Author: jnioche
Date: Fri Jul 2 10:53:24 2010
New Revision: 959948
URL: http://svn.apache.org/viewvc?rev=959948&view=rev
Log:
NUTCH-836 : Remove deprecated parse plugins
Removed:
nutch/trunk/src/plugin/lib-jakarta-poi/
nutch/trunk/src/plugin/lib-parsems/
nutch/trunk/src/plugin/parse-html/
nutch/trunk/src/plugin/parse-msexcel/
nutch/trunk/src/plugin/parse-mspowerpoint/
nutch/trunk/src/plugin/parse-msword/
nutch/trunk/src/plugin/parse-oo/
nutch/trunk/src/plugin/parse-pdf/
nutch/trunk/src/plugin/parse-text/
Modified:
nutch/trunk/conf/nutch-default.xml
nutch/trunk/conf/parse-plugins.xml
nutch/trunk/default.properties
nutch/trunk/src/plugin/build.xml
nutch/trunk/src/plugin/creativecommons/build.xml
nutch/trunk/src/plugin/creativecommons/src/test/org/creativecommons/nutch/TestCCParseFilter.java
nutch/trunk/src/plugin/languageidentifier/build.xml
nutch/trunk/src/plugin/parse-tika/plugin.xml
nutch/trunk/src/plugin/parse-zip/build.xml
nutch/trunk/src/test/crawl-tests.xml
nutch/trunk/src/test/org/apache/nutch/parse/TestParserFactory.java
nutch/trunk/src/test/org/apache/nutch/parse/parse-plugin-test.xml
Modified: nutch/trunk/conf/nutch-default.xml
URL:
http://svn.apache.org/viewvc/nutch/trunk/conf/nutch-default.xml?rev=959948&r1=959947&r2=959948&view=diff
==============================================================================
--- nutch/trunk/conf/nutch-default.xml (original)
+++ nutch/trunk/conf/nutch-default.xml Fri Jul 2 10:53:24 2010
@@ -956,17 +956,14 @@
<property>
<name>plugin.includes</name>
-
<value>protocol-http|urlfilter-regex|parse-(text|html|js|tika)|index-(basic|anchor)|query-(basic|site|url)|response-(json|xml)|summary-basic|scoring-opic|urlnormalizer-(pass|regex|basic)</value>
+
<value>protocol-http|urlfilter-regex|parse-tika|index-(basic|anchor)|query-(basic|site|url)|response-(json|xml)|summary-basic|scoring-opic|urlnormalizer-(pass|regex|basic)</value>
<description>Regular expression naming plugin directory names to
include. Any plugin not matching this expression is excluded.
In any case you need at least include the nutch-extensionpoints plugin. By
default Nutch includes crawling just HTML and plain text via HTTP,
and basic indexing and search plugins. In order to use HTTPS please enable
protocol-httpclient, but be aware of possible intermittent problems with the
- underlying commons-httpclient library. Nutch now also includes integration
with Tika
- to leverage Tika's parsing capabilities for multiple content types. The
existing Nutch
- parser implementations will likely be phased out in the next release or so,
as such, it is
- a good idea to begin migrating away from anything not provided by parse-tika.
+ underlying commons-httpclient library.
</description>
</property>
Modified: nutch/trunk/conf/parse-plugins.xml
URL:
http://svn.apache.org/viewvc/nutch/trunk/conf/parse-plugins.xml?rev=959948&r1=959947&r2=959948&view=diff
==============================================================================
--- nutch/trunk/conf/parse-plugins.xml (original)
+++ nutch/trunk/conf/parse-plugins.xml Fri Jul 2 10:53:24 2010
@@ -22,126 +22,10 @@
<parse-plugins>
-<!--
- The following mimetype are now handled by the default parser
(parse-tika).
