Merge pull request #775 from tballison/NUTCH-2998

NUTCH-2998 -- Remove Any23 from Nutch

LICENSE-binary

@@ -327,11 +327,6 @@ net.sourceforge.owlapi:owlapi-impl
 net.sourceforge.owlapi:owlapi-parsers
 net.sourceforge.owlapi:owlapi-rio
 net.sourceforge.owlapi:owlapi-tools
-org.apache.any23:apache-any23-api
-org.apache.any23:apache-any23-core
-org.apache.any23:apache-any23-csvutils
-org.apache.any23:apache-any23-encoding
-org.apache.any23:apache-any23-mime
 org.apache.avro:avro
 org.apache.commons:commons-collections4
 org.apache.commons:commons-compress

NOTICE-binary

@@ -29,7 +29,7 @@ code and source code.
 
 The following provides more details on the included cryptographic software:
 
-The plugins parse-tika and any23 use Apache Tika and the Bouncy Castle
+The parse-tika plugin uses Apache Tika and the Bouncy Castle
 generic encryption libraries for extracting text content and metadata
 from encrypted PDF files. See <https://www.bouncycastle.org/> for more
 details on Bouncy Castle and <https://tika.apache.org/> for details
@@ -46,9 +46,6 @@ on Apache Tika.
 Apache projects
 ---------------
 
-Apache Any23 (https://any23.apache.org/)
-  see https://github.com/apache/any23/blob/master/NOTICE.txt
-
 Apache Avro (https://avro.apache.org)
   see https://github.com/apache/avro/blob/master/NOTICE.txt
 

NOTICE.txt

@@ -29,7 +29,7 @@ code and source code.
 
 The following provides more details on the included cryptographic software:
 
-The plugins parse-tika and any23 use Apache Tika and the Bouncy Castle
+The parse-tika plugin uses Apache Tika and the Bouncy Castle
 generic encryption libraries for extracting text content and metadata
 from encrypted PDF files. See <https://www.bouncycastle.org/> for more
 details on Bouncy Castle and <https://tika.apache.org/> for details

build.xml

@@ -202,7 +202,6 @@
       <arg value="--no-module-directories" if:set="using.jdk.11"/>
 
       <packageset dir="${src.dir}"/>
-      <packageset dir="${plugins.dir}/any23/src/java/"/>
       <packageset dir="${plugins.dir}/creativecommons/src/java"/>
       <packageset dir="${plugins.dir}/feed/src/java"/>
       <packageset dir="${plugins.dir}/headings/src/java"/>
@@ -687,7 +686,6 @@
       <arg value="--no-module-directories" if:set="using.jdk.11"/>
 
       <packageset dir="${src.dir}"/>
-      <packageset dir="${plugins.dir}/any23/src/java/" />
       <packageset dir="${plugins.dir}/creativecommons/src/java"/>
       <packageset dir="${plugins.dir}/feed/src/java"/>
       <packageset dir="${plugins.dir}/headings/src/java"/>
@@ -772,7 +770,6 @@
       <classpath>
         <fileset dir="${build.plugins}" >
           <include name="**/*.jar"/>
-          <exclude name="any23/javax.annotation-api*.jar"/>
         </fileset>
       </classpath>
 
@@ -1180,8 +1177,6 @@
         <source path="${basedir}/src/java/" />
         <source path="${basedir}/src/test/" output="build/test/classes" />
 
-        <source path="${plugins.dir}/any23/src/java/" />
-        <source path="${plugins.dir}/any23/src/test/" />
         <source path="${plugins.dir}/creativecommons/src/java/" />
         <source path="${plugins.dir}/creativecommons/src/test/" />
         <source path="${plugins.dir}/feed/src/java/" />

conf/nutch-default.xml

@@ -1353,22 +1353,6 @@
   </description>
 </property>
 
-
-
-<!--  any23 plugin properties -->
-
-<property>
-    <name>any23.extractors</name>
-    <value>html-microdata</value>
-    <description>Comma-separated list of Any23 extractors (a list of extractors is available here: http://any23.apache.org/getting-started.html)</description>
-</property>
-
-<property>
-    <name>any23.content_types</name>
-    <value>text/html,application/xhtml+xml</value>
-    <description>Comma-separated list of content-types onto which Any23 extractors should be applied (see http://www.iana.org/assignments/media-types/). If empty, all content-types are supported.</description>
-</property>
-
 <!-- moreindexingfilter plugin properties -->
 
 <property>

default.properties

@@ -210,6 +210,4 @@ plugins.misc=\
    org.apache.nutch.collection*:\
    org.apache.nutch.analysis.lang*:\
    org.creativecommons.nutch*:\
-   org.apache.nutch.microformats.reltag*:\
-   org.apache.nutch.any23*
-
+   org.apache.nutch.microformats.reltag*: