Ubuntu环境下：Hadoop, HBase, nutch各个配置文件一览

 1 # Set Hadoop-specific environment variables here.
 2 
 3 # The only required environment variable is JAVA_HOME.  All others are
 4 # optional.  When running a distributed configuration it is best to
 5 # set JAVA_HOME in this file, so that it is correctly defined on
 6 # remote nodes.
 7 
 8 # The java implementation to use.  Required.
 9 # export JAVA_HOME=/usr/lib/j2sdk1.5-sun
10 
11 # Extra Java CLASSPATH elements.  Optional.
12 # export HADOOP_CLASSPATH=
13 
14 # The maximum amount of heap to use, in MB. Default is 1000.
15 # export HADOOP_HEAPSIZE=2000
16 
17 # Extra Java runtime options.  Empty by default.
18 # export HADOOP_OPTS=-server
19 
20 # Command specific options appended to HADOOP_OPTS when specified
21 export HADOOP_NAMENODE_OPTS="-Dcom.sun.management.jmxremote $HADOOP_NAMENODE_OPTS"
22 export HADOOP_SECONDARYNAMENODE_OPTS="-Dcom.sun.management.jmxremote $HADOOP_SECONDARYNAMENODE_OPTS"
23 export HADOOP_DATANODE_OPTS="-Dcom.sun.management.jmxremote $HADOOP_DATANODE_OPTS"
24 export HADOOP_BALANCER_OPTS="-Dcom.sun.management.jmxremote $HADOOP_BALANCER_OPTS"
25 export HADOOP_JOBTRACKER_OPTS="-Dcom.sun.management.jmxremote $HADOOP_JOBTRACKER_OPTS"
26 # export HADOOP_TASKTRACKER_OPTS=
27 # The following applies to multiple commands (fs, dfs, fsck, distcp etc)
28 # export HADOOP_CLIENT_OPTS
29 
30 # Extra ssh options.  Empty by default.
31 # export HADOOP_SSH_OPTS="-o ConnectTimeout=1 -o SendEnv=HADOOP_CONF_DIR"
32 
33 # Where log files are stored.  $HADOOP_HOME/logs by default.
34 # export HADOOP_LOG_DIR=${HADOOP_HOME}/logs
35 
36 # File naming remote slave hosts.  $HADOOP_HOME/conf/slaves by default.
37 # export HADOOP_SLAVES=${HADOOP_HOME}/conf/slaves
38 
39 # host:path where hadoop code should be rsync‘d from.  Unset by default.
40 # export HADOOP_MASTER=master:/home/$USER/src/hadoop
41 
42 # Seconds to sleep between slave commands.  Unset by default.  This
43 # can be useful in large clusters, where, e.g., slave rsyncs can
44 # otherwise arrive faster than the master can service them.
45 # export HADOOP_SLAVE_SLEEP=0.1
46 
47 # The directory where pid files are stored. /tmp by default.
48 # NOTE: this should be set to a directory that can only be written to by 
49 #       the users that are going to run the hadoop daemons.  Otherwise there is
50 #       the potential for a symlink attack.
51 # export HADOOP_PID_DIR=/var/hadoop/pids
52 
53 # A string representing this instance of hadoop. $USER by default.
54 # export HADOOP_IDENT_STRING=$USER
55 
56 # The scheduling priority for daemon processes.  See ‘man nice‘.
57 # export HADOOP_NICENESS=10
58 
59 export JAVA_HOME=/usr/lib/jvm 
60 
61 export HADOOP_HOME=/home/xxx0624/hadoop
62 
63 export PATH=$PATH:/home/xxx0624/hadoop/bin
64 
65 export HBASE_CLASSPATH=/home/xxx0624/hadoop/conf

  1 #
  2 #/**
  3 # * Copyright 2007 The Apache Software Foundation
  4 # *
  5 # * Licensed to the Apache Software Foundation (ASF) under one
  6 # * or more contributor license agreements.  See the NOTICE file
  7 # * distributed with this work for additional information
  8 # * regarding copyright ownership.  The ASF licenses this file
  9 # * to you under the Apache License, Version 2.0 (the
 10 # * "License"); you may not use this file except in compliance
 11 # * with the License.  You may obtain a copy of the License at
 12 # *
 13 # *     http://www.apache.org/licenses/LICENSE-2.0
 14 # *
 15 # * Unless required by applicable law or agreed to in writing, software
 16 # * distributed under the License is distributed on an "AS IS" BASIS,
 17 # * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 18 # * See the License for the specific language governing permissions and
 19 # * limitations under the License.
 20 # */
 21 
 22 # Set environment variables here.
 23 
 24 # This script sets variables multiple times over the course of starting an hbase process,
 25 # so try to keep things idempotent unless you want to take an even deeper look
 26 # into the startup scripts (bin/hbase, etc.)
 27 
 28 # The java implementation to use.  Java 1.6 required.
 29 # export JAVA_HOME=/usr/java/jdk1.6.0/
 30 
 31 # Extra Java CLASSPATH elements.  Optional.
 32 # export HBASE_CLASSPATH=
 33 
 34 # The maximum amount of heap to use, in MB. Default is 1000.
 35 # export HBASE_HEAPSIZE=1000
 36 
 37 # Extra Java runtime options.
 38 # Below are what we set by default.  May only work with SUN JVM.
 39 # For more on why as well as other possible settings,
 40 # see http://wiki.apache.org/hadoop/PerformanceTuning
 41 export HBASE_OPTS="-XX:+UseConcMarkSweepGC"
 42 
 43 # Uncomment one of the below three options to enable java garbage collection logging for the server-side processes.
 44 
 45 # This enables basic gc logging to the .out file.
 46 # export SERVER_GC_OPTS="-verbose:gc -XX:+PrintGCDetails -XX:+PrintGCDateStamps"
 47 
 48 # This enables basic gc logging to its own file.
 49 # If FILE-PATH is not replaced, the log file(.gc) would still be generated in the HBASE_LOG_DIR .
 50 # export SERVER_GC_OPTS="-verbose:gc -XX:+PrintGCDetails -XX:+PrintGCDateStamps -Xloggc:<FILE-PATH>"
 51 
 52 # This enables basic GC logging to its own file with automatic log rolling. Only applies to jdk 1.6.0_34+ and 1.7.0_2+.
 53 # If FILE-PATH is not replaced, the log file(.gc) would still be generated in the HBASE_LOG_DIR .
 54 # export SERVER_GC_OPTS="-verbose:gc -XX:+PrintGCDetails -XX:+PrintGCDateStamps -Xloggc:<FILE-PATH> -XX:+UseGCLogFileRotation -XX:NumberOfGCLogFiles=1 -XX:GCLogFileSize=512M"
 55 
 56 # Uncomment one of the below three options to enable java garbage collection logging for the client processes.
 57 
 58 # This enables basic gc logging to the .out file.
 59 # export CLIENT_GC_OPTS="-verbose:gc -XX:+PrintGCDetails -XX:+PrintGCDateStamps"
 60 
 61 # This enables basic gc logging to its own file.
 62 # If FILE-PATH is not replaced, the log file(.gc) would still be generated in the HBASE_LOG_DIR .
 63 # export CLIENT_GC_OPTS="-verbose:gc -XX:+PrintGCDetails -XX:+PrintGCDateStamps -Xloggc:<FILE-PATH>"
 64 
 65 # This enables basic GC logging to its own file with automatic log rolling. Only applies to jdk 1.6.0_34+ and 1.7.0_2+.
 66 # If FILE-PATH is not replaced, the log file(.gc) would still be generated in the HBASE_LOG_DIR .
 67 # export CLIENT_GC_OPTS="-verbose:gc -XX:+PrintGCDetails -XX:+PrintGCDateStamps -Xloggc:<FILE-PATH> -XX:+UseGCLogFileRotation -XX:NumberOfGCLogFiles=1 -XX:GCLogFileSize=512M"
 68 
 69 # Uncomment below if you intend to use the EXPERIMENTAL off heap cache.
 70 # export HBASE_OPTS="$HBASE_OPTS -XX:MaxDirectMemorySize="
 71 # Set hbase.offheapcache.percentage in hbase-site.xml to a nonzero value.
 72 
 73 
 74 # Uncomment and adjust to enable JMX exporting
 75 # See jmxremote.password and jmxremote.access in $JRE_HOME/lib/management to configure remote password access.