- You can uncomment the associations below to override parse-tika
- and chose which plugin should be used for a given content type
-
- <mimeType name="application/msword">
- <plugin id="parse-msword" />
- </mimeType>
-
- <mimeType name="application/pdf">
- <plugin id="parse-pdf" />
- </mimeType>
-
- <mimeType name="application/postscript">
- <plugin id="parse-pdf" />
- </mimeType>
-
- <mimeType name="application/vnd.ms-excel">
- <plugin id="parse-msexcel" />
- </mimeType>
-
- <mimeType name="application/vnd.ms-powerpoint">
- <plugin id="parse-mspowerpoint" />
- </mimeType>
-
- <mimeType name="application/vnd.oasis.opendocument.text">
- <plugin id="parse-oo" />
- </mimeType>
-
- <mimeType name="application/vnd.oasis.opendocument.text-template">
- <plugin id="parse-oo" />
- </mimeType>
-
- <mimeType name="application/vnd.oasis.opendocument.text-master">
- <plugin id="parse-oo" />
- </mimeType>
-
- <mimeType name="application/vnd.oasis.opendocument.text-web">
- <plugin id="parse-oo" />
- </mimeType>
-
- <mimeType name="application/vnd.oasis.opendocument.presentation">
- <plugin id="parse-oo" />
- </mimeType>
-
- <mimeType
name="application/vnd.oasis.opendocument.presentation-template">
- <plugin id="parse-oo" />
- </mimeType>
-
- <mimeType name="application/vnd.oasis.opendocument.spreadsheet">
- <plugin id="parse-oo" />
- </mimeType>
-
- <mimeType
name="application/vnd.oasis.opendocument.spreadsheet-template">
- <plugin id="parse-oo" />
- </mimeType>
-
- <mimeType name="application/vnd.sun.xml.calc">
- <plugin id="parse-oo" />
- </mimeType>
-
- <mimeType name="application/vnd.sun.xml.calc.template">
- <plugin id="parse-oo" />
- </mimeType>
-
- <mimeType name="application/vnd.sun.xml.impress">
- <plugin id="parse-oo" />
- </mimeType>
-
- <mimeType name="application/vnd.sun.xml.impress.template">
- <plugin id="parse-oo" />
- </mimeType>
-
- <mimeType name="application/vnd.sun.xml.writer">
- <plugin id="parse-oo" />
- </mimeType>
-
- <mimeType name="application/vnd.sun.xml.writer.template">
- <plugin id="parse-oo" />
- </mimeType>
-
- <mimeType name="application/x-csh">
- <plugin id="parse-text" />
- </mimeType>
-
- <mimeType name="application/x-kword">
- <plugin id="parse-msword" />
- </mimeType>
-
- <mimeType name="application/x-kspread">
- <plugin id="parse-msexcel" />
- </mimeType>
-
- <mimeType name="text/html">
- <plugin id="parse-html" />
- </mimeType>
-
- <mimeType name="text/plain">
- <plugin id="parse-text" />
- </mimeType>
-
- <mimeType name="text/richtext">
- <plugin id="parse-tika" />
- </mimeType>
-
- <mimeType name="text/rtf">
- <plugin id="parse-tika" />
- </mimeType>
-
- <mimeType name="text/sgml">
- <plugin id="parse-html" />
- </mimeType>
-
- <mimeType name="text/tab-separated-values">
- <plugin id="parse-msexcel" />
- </mimeType>
--->
-
- <mimeType name="application/xhtml+xml">
- <plugin id="parse-html" />
+ <!-- by default if the mimeType is set to *, or
+ if it can't be determined, use parse-tika -->
+ <mimeType name="*">
+ <plugin id="parse-tika" />
</mimeType>
<mimeType name="application/rss+xml">
@@ -172,9 +56,9 @@
</mimeType>
<mimeType name="text/xml">
- <plugin id="parse-html" />
+ <plugin id="parse-tika" />
<plugin id="parse-rss" />
- <plugin id="feed" />
+ <plugin id="feed" />
</mimeType>
<!-- Types for parse-ext plugin: required for unit tests to pass. -->
@@ -190,30 +74,16 @@
<!-- alias mappings for parse-xxx names to the actual extension
implementation
ids described in each plugin's plugin.xml file -->
<aliases>
- <alias name="parse-tika"
- extension-id="org.apache.nutch.parse.tika.Parser" />
+ <alias name="parse-tika"
+ extension-id="org.apache.nutch.parse.tika.TikaParser" />
<alias name="parse-ext" extension-id="ExtParser" />
