata (such as title) should not be part of content
+18. TIKA-113 - Metadata (such as title) should not be part of content
(Jukka Zitting)
-20. TIKA-139 - Add a composite parser (Jukka Zitting)
+19. TIKA-139 - Add a composite parser (Jukka Zitting)
-21. TIKA-142 - Include application/xhtml+xml as valid mime type for XMLParser
+20. TIKA-142 - Include application/xhtml+xml as valid mime type for XMLParser
(mattmann)
-22. TIKA-143 - Add ParsingReader (Jukka Zitting)
+21. TIKA-143 - Add ParsingReader (Jukka Zitting)
-23. TIKA-144 - Upgrade nekohtml dependency (Jukka Zitting)
+22. TIKA-144 - Upgrade nekohtml dependency (Jukka Zitting)
-24. TIKA-145 - Separate NOTICEs and LICENSEs for binary and source packages
+23. TIKA-145 - Separate NOTICEs and LICENSEs for binary and source packages
(Jukka Zitting)
-25. TIKA-146 - Upgrade to POI 3.1 (Jukka Zitting)
+24. TIKA-146 - Upgrade to POI 3.1 (Jukka Zitting)
-26. TIKA-99 - Support external parser programs (Jukka Zitting)
+25. TIKA-99 - Support external parser programs (Jukka Zitting)
-27. TIKA-149 - Parser for Zip files (Dave Meikle & Jukka Zitting)
+26. TIKA-149 - Parser for Zip files (Dave Meikle & Jukka Zitting)
-28. TIKA-150 - Parser for tar files (Jukka Zitting)
+27. TIKA-150 - Parser for tar files (Jukka Zitting)
-29. TIKA-151 - Stream compression support (Jukka Zitting)
+28. TIKA-151 - Stream compression support (Jukka Zitting)
-30. TIKA-156 - Some MIME magic patterns are ignored by MimeTypes
+29. TIKA-156 - Some MIME magic patterns are ignored by MimeTypes
(Jukka Zitting)
-31. TIKA-155 - Java class file parser (Dave Brosius & Jukka Zitting)
+30. TIKA-155 - Java class file parser (Dave Brosius & Jukka Zitting)
-32. TIKA-108 - New Tika logos (Yongqian Li & Jukka Zitting)
+31. TIKA-108 - New Tika logos (Yongqian Li & Jukka Zitting)
-33. TIKA-120 - Add support for retrieving ID3 tags from MP3 files
+32. TIKA-120 - Add support for retrieving ID3 tags from MP3 files
(Dave Meikle & Jukka Zitting)
-34. TIKA-54 - Outlook msg parser
+33. TIKA-54 - Outlook msg parser
(Rida Benjelloun, Dave Meikle & Jukka Zitting)
-35. TIKA-114 - PDFParser : Getting content of the document using
+34. TIKA-114 - PDFParser : Getting content of the document using
"writer.ToString ()" , some words are stuck together
(Dave Meikle)
-36. TIKA-161 - Enable PMD reports (Jukka Zitting)
+35. TIKA-161 - Enable PMD reports (Jukka Zitting)
Release 0.1-incubating - 12/27/2007
Modified:
incubator/tika/trunk/src/main/java/org/apache/tika/parser/CompositeParser.java
URL:
http://svn.apache.org/viewvc/incubator/tika/trunk/src/main/java/org/apache/tika/parser/CompositeParser.java?rev=697265&r1=697264&r2=697265&view=diff
==============================================================================
---
incubator/tika/trunk/src/main/java/org/apache/tika/parser/CompositeParser.java
(original)
+++
incubator/tika/trunk/src/main/java/org/apache/tika/parser/CompositeParser.java
Fri Sep 19 15:06:53 2008
@@ -102,14 +102,6 @@
/**
* Delegates the call to the matching component parser.
*/
- public void parse(InputStream stream, Metadata metadata)
- throws IOException, TikaException {
- getParser(metadata).parse(stream, metadata);
- }
-
- /**
- * Delegates the call to the matching component parser.
- */
public void parse(
InputStream stream, ContentHandler handler, Metadata metadata)
throws IOException, SAXException, TikaException {
Modified:
incubator/tika/trunk/src/main/java/org/apache/tika/parser/EmptyParser.java
URL:
http://svn.apache.org/viewvc/incubator/tika/trunk/src/main/java/org/apache/tika/parser/EmptyParser.java?rev=697265&r1=697264&r2=697265&view=diff
==============================================================================
--- incubator/tika/trunk/src/main/java/org/apache/tika/parser/EmptyParser.java
(original)
+++ incubator/tika/trunk/src/main/java/org/apache/tika/parser/EmptyParser.java
Fri Sep 19 15:06:53 2008
@@ -30,9 +30,6 @@
*/
public class EmptyParser implements Parser {
- public void parse(InputStream stream, Metadata metadata) {
- }
-
public void parse(
InputStream stream, ContentHandler handler, Metadata metadata)
throws SAXException {
Modified:
incubator/tika/trunk/src/main/java/org/apache/tika/parser/ErrorParser.java
URL:
http://svn.apache.org/viewvc/incubator/tika/trunk/src/main/java/org/apache/tika/parser/ErrorParser.java?rev=697265&r1=697264&r2=697265&view=diff
==============================================================================
--- incubator/tika/trunk/src/main/java/org/apache/tika/parser/ErrorParser.java
(original)
+++ incubator/tika/trunk/src/main/java/org/apache/tika/parser/ErrorParser.java
Fri Sep 19 15:06:53 2008
@@ -30,12 +30,6 @@
public class ErrorParser implements Parser {
public void parse(
- InputStream stream, Metadata metadata)
- throws TikaException {
- throw new TikaException("Parse error");
- }
-
- public void parse(
InputStream stream, ContentHandler handler, Metadata metadata)
throws TikaException {
throw new TikaException("Parse error");
Modified:
incubator/tika/trunk/src/main/java/org/apache/tika/parser/ExternalParser.java
URL:
http://svn.apache.org/viewvc/incubator/tika/trunk/src/main/java/org/apache/tika/parser/ExternalParser.java?rev=697265&r1=697264&r2=697265&view=diff
==============================================================================
---
incubator/tika/trunk/src/main/java/org/apache/tika/parser/ExternalParser.java
(original)
+++
incubator/tika/trunk/src/main/java/org/apache/tika/parser/ExternalParser.java
Fri Sep 19 15:06:53 2008
@@ -43,12 +43,6 @@
private String command = "cat";
/**
- * Does nothing as metadata extraction is not supported by this parser.
- */
- public void parse(InputStream stream, Metadata metadata) {
- }
-
- /**
* Executes the configured external command and passes the given document
* stream as a simple XHTML document to the given SAX content handler.
* No metadata is extracted.
Modified: incubator/tika/trunk/src/main/java/org/apache/tika/parser/Parser.java
URL:
http://svn.apache.org/viewvc/incubator/tika/trunk/src/main/java/org/apache/tika/parser/Parser.java?rev=697265&r1=697264&r2=697265&view=diff
==============================================================================
--- incubator/tika/trunk/src/main/java/org/apache/tika/parser/Parser.java
(original)
+++ incubator/tika/trunk/src/main/java/org/apache/tika/parser/Parser.java Fri
Sep 19 15:06:53 2008
@@ -30,20 +30,6 @@
public interface Parser {
/**
- * Parses document metadata from the given document stream.
- * <p>
- * The given document stream is consumed but not closed by this method.
- * The responsibility to close the stream remains on the caller.
- *
- * @param stream the document stream (input)
- * @param metadata document metadata (input and output)
- * @throws IOException if the document stream could not be read
- * @throws TikaException if the document could not be parsed
- */
- void parse(InputStream stream, Metadata metadata)
- throws IOException, TikaException;
-
- /**
* Parses a document stream into a sequence of XHTML SAX events.
* Fills in related document metadata in the given metadata object.
* <p>
Modified:
incubator/tika/trunk/src/main/java/org/apache/tika/parser/ParserDecorator.java
URL:
http://svn.apache.org/viewvc/incubator/tika/trunk/src/main/java/org/apache/tika/parser/ParserDecorator.java?rev=697265&r1=697264&r2=697265&view=diff
==============================================================================
---
incubator/tika/trunk/src/main/java/org/apache/tika/parser/ParserDecorator.java
(original)
+++
incubator/tika/trunk/src/main/java/org/apache/tika/parser/ParserDecorator.java
Fri Sep 19 15:06:53 2008
@@ -30,7 +30,7 @@
* instance. Subclasses can provide extra decoration by overriding the
* parse method.
*/
-public class ParserDecorator extends AbstractParser {
+public class ParserDecorator implements Parser {
/**
* The decorated parser instance.
Modified:
incubator/tika/trunk/src/main/java/org/apache/tika/parser/asm/ClassParser.java
URL:
http://svn.apache.org/viewvc/incubator/tika/trunk/src/main/java/org/apache/tika/parser/asm/ClassParser.java?rev=697265&r1=697264&r2=697265&view=diff
==============================================================================
---
incubator/tika/trunk/src/main/java/org/apache/tika/parser/asm/ClassParser.java
(original)
+++
incubator/tika/trunk/src/main/java/org/apache/tika/parser/asm/ClassParser.java
Fri Sep 19 15:06:53 2008
@@ -21,7 +21,7 @@
import org.apache.tika.exception.TikaException;
import org.apache.tika.metadata.Metadata;
-import org.apache.tika.parser.AbstractParser;
+import org.apache.tika.parser.Parser;
import org.objectweb.asm.ClassReader;
import org.objectweb.asm.ClassVisitor;
import org.xml.sax.ContentHandler;
@@ -30,7 +30,7 @@
/**
* Parser for Java .class files.
*/
-public class ClassParser extends AbstractParser {
+public class ClassParser implements Parser {
public void parse(
InputStream stream, ContentHandler handler, Metadata metadata)
Modified:
incubator/tika/trunk/src/main/java/org/apache/tika/parser/html/HtmlParser.java
URL:
http://svn.apache.org/viewvc/incubator/tika/trunk/src/main/java/org/apache/tika/parser/html/HtmlParser.java?rev=697265&r1=697264&r2=697265&view=diff
==============================================================================
---
incubator/tika/trunk/src/main/java/org/apache/tika/parser/html/HtmlParser.java
(original)
+++
incubator/tika/trunk/src/main/java/org/apache/tika/parser/html/HtmlParser.java
Fri Sep 19 15:06:53 2008
@@ -26,7 +26,7 @@
import org.apache.commons.io.input.CloseShieldInputStream;
import org.apache.tika.exception.TikaException;
import org.apache.tika.metadata.Metadata;
-import org.apache.tika.parser.AbstractParser;
+import org.apache.tika.parser.Parser;
import org.apache.tika.sax.TeeContentHandler;
import org.apache.tika.sax.TextContentHandler;
import org.apache.tika.sax.WriteOutContentHandler;
@@ -46,7 +46,7 @@
* and post-processes the events to produce XHTML and metadata expected by
* Tika clients.