 76 # More details at: http://java.sun.com/javase/6/docs/technotes/guides/management/agent.html
 77 #
 78 # export HBASE_JMX_BASE="-Dcom.sun.management.jmxremote.ssl=false -Dcom.sun.management.jmxremote.authenticate=false"
 79 # export HBASE_MASTER_OPTS="$HBASE_MASTER_OPTS $HBASE_JMX_BASE -Dcom.sun.management.jmxremote.port=10101"
 80 # export HBASE_REGIONSERVER_OPTS="$HBASE_REGIONSERVER_OPTS $HBASE_JMX_BASE -Dcom.sun.management.jmxremote.port=10102"
 81 # export HBASE_THRIFT_OPTS="$HBASE_THRIFT_OPTS $HBASE_JMX_BASE -Dcom.sun.management.jmxremote.port=10103"
 82 # export HBASE_ZOOKEEPER_OPTS="$HBASE_ZOOKEEPER_OPTS $HBASE_JMX_BASE -Dcom.sun.management.jmxremote.port=10104"
 83 
 84 # File naming hosts on which HRegionServers will run.  $HBASE_HOME/conf/regionservers by default.
 85 # export HBASE_REGIONSERVERS=${HBASE_HOME}/conf/regionservers
 86 
 87 # File naming hosts on which backup HMaster will run.  $HBASE_HOME/conf/backup-masters by default.
 88 # export HBASE_BACKUP_MASTERS=${HBASE_HOME}/conf/backup-masters
 89 
 90 # Extra ssh options.  Empty by default.
 91 # export HBASE_SSH_OPTS="-o ConnectTimeout=1 -o SendEnv=HBASE_CONF_DIR"
 92 
 93 # Where log files are stored.  $HBASE_HOME/logs by default.
 94 # export HBASE_LOG_DIR=${HBASE_HOME}/logs
 95 
 96 # Enable remote JDWP debugging of major HBase processes. Meant for Core Developers 
 97 # export HBASE_MASTER_OPTS="$HBASE_MASTER_OPTS -Xdebug -Xrunjdwp:transport=dt_socket,server=y,suspend=n,address=8070"
 98 # export HBASE_REGIONSERVER_OPTS="$HBASE_REGIONSERVER_OPTS -Xdebug -Xrunjdwp:transport=dt_socket,server=y,suspend=n,address=8071"
 99 # export HBASE_THRIFT_OPTS="$HBASE_THRIFT_OPTS -Xdebug -Xrunjdwp:transport=dt_socket,server=y,suspend=n,address=8072"
100 # export HBASE_ZOOKEEPER_OPTS="$HBASE_ZOOKEEPER_OPTS -Xdebug -Xrunjdwp:transport=dt_socket,server=y,suspend=n,address=8073"
101 
102 # A string representing this instance of hbase. $USER by default.
103 # export HBASE_IDENT_STRING=$USER
104 
105 # The scheduling priority for daemon processes.  See ‘man nice‘.
106 # export HBASE_NICENESS=10
107 
108 # The directory where pid files are stored. /tmp by default.
109 # export HBASE_PID_DIR=/var/hadoop/pids
110 
111 # Seconds to sleep between slave commands.  Unset by default.  This
112 # can be useful in large clusters, where, e.g., slave rsyncs can
113 # otherwise arrive faster than the master can service them.
114 # export HBASE_SLAVE_SLEEP=0.1
115 
116 # Tell HBase whether it should manage it‘s own instance of Zookeeper or not.
117 # export HBASE_MANAGES_ZK=true
118 
119 export JAVA_HOME=/usr/lib/jvm
120 export HBASE_MANAGES_ZK=true
121 export HBASE_CLASSPATH=/home/xxx0624/hadoop/conf

 1 # Licensed to the Apache Software Foundation (ASF) under one or more
 2 # contributor license agreements.  See the NOTICE file distributed with
 3 # this work for additional information regarding copyright ownership.
 4 # The ASF licenses this file to You under the Apache License, Version 2.0
 5 # (the "License"); you may not use this file except in compliance with
 6 # the License.  You may obtain a copy of the License at
 7 #
 8 #     http://www.apache.org/licenses/LICENSE-2.0
 9 #
10 # Unless required by applicable law or agreed to in writing, software
11 # distributed under the License is distributed on an "AS IS" BASIS,
12 # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
13 # See the License for the specific language governing permissions and
14 # limitations under the License.
15 
16 #gora.datastore.default=org.apache.gora.mock.store.MockDataStore
17 #gora.datastore.autocreateschema=true
18 
19 ###############################
20 # Default SqlStore properties #
21 ###############################
22 
23 #gora.sqlstore.jdbc.driver=org.hsqldb.jdbc.JDBCDriver
24 #gora.sqlstore.jdbc.url=jdbc:hsqldb:hsql://localhost/nutchtest
25 #gora.sqlstore.jdbc.user=sa
26 #gora.sqlstore.jdbc.password=
27 
28 ################################
29 # Default AvroStore properties #
30 ################################
31 
32 # gora.avrostore.codec.type=BINARY||JSON
33 # gora.avrostore.output.path=file:///tmp/gora.avrostore.test.output
34 
35 ################################
36 # DatafileAvroStore properties #
37 ################################
38 # DataFileAvroStore is file based store which uses Avro‘s 
39 # DataFile{Writer,Reader}‘s as a backend. This datastore supports 
40 # mapreduce.