- <alias name="parse-html"
- extension-id="org.apache.nutch.parse.html.HtmlParser" />
<alias name="parse-js" extension-id="JSParser" />
- <alias name="parse-msexcel"
-
extension-id="org.apache.nutch.parse.msexcel.MSExcelParser" />
- <alias name="parse-mspowerpoint"
-
extension-id="org.apache.nutch.parse.mspowerpoint.MSPowerPointParser" />
- <alias name="parse-msword"
-
extension-id="org.apache.nutch.parse.msword.MSWordParser" />
- <alias name="parse-oo"
-
extension-id="org.apache.nutch.parse.oo.OpenDocument.Text" />
- <alias name="parse-pdf"
- extension-id="org.apache.nutch.parse.pdf.PdfParser" />
<alias name="parse-rss"
extension-id="org.apache.nutch.parse.rss.RSSParser" />
<alias name="feed"
extension-id="org.apache.nutch.parse.feed.FeedParser" />
<alias name="parse-swf"
extension-id="org.apache.nutch.parse.swf.SWFParser" />
- <alias name="parse-text"
- extension-id="org.apache.nutch.parse.text.TextParser" />
<alias name="parse-zip"
extension-id="org.apache.nutch.parse.zip.ZipParser" />
</aliases>
Modified: nutch/trunk/default.properties
URL:
http://svn.apache.org/viewvc/nutch/trunk/default.properties?rev=959948&r1=959947&r2=959948&view=diff
==============================================================================
--- nutch/trunk/default.properties (original)
+++ nutch/trunk/default.properties Fri Jul 2 10:53:24 2010
@@ -52,8 +52,7 @@ javac.version= 1.5
#
plugins.api=\
org.apache.nutch.protocol.http.api*:\
- org.apache.nutch.urlfilter.api*:\
- org.apache.nutch.parse.ms*
+ org.apache.nutch.urlfilter.api*
#
# Protocol Plugins
@@ -83,18 +82,10 @@ plugins.scoring=\
#
plugins.parse=\
org.apache.nutch.parse.ext*:\
- org.apache.nutch.parse.html*:\
org.apache.nutch.parse.js:\
- org.apache.nutch.parse.mp3*:\
- org.apache.nutch.parse.msexcel*:\
- org.apache.nutch.parse.mspowerpoint*:\
- org.apache.nutch.parse.msword*:\
- org.apache.nutch.parse.oo*:\
- org.apache.nutch.parse.pdf*:\
- org.apache.nutch.parse.rtf*:\
org.apache.nutch.parse.rss*:\
org.apache.nutch.parse.swf*:\
- org.apache.nutch.parse.text:\
+ org.apache.nutch.parse.tika:\
org.apache.nutch.parse.zip
#
Modified: nutch/trunk/src/plugin/build.xml
URL:
http://svn.apache.org/viewvc/nutch/trunk/src/plugin/build.xml?rev=959948&r1=959947&r2=959948&view=diff
==============================================================================
--- nutch/trunk/src/plugin/build.xml (original)
+++ nutch/trunk/src/plugin/build.xml Fri Jul 2 10:53:24 2010
@@ -36,10 +36,8 @@
<ant dir="field-boost" target="deploy"/>
<ant dir="languageidentifier" target="deploy"/>
<ant dir="lib-http" target="deploy"/>
- <ant dir="lib-jakarta-poi" target="deploy"/>
<ant dir="lib-lucene-analyzers" target="deploy"/>
<ant dir="lib-nekohtml" target="deploy"/>
- <ant dir="lib-parsems" target="deploy"/>
<ant dir="lib-regex-filter" target="deploy"/>
<ant dir="lib-xml" target="deploy"/>
<ant dir="microformats-reltag" target="deploy"/>
@@ -50,16 +48,9 @@
<ant dir="protocol-http" target="deploy"/>
<ant dir="protocol-httpclient" target="deploy"/>
<ant dir="parse-ext" target="deploy"/>
- <ant dir="parse-html" target="deploy"/>
<ant dir="parse-js" target="deploy"/>
- <ant dir="parse-msexcel" target="deploy"/>
- <ant dir="parse-mspowerpoint" target="deploy"/>
- <ant dir="parse-msword" target="deploy"/>
- <ant dir="parse-oo" target="deploy"/>
- <ant dir="parse-pdf" target="deploy"/>
<ant dir="parse-rss" target="deploy"/>
<ant dir="parse-swf" target="deploy"/>
- <ant dir="parse-text" target="deploy"/>
<ant dir="parse-tika" target="deploy"/>
<ant dir="parse-zip" target="deploy"/>
<ant dir="query-basic" target="deploy"/>
@@ -98,12 +89,6 @@
<ant dir="ontology" target="test"/>
<ant dir="protocol-httpclient" target="test"/>
<!--ant dir="parse-ext" target="test"/-->
- <ant dir="parse-html" target="test"/>
- <ant dir="parse-msexcel" target="test"/>
- <ant dir="parse-mspowerpoint" target="test"/>
- <ant dir="parse-msword" target="test"/>
- <ant dir="parse-oo" target="test"/>
- <ant dir="parse-pdf" target="test"/>
<ant dir="parse-rss" target="test"/>
<ant dir="feed" target="test"/>
<ant dir="parse-swf" target="test"/>
@@ -138,10 +123,8 @@
<ant dir="languageidentifier" target="clean"/>
<ant dir="lib-commons-httpclient" target="clean"/>
<ant dir="lib-http" target="clean"/>
- <ant dir="lib-jakarta-poi" target="clean"/>
<ant dir="lib-lucene-analyzers" target="clean"/>
<ant dir="lib-nekohtml" target="clean"/>
- <ant dir="lib-parsems" target="clean"/>
<ant dir="lib-regex-filter" target="clean"/>
<ant dir="lib-xml" target="clean"/>
<ant dir="microformats-reltag" target="clean"/>
@@ -152,18 +135,9 @@
<ant dir="protocol-http" target="clean"/>
<ant dir="protocol-httpclient" target="clean"/>
<ant dir="parse-ext" target="clean"/>
- <ant dir="parse-html" target="clean"/>
<ant dir="parse-js" target="clean"/>
- <ant dir="parse-mp3" target="clean"/>
- <ant dir="parse-msexcel" target="clean"/>
- <ant dir="parse-mspowerpoint" target="clean"/>
- <ant dir="parse-msword" target="clean"/>
- <ant dir="parse-oo" target="clean"/>
- <ant dir="parse-pdf" target="clean"/>
<ant dir="parse-rss" target="clean"/>
- <ant dir="parse-rtf" target="clean"/>
<ant dir="parse-swf" target="clean"/>
- <ant dir="parse-text" target="clean"/>
<ant dir="parse-tika" target="clean"/>
<ant dir="parse-zip" target="clean"/>
<ant dir="query-basic" target="clean"/>
Modified: nutch/trunk/src/plugin/creativecommons/build.xml
URL:
http://svn.apache.org/viewvc/nutch/trunk/src/plugin/creativecommons/build.xml?rev=959948&r1=959947&r2=959948&view=diff
==============================================================================
--- nutch/trunk/src/plugin/creativecommons/build.xml (original)
+++ nutch/trunk/src/plugin/creativecommons/build.xml Fri Jul 2 10:53:24 2010
@@ -22,7 +22,7 @@
<!-- Deploy Unit test dependencies -->
<target name="deps-test">
<ant target="deploy" inheritall="false" dir="../nutch-extensionpoints"/>
- <ant target="deploy" inheritall="false" dir="../parse-html"/>
+ <!-- <ant target="deploy" inheritall="false" dir="../parse-html"/> -->
</target>
Modified:
nutch/trunk/src/plugin/creativecommons/src/test/org/creativecommons/nutch/TestCCParseFilter.java
URL:
http://svn.apache.org/viewvc/nutch/trunk/src/plugin/creativecommons/src/test/org/creativecommons/nutch/TestCCParseFilter.java?rev=959948&r1=959947&r2=959948&view=diff
==============================================================================
---
nutch/trunk/src/plugin/creativecommons/src/test/org/creativecommons/nutch/TestCCParseFilter.java
(original)
+++
nutch/trunk/src/plugin/creativecommons/src/test/org/creativecommons/nutch/TestCCParseFilter.java
Fri Jul 2 10:53:24 2010
@@ -38,10 +38,14 @@ public class TestCCParseFilter extends T
public void testPages() throws Exception {
pageTest(new File(testDir, "anchor.html"), "http://foo.com/";,
"http://creativecommons.org/licenses/by-nc-sa/1.0";, "a", null);
+ // Tika returns <a> whereas parse-html returns <rel>
+ // check later
pageTest(new File(testDir, "rel.html"), "http://foo.com/";,
- "http://creativecommons.org/licenses/by-nc/2.0";, "rel", null);
+ "http://creativecommons.org/licenses/by-nc/2.0";, "a", null);
+ // Tika returns <a> whereas parse-html returns <rdf>
+ // check later
pageTest(new File(testDir, "rdf.html"), "http://foo.com/";,
- "http://creativecommons.org/licenses/by-nc/1.0";, "rdf", "text");
+ "http://creativecommons.org/licenses/by-nc/1.0";, "a", null);
}