*/
-public class HtmlParser extends AbstractParser {
+public class HtmlParser implements Parser {
/**
* Set of safe mappings from incoming HTML elements to outgoing
Modified:
incubator/tika/trunk/src/main/java/org/apache/tika/parser/image/ImageParser.java
URL:
http://svn.apache.org/viewvc/incubator/tika/trunk/src/main/java/org/apache/tika/parser/image/ImageParser.java?rev=697265&r1=697264&r2=697265&view=diff
==============================================================================
---
incubator/tika/trunk/src/main/java/org/apache/tika/parser/image/ImageParser.java
(original)
+++
incubator/tika/trunk/src/main/java/org/apache/tika/parser/image/ImageParser.java
Fri Sep 19 15:06:53 2008
@@ -33,8 +33,9 @@
public class ImageParser implements Parser {
- public void parse(InputStream stream, Metadata metadata)
- throws IOException, TikaException {
+ public void parse(
+ InputStream stream, ContentHandler handler, Metadata metadata)
+ throws IOException, SAXException, TikaException {
String type = metadata.get(Metadata.CONTENT_TYPE);
if (type != null) {
Iterator<ImageReader> iterator =
@@ -48,12 +49,7 @@
reader.dispose();
}
}
- }
- public void parse(
- InputStream stream, ContentHandler handler, Metadata metadata)
- throws IOException, SAXException, TikaException {
- parse(stream, metadata);
XHTMLContentHandler xhtml = new XHTMLContentHandler(handler, metadata);
xhtml.startDocument();
xhtml.endDocument();
Modified:
incubator/tika/trunk/src/main/java/org/apache/tika/parser/microsoft/OfficeParser.java
URL:
http://svn.apache.org/viewvc/incubator/tika/trunk/src/main/java/org/apache/tika/parser/microsoft/OfficeParser.java?rev=697265&r1=697264&r2=697265&view=diff
==============================================================================
---
incubator/tika/trunk/src/main/java/org/apache/tika/parser/microsoft/OfficeParser.java
(original)
+++
incubator/tika/trunk/src/main/java/org/apache/tika/parser/microsoft/OfficeParser.java
Fri Sep 19 15:06:53 2008
@@ -53,35 +53,6 @@
DocumentSummaryInformation.DEFAULT_STREAM_NAME;
/**
- * Extracts properties from an MS Document input stream
- */
- public void parse(InputStream stream, Metadata metadata)
- throws IOException, TikaException {
- POIFSFileSystem filesystem = new POIFSFileSystem(stream);
- Iterator<?> entries = filesystem.getRoot().getEntries();
- while (entries.hasNext()) {
- Entry entry = (Entry) entries.next();
- String name = entry.getName();
- if (!(entry instanceof DocumentEntry)) {
- // Skip directory entries
- } else if (SUMMARY_INFORMATION.equals(name)
- || DOCUMENT_SUMMARY_INFORMATION.equals(name)) {
- parse((DocumentEntry) entry, metadata);
- } else if ("WordDocument".equals(name)) {
- setType(metadata, "application/msword");
- } else if ("PowerPoint Document".equals(name)) {
- setType(metadata, "application/vnd.ms-powerpoint");
- } else if ("Workbook".equals(name)) {
- setType(metadata, "application/vnd.ms-excel");
- } else if ("VisioDocument".equals(name)) {
- setType(metadata, "application/vnd.visio");
- } else if (name.startsWith("__substg1.0_")) {
- setType(metadata, "application/vnd.ms-outlook");
- }
- }
- }
-
- /**
* Extracts properties and text from an MS Document input stream
*/
public void parse(
Modified:
incubator/tika/trunk/src/main/java/org/apache/tika/parser/mp3/Mp3Parser.java
URL:
http://svn.apache.org/viewvc/incubator/tika/trunk/src/main/java/org/apache/tika/parser/mp3/Mp3Parser.java?rev=697265&r1=697264&r2=697265&view=diff
==============================================================================
---
incubator/tika/trunk/src/main/java/org/apache/tika/parser/mp3/Mp3Parser.java
(original)
+++
incubator/tika/trunk/src/main/java/org/apache/tika/parser/mp3/Mp3Parser.java
Fri Sep 19 15:06:53 2008
@@ -22,7 +22,7 @@
import org.apache.tika.exception.TikaException;
import org.apache.tika.metadata.Metadata;
-import org.apache.tika.parser.AbstractParser;
+import org.apache.tika.parser.Parser;
import org.apache.tika.sax.XHTMLContentHandler;
import org.xml.sax.ContentHandler;
import org.xml.sax.SAXException;
@@ -33,7 +33,7 @@
*
* @see http://www.id3.org/ID3v1
*/
-public class Mp3Parser extends AbstractParser {
+public class Mp3Parser implements Parser {
/**
* List of predefined genres.