41 
42 # gora.datafileavrostore.###=
43 
44 #########################
45 # HBaseStore properties #
46 #########################
47 # HBase requires that the Configuration has a valid "hbase.zookeeper.quorum"
48 # property. It should be included within hbase-site.xml on the classpath. When
49 # this property is omitted, it expects Zookeeper to run on localhost:2181.
50 
51 # To greatly improve scan performance, increase the hbase-site Configuration
52 # property "hbase.client.scanner.caching". This sets the number of rows to grab
53 # per request.
54 
55 # HBase autoflushing. Enabling autoflush decreases write performance. 
56 # Available since Gora 0.2. Defaults to disabled.
57 # hbase.client.autoflush.default=false
58 
59 #############################
60 # CassandraStore properties #
61 #############################
62 
63 # gora.cassandrastore.servers=localhost:9160
64 
65 #######################
66 # MemStore properties #
67 #######################
68 # This is a memory based {@link DataStore} implementation for tests.
69 
70 # gora.memstore.###=
71 
72 ############################
73 # AccumuloStore properties #
74 ############################
75 gora.datastore.default=org.apache.gora.hbase.store.HBaseStore
76 #gora.datastore.default=org.apache.gora.accumulo.store.AccumuloStore
77 #gora.datastore.accumulo.mock=true
78 #gora.datastore.accumulo.instance=a14
79 #gora.datastore.accumulo.zookeepers=localhost
80 #gora.datastore.accumulo.user=root
81 #gora.datastore.accumulo.password=secret

  1 <?xml version="1.0"?>
  2 <!--
  3  Licensed to the Apache Software Foundation (ASF) under one or more
  4  contributor license agreements.  See the NOTICE file distributed with
  5  this work for additional information regarding copyright ownership.
  6  The ASF licenses this file to You under the Apache License, Version 2.0
  7  (the "License"); you may not use this file except in compliance with
  8  the License.  You may obtain a copy of the License at
  9 
 10      http://www.apache.org/licenses/LICENSE-2.0
 11 
 12  Unless required by applicable law or agreed to in writing, software
 13  distributed under the License is distributed on an "AS IS" BASIS,
 14  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 15  See the License for the specific language governing permissions and
 16  limitations under the License.
 17 -->
 18 
 19 <project name="${name}" default="runtime" xmlns:ivy="antlib:org.apache.ivy.ant" xmlns:artifact="antlib:org.apache.maven.artifact.ant">
 20 
 21  <!-- Load all the default properties, and any the user wants -->
 22  <!-- to contribute (without having to type -D or edit this file -->
 23  <property file="${user.home}/build.properties" />
 24  <property file="${basedir}/build.properties" />
 25  <property file="${basedir}/default.properties" />
 26  <property name="test.junit.output.format" value="plain" />
 27  <property name="release.dir" value="${build.dir}/release" />
 28 
 29  <!-- define Maven coordinates, repository url and artifacts name etc -->
 30  <property name="groupId" value="org.apache.nutch" />
 31  <property name="artifactId" value="nutch" />
 32  <property name="maven-repository-url" value="https://repository.apache.org/service/local/staging/deploy/maven2" />
 33  <property name="maven-repository-id" value="apache.releases.https" />
 34  <property name="maven-jar" value="${release.dir}/${artifactId}-${version}.jar" />
 35  <property name="maven-javadoc-jar" value="${release.dir}/${artifactId}-${version}-javadoc.jar" />
 36  <property name="maven-sources-jar" value="${release.dir}/${artifactId}-${version}-sources.jar" />
 37 
 38 
 39  <!-- the normal classpath -->
 40  <path id="classpath">
 41   <pathelement location="${build.classes}" />
 42   <fileset dir="${build.lib.dir}">
 43    <include name="*.jar" />
 44   </fileset>
 45  </path>
 46 
 47  <presetdef name="javac">
 48    <javac includeantruntime="false" />
 49  </presetdef>
 50 
 51  <!-- the unit test classpath -->
 52  <dirname property="plugins.classpath.dir" file="${build.plugins}" />
 53  <path id="test.classpath">
 54   <pathelement location="${test.build.classes}" />