public void pageTest(File file, String url,
@@ -62,8 +66,8 @@ public class TestCCParseFilter extends T
Content content =
new Content(url, url, bytes, contentType, new Metadata(), conf);
- Parse parse = new ParseUtil(conf).parseByExtensionId("parse-html",
content).get(content.getUrl());
-
+ Parse parse = new ParseUtil(conf).parse(content).get(content.getUrl());
+
Metadata metadata = parse.getData().getParseMeta();
assertEquals(license, metadata.get("License-Url"));
assertEquals(location, metadata.get("License-Location"));
Modified: nutch/trunk/src/plugin/languageidentifier/build.xml
URL:
http://svn.apache.org/viewvc/nutch/trunk/src/plugin/languageidentifier/build.xml?rev=959948&r1=959947&r2=959948&view=diff
==============================================================================
--- nutch/trunk/src/plugin/languageidentifier/build.xml (original)
+++ nutch/trunk/src/plugin/languageidentifier/build.xml Fri Jul 2 10:53:24 2010
@@ -33,7 +33,6 @@
<!-- Deploy Unit test dependencies -->
<target name="deps-test">
<ant target="deploy" inheritall="false" dir="../nutch-extensionpoints"/>
- <ant target="deploy" inheritall="false" dir="../parse-html"/>
</target>
</project>
Modified: nutch/trunk/src/plugin/parse-tika/plugin.xml
URL:
http://svn.apache.org/viewvc/nutch/trunk/src/plugin/parse-tika/plugin.xml?rev=959948&r1=959947&r2=959948&view=diff
==============================================================================
--- nutch/trunk/src/plugin/parse-tika/plugin.xml (original)
+++ nutch/trunk/src/plugin/parse-tika/plugin.xml Fri Jul 2 10:53:24 2010
@@ -58,7 +58,7 @@
id="org.apache.nutch.parse.tika"
name="TikaParser">
- <implementation id="org.apache.nutch.parse.tika.Parser"
+ <implementation id="org.apache.nutch.parse.tika.TikaParser"
class="org.apache.nutch.parse.tika.TikaParser">
<parameter name="contentType" value="*"/>
</implementation>
Modified: nutch/trunk/src/plugin/parse-zip/build.xml
URL:
http://svn.apache.org/viewvc/nutch/trunk/src/plugin/parse-zip/build.xml?rev=959948&r1=959947&r2=959948&view=diff
==============================================================================
--- nutch/trunk/src/plugin/parse-zip/build.xml (original)
+++ nutch/trunk/src/plugin/parse-zip/build.xml Fri Jul 2 10:53:24 2010
@@ -23,7 +23,7 @@
<target name="deps-test">
<ant target="deploy" inheritall="false" dir="../nutch-extensionpoints"/>
<ant target="deploy" inheritall="false" dir="../protocol-file"/>
- <ant target="deploy" inheritall="false" dir="../parse-text"/>
+ <!-- <ant target="deploy" inheritall="false" dir="../parse-text"/>-->
</target>
Modified: nutch/trunk/src/test/crawl-tests.xml
URL:
http://svn.apache.org/viewvc/nutch/trunk/src/test/crawl-tests.xml?rev=959948&r1=959947&r2=959948&view=diff
==============================================================================
--- nutch/trunk/src/test/crawl-tests.xml (original)
+++ nutch/trunk/src/test/crawl-tests.xml Fri Jul 2 10:53:24 2010
@@ -6,7 +6,7 @@
<property>
<name>plugin.includes</name>
- <value>parse-html|protocol-http|urlfilter-suffix|scoring-opic</value>
+ <value>parse-tika|protocol-http|urlfilter-suffix|scoring-opic</value>
<description>Enable required plugins.</description>
</property>
Modified: nutch/trunk/src/test/org/apache/nutch/parse/TestParserFactory.java
URL:
http://svn.apache.org/viewvc/nutch/trunk/src/test/org/apache/nutch/parse/TestParserFactory.java?rev=959948&r1=959947&r2=959948&view=diff
==============================================================================
--- nutch/trunk/src/test/org/apache/nutch/parse/TestParserFactory.java
(original)
+++ nutch/trunk/src/test/org/apache/nutch/parse/TestParserFactory.java Fri Jul
2 10:53:24 2010
@@ -50,9 +50,9 @@ public class TestParserFactory extends T