Modified:
incubator/tika/trunk/src/main/java/org/apache/tika/parser/opendocument/OpenOfficeParser.java
URL:
http://svn.apache.org/viewvc/incubator/tika/trunk/src/main/java/org/apache/tika/parser/opendocument/OpenOfficeParser.java?rev=697265&r1=697264&r2=697265&view=diff
==============================================================================
---
incubator/tika/trunk/src/main/java/org/apache/tika/parser/opendocument/OpenOfficeParser.java
(original)
+++
incubator/tika/trunk/src/main/java/org/apache/tika/parser/opendocument/OpenOfficeParser.java
Fri Sep 19 15:06:53 2008
@@ -28,6 +28,7 @@
import org.apache.tika.parser.Parser;
import org.xml.sax.ContentHandler;
import org.xml.sax.SAXException;
+import org.xml.sax.helpers.DefaultHandler;
/**
* OpenOffice parser
@@ -54,21 +55,6 @@
this.content = content;
}
- public void parse(InputStream stream, Metadata metadata)
- throws IOException, TikaException {
- ZipInputStream zip = new ZipInputStream(stream);
- ZipEntry entry = zip.getNextEntry();
- while (entry != null) {
- if (entry.getName().equals("mimetype")) {
- String type = IOUtils.toString(zip, "UTF-8");
- metadata.set(Metadata.CONTENT_TYPE, type);
- } else if (entry.getName().equals("meta.xml")) {
- meta.parse(zip, metadata);
- }
- entry = zip.getNextEntry();
- }
- }
-
public void parse(
InputStream stream, ContentHandler handler, Metadata metadata)
throws IOException, SAXException, TikaException {
@@ -79,7 +65,7 @@
String type = IOUtils.toString(zip, "UTF-8");
metadata.set(Metadata.CONTENT_TYPE, type);
} else if (entry.getName().equals("meta.xml")) {
- meta.parse(zip, metadata);
+ meta.parse(zip, new DefaultHandler(), metadata);
} else if (entry.getName().equals("content.xml")) {
content.parse(zip, handler, metadata);
}
Modified:
incubator/tika/trunk/src/main/java/org/apache/tika/parser/pdf/PDFParser.java
URL:
http://svn.apache.org/viewvc/incubator/tika/trunk/src/main/java/org/apache/tika/parser/pdf/PDFParser.java?rev=697265&r1=697264&r2=697265&view=diff
==============================================================================
---
incubator/tika/trunk/src/main/java/org/apache/tika/parser/pdf/PDFParser.java
(original)
+++
incubator/tika/trunk/src/main/java/org/apache/tika/parser/pdf/PDFParser.java
Fri Sep 19 15:06:53 2008
@@ -34,24 +34,6 @@
*/
public class PDFParser implements Parser {
- public void parse(InputStream stream, Metadata metadata)
- throws IOException, TikaException {
- PDDocument pdfDocument = PDDocument.load(stream);
- try {
- if (pdfDocument.isEncrypted()) {
- try {
- pdfDocument.decrypt("");
- } catch (Exception e) {
- // Ignore
- }
- }
- metadata.add(Metadata.CONTENT_TYPE, "application/pdf");
- extractMetadata(pdfDocument, metadata);
- } finally {
- pdfDocument.close();
- }
- }
-
public void parse(
InputStream stream, ContentHandler handler, Metadata metadata)
throws IOException, SAXException, TikaException {
Modified:
incubator/tika/trunk/src/main/java/org/apache/tika/parser/pkg/PackageParser.java
URL:
http://svn.apache.org/viewvc/incubator/tika/trunk/src/main/java/org/apache/tika/parser/pkg/PackageParser.java?rev=697265&r1=697264&r2=697265&view=diff
==============================================================================
---
incubator/tika/trunk/src/main/java/org/apache/tika/parser/pkg/PackageParser.java
(original)
+++
incubator/tika/trunk/src/main/java/org/apache/tika/parser/pkg/PackageParser.java
Fri Sep 19 15:06:53 2008
@@ -22,7 +22,6 @@
import org.apache.commons.io.input.CloseShieldInputStream;
import org.apache.tika.exception.TikaException;
import org.apache.tika.metadata.Metadata;
-import org.apache.tika.parser.AbstractParser;
import org.apache.tika.parser.AutoDetectParser;
import org.apache.tika.parser.Parser;
import org.apache.tika.sax.BodyContentHandler;
@@ -39,7 +38,7 @@
* (optional) entry name as a <h1> element and the full
* structured body content of the parsed entry.