 55   <pathelement location="${conf.dir}" />
 56   <pathelement location="${test.src.dir}" />
 57   <pathelement location="${plugins.classpath.dir}" />
 58   <path refid="classpath" />
 59   <pathelement location="${build.dir}/${final.name}.job" />
 60   <fileset dir="${build.lib.dir}">
 61    <include name="*.jar" />
 62   </fileset>
 63  </path>
 64 
 65  <!-- ====================================================== -->
 66  <!-- Stuff needed by all targets -->
 67  <!-- ====================================================== -->
 68  <target name="init" depends="ivy-init" description="--> stuff required by all targets">
 69   <mkdir dir="${build.dir}" />
 70   <mkdir dir="${build.classes}" />
 71   <mkdir dir="${release.dir}" />
 72 
 73   <mkdir dir="${test.build.dir}" />
 74   <mkdir dir="${test.build.classes}" />
 75 
 76   <touch datetime="01/25/1971 2:00 pm">
 77    <fileset dir="${conf.dir}" includes="**/*.template" />
 78   </touch>
 79 
 80   <copy todir="${conf.dir}" verbose="true">
 81    <fileset dir="${conf.dir}" includes="**/*.template" />
 82    <mapper type="glob" from="*.template" to="*" />
 83   </copy>
 84  </target>
 85 
 86  <!-- ====================================================== -->
 87  <!-- Compile the Java files -->
 88  <!-- ====================================================== -->
 89  <target name="compile" depends="compile-core, compile-plugins" description="--> compile all Java files"/>
 90 
 91  <target name="compile-core" depends="init, resolve-default" description="--> compile core Java files only">
 92   <javac 
 93       encoding="${build.encoding}" 
 94       srcdir="${src.dir}"
 95     includes="org/apache/nutch/**/*.java" 
 96       destdir="${build.classes}"
 97     debug="${javac.debug}" 
 98       optimize="${javac.optimize}" 
 99       target="${javac.version}"
100     source="${javac.version}" 
101       deprecation="${javac.deprecation}">
102    <compilerarg value="-Xlint:-path"/>
103    <classpath refid="classpath" />
104   </javac>
105  </target>
106 
107  <target name="compile-plugins" depends="init, resolve-default" description="--> compile plugins only">
108   <ant dir="src/plugin" target="deploy" inheritAll="false" />
109  </target>
110 
111  <!-- ================================================================== -->
112  <!-- Make nutch.jar -->
113  <!-- ================================================================== -->
114  <!-- -->
115  <!-- ================================================================== -->
116  <target name="jar" depends="compile-core" description="--> make nutch.jar">
117   <copy file="${conf.dir}/nutch-default.xml" todir="${build.classes}" />
118   <copy file="${conf.dir}/nutch-site.xml" todir="${build.classes}" />
119   <jar jarfile="${build.dir}/${final.name}.jar" basedir="${build.classes}">
120    <manifest>
121    </manifest>
122   </jar>
123  </target>
124 
125  <!-- ================================================================== -->
126  <!-- Make Maven Central Release -->
127  <!-- ================================================================== -->
128  <!-- -->
129  <!-- ================================================================== -->
130  <target name="release" depends="compile-core"
131   description="--> generate the release distribution">
132   <copy file="${conf.dir}/nutch-default.xml" todir="${build.classes}" />
133   <copy file="${conf.dir}/nutch-site.xml" todir="${build.classes}" />
134 
135   <!-- build the main artifact -->
136   <jar jarfile="${maven-jar}" basedir="${build.classes}" />
137 
138   <!-- build the javadoc artifact -->
139   <javadoc 
140       destdir="${release.dir}/javadoc" 
141       overview="${src.dir}/overview.html"
142     author="true" 
143       version="true" 
144       use="true" 
145       windowtitle="${name} ${version} API"
146     doctitle="${name} ${version} API"
147     bottom="Copyright &amp;copy; ${year} The Apache Software Foundation">
148    <arg value="${javadoc.proxy.host}" />
149    <arg value="${javadoc.proxy.port}" />
150 
151    <packageset dir="${src.dir}"/>
152       <packageset dir="${plugins.dir}/creativecommons/src/java"/>