/** Unit test for <code>getExtensions(String)</code> method. */
public void testGetExtensions() throws Exception {
Extension ext = (Extension)parserFactory.getExtensions("text/html").get(0);
- assertEquals("parse-html", ext.getDescriptor().getPluginId());
+ assertEquals("parse-tika", ext.getDescriptor().getPluginId());
ext = (Extension) parserFactory.getExtensions("text/html;
charset=ISO-8859-1").get(0);
- assertEquals("parse-html", ext.getDescriptor().getPluginId());
+ assertEquals("parse-tika", ext.getDescriptor().getPluginId());
ext = (Extension)parserFactory.getExtensions("foo/bar").get(0);
assertEquals("parse-tika", ext.getDescriptor().getPluginId());
}
@@ -62,14 +62,14 @@ public class TestParserFactory extends T
Parser [] parsers = parserFactory.getParsers("text/html",
"http://foo.com";);
assertNotNull(parsers);
assertEquals(1, parsers.length);
- assertEquals("org.apache.nutch.parse.html.HtmlParser",
+ assertEquals("org.apache.nutch.parse.tika.TikaParser",
parsers[0].getClass().getName());
parsers = parserFactory.getParsers("text/html; charset=ISO-8859-1",
"http://foo.com";);
assertNotNull(parsers);
assertEquals(1, parsers.length);
- assertEquals("org.apache.nutch.parse.html.HtmlParser",
+ assertEquals("org.apache.nutch.parse.tika.TikaParser",
parsers[0].getClass().getName());
parsers = parserFactory.getParsers("application/x-javascript",
@@ -82,7 +82,7 @@ public class TestParserFactory extends T
parsers = parserFactory.getParsers("text/plain", "http://foo.com";);
assertNotNull(parsers);
assertEquals(1, parsers.length);
- assertEquals("org.apache.nutch.parse.text.TextParser",
+ assertEquals("org.apache.nutch.parse.tika.TikaParser",
parsers[0].getClass().getName());
Parser parser1 = parserFactory.getParsers("text/plain",
"http://foo.com";)[0];
Modified: nutch/trunk/src/test/org/apache/nutch/parse/parse-plugin-test.xml
URL:
http://svn.apache.org/viewvc/nutch/trunk/src/test/org/apache/nutch/parse/parse-plugin-test.xml?rev=959948&r1=959947&r2=959948&view=diff
==============================================================================
--- nutch/trunk/src/test/org/apache/nutch/parse/parse-plugin-test.xml (original)
+++ nutch/trunk/src/test/org/apache/nutch/parse/parse-plugin-test.xml Fri Jul
2 10:53:24 2010
@@ -22,9 +22,9 @@
<parse-plugins>
<!-- by default if the mimeType is set to *, or
- can't be determined, use parse-text -->
+ if it can't be determined, use parse-tika -->
<mimeType name="*">
- <plugin id="parse-text" />
+ <plugin id="parse-tika" />
</mimeType>
<!-- test these 4 plugins -->
@@ -34,12 +34,6 @@
! it should not block the process and then the next one is used
!-->
<plugin id="parse-plugin-that-not-exist"/>
- <plugin id="parse-html"/>
- </mimeType>
-
- <mimeType name="text/plain">
- <!-- Test that an extension-id can be directly used here -->
- <plugin id="org.apache.nutch.parse.text.TextParser"/>
</mimeType>
<mimeType name="application/x-javascript">
@@ -47,19 +41,18 @@
</mimeType>
<mimeType name="text/rss">
- <plugin id="parse-rss"/>
+ <!-- Test that an extension-id can be directly used here -->
+ <plugin id="org.apache.nutch.parse.rss.RSSParser"/>
</mimeType>
<!-- alias mappings for parse-xxx names to the actual extension
implementation
ids described in each plugin's plugin.xml file -->
<aliases>
- <alias name="parse-html"
- extension-id="org.apache.nutch.parse.html.HtmlParser" />
<alias name="parse-js"
extension-id="JSParser" />
<alias name="parse-rss"
extension-id="org.apache.nutch.parse.rss.RSSParser" />
- <alias name="parse-text"
- extension-id="org.apache.nutch.parse.text.TextParser" />
+ <alias name="parse-tika"
+ extension-id="org.apache.nutch.parse.tika.TikaParser" />
</aliases>
</parse-plugins>