*/
-public abstract class PackageParser extends AbstractParser {
+public abstract class PackageParser implements Parser {
/**
* The parser instance used to parse package entries.
Modified:
incubator/tika/trunk/src/main/java/org/apache/tika/parser/rtf/RTFParser.java
URL:
http://svn.apache.org/viewvc/incubator/tika/trunk/src/main/java/org/apache/tika/parser/rtf/RTFParser.java?rev=697265&r1=697264&r2=697265&view=diff
==============================================================================
---
incubator/tika/trunk/src/main/java/org/apache/tika/parser/rtf/RTFParser.java
(original)
+++
incubator/tika/trunk/src/main/java/org/apache/tika/parser/rtf/RTFParser.java
Fri Sep 19 15:06:53 2008
@@ -25,7 +25,7 @@
import org.apache.tika.exception.TikaException;
import org.apache.tika.metadata.Metadata;
-import org.apache.tika.parser.AbstractParser;
+import org.apache.tika.parser.Parser;
import org.apache.tika.sax.XHTMLContentHandler;
import org.xml.sax.ContentHandler;
import org.xml.sax.SAXException;
@@ -33,7 +33,7 @@
/**
* RTF parser
*/
-public class RTFParser extends AbstractParser {
+public class RTFParser implements Parser {
public void parse(
InputStream stream, ContentHandler handler, Metadata metadata)
Modified:
incubator/tika/trunk/src/main/java/org/apache/tika/parser/txt/TXTParser.java
URL:
http://svn.apache.org/viewvc/incubator/tika/trunk/src/main/java/org/apache/tika/parser/txt/TXTParser.java?rev=697265&r1=697264&r2=697265&view=diff
==============================================================================
---
incubator/tika/trunk/src/main/java/org/apache/tika/parser/txt/TXTParser.java
(original)
+++
incubator/tika/trunk/src/main/java/org/apache/tika/parser/txt/TXTParser.java
Fri Sep 19 15:06:53 2008
@@ -22,7 +22,7 @@
import org.apache.tika.exception.TikaException;
import org.apache.tika.metadata.Metadata;
-import org.apache.tika.parser.AbstractParser;
+import org.apache.tika.parser.Parser;
import org.apache.tika.sax.XHTMLContentHandler;
import org.apache.tika.utils.Utils;
import org.xml.sax.ContentHandler;
@@ -31,7 +31,7 @@
/**
* Text parser
*/
-public class TXTParser extends AbstractParser {
+public class TXTParser implements Parser {
public void parse(
InputStream stream, ContentHandler handler, Metadata metadata)
Modified:
incubator/tika/trunk/src/main/java/org/apache/tika/parser/xml/XMLParser.java
URL:
http://svn.apache.org/viewvc/incubator/tika/trunk/src/main/java/org/apache/tika/parser/xml/XMLParser.java?rev=697265&r1=697264&r2=697265&view=diff
==============================================================================
---
incubator/tika/trunk/src/main/java/org/apache/tika/parser/xml/XMLParser.java
(original)
+++
incubator/tika/trunk/src/main/java/org/apache/tika/parser/xml/XMLParser.java
Fri Sep 19 15:06:53 2008
@@ -26,7 +26,7 @@
import org.apache.commons.io.input.CloseShieldInputStream;
import org.apache.tika.exception.TikaException;
import org.apache.tika.metadata.Metadata;
-import org.apache.tika.parser.AbstractParser;
+import org.apache.tika.parser.Parser;
import org.apache.tika.sax.TextContentHandler;
import org.apache.tika.sax.XHTMLContentHandler;
import org.xml.sax.ContentHandler;
@@ -36,7 +36,7 @@
/**
* XML parser
*/
-public class XMLParser extends AbstractParser {
+public class XMLParser implements Parser {
public void parse(
InputStream stream, ContentHandler handler, Metadata metadata)