153       <packageset dir="${plugins.dir}/feed/src/java"/>
154       <packageset dir="${plugins.dir}/index-anchor/src/java"/>
155       <packageset dir="${plugins.dir}/index-basic/src/java"/>
156       <packageset dir="${plugins.dir}/index-more/src/java"/>
157       <packageset dir="${plugins.dir}/language-identifier/src/java"/>
158       <packageset dir="${plugins.dir}/lib-http/src/java"/>
159       <packageset dir="${plugins.dir}/lib-regex-filter/src/java"/>
160       <packageset dir="${plugins.dir}/microformats-reltag/src/java"/>
161       <packageset dir="${plugins.dir}/parse-ext/src/java"/>
162       <packageset dir="${plugins.dir}/parse-html/src/java"/>
163       <packageset dir="${plugins.dir}/parse-js/src/java"/>
164       <packageset dir="${plugins.dir}/parse-swf/src/java"/>
165       <packageset dir="${plugins.dir}/parse-tika/src/java"/>
166       <packageset dir="${plugins.dir}/parse-zip/src/java"/>
167       <packageset dir="${plugins.dir}/protocol-file/src/java"/>
168       <packageset dir="${plugins.dir}/protocol-ftp/src/java"/>
169       <packageset dir="${plugins.dir}/protocol-http/src/java"/>
170       <packageset dir="${plugins.dir}/protocol-httpclient/src/java"/>
171       <packageset dir="${plugins.dir}/protocol-sftp/src/java"/>
172       <packageset dir="${plugins.dir}/scoring-link/src/java"/>
173       <packageset dir="${plugins.dir}/scoring-opic/src/java"/>
174       <packageset dir="${plugins.dir}/subcollection/src/java"/>
175       <packageset dir="${plugins.dir}/tld/src/java"/>
176       <packageset dir="${plugins.dir}/urlfilter-automaton/src/java"/>
177       <packageset dir="${plugins.dir}/urlfilter-domain/src/java"/>
178       <packageset dir="${plugins.dir}/urlfilter-prefix/src/java"/>
179       <packageset dir="${plugins.dir}/urlfilter-regex/src/java"/>
180       <packageset dir="${plugins.dir}/urlfilter-suffix/src/java"/>
181       <packageset dir="${plugins.dir}/urlfilter-validator/src/java"/>
182       <packageset dir="${plugins.dir}/urlnormalizer-basic/src/java"/>
183       <packageset dir="${plugins.dir}/urlnormalizer-pass/src/java"/>
184       <packageset dir="${plugins.dir}/urlnormalizer-regex/src/java"/>
185 
186    <link href="${javadoc.link.java}" />
187    <link href="${javadoc.link.lucene}" />
188    <link href="${javadoc.link.hadoop}" />
189 
190    <classpath refid="classpath" />
191    <classpath>
192     <fileset dir="${plugins.dir}">
193      <include name="**/*.jar" />
194     </fileset>
195    </classpath>
196 
197    <group title="Core" packages="org.apache.nutch.*" />
198    <group title="Plugins API" packages="${plugins.api}" />
199    <group title="Protocol Plugins" packages="${plugins.protocol}" />
200    <group title="URL Filter Plugins" packages="${plugins.urlfilter}" />
201    <group title="Scoring Plugins" packages="${plugins.scoring}" />
202    <group title="Parse Plugins" packages="${plugins.parse}" />
203    <group title="Indexing Filter Plugins" packages="${plugins.index}" />
204    <group title="Misc. Plugins" packages="${plugins.misc}" />
205   </javadoc>
206   <jar jarfile="${maven-javadoc-jar}">
207    <fileset dir="${release.dir}/javadoc" />
208   </jar>
209 
210   <!-- build the sources artifact -->
211   <jar jarfile="${maven-sources-jar}">
212    <fileset dir="${src.dir}" />
213   </jar>
214  </target>
215 
216  <!-- ================================================================== -->
217  <!-- Deploy to Apache Nexus -->
218  <!-- ================================================================== -->
219  <!-- -->
220  <!-- ================================================================== -->
221  <target name="deploy" depends="release" description="--> deploy to Apache Nexus">
222 
223   <!-- generate a pom file -->
224   <ivy:makepom ivyfile="${ivy.file}" pomfile="${basedir}/pom.xml"
225    templatefile="ivy/mvn.template">
226    <mapping conf="default" scope="compile" />
227    <mapping conf="runtime" scope="runtime" />
228   </ivy:makepom>
229 
230   <!-- sign and deploy the main artifact -->
231   <artifact:mvn>
232    <arg
233     value="org.apache.maven.plugins:maven-gpg-plugin:1.4:sign-and-deploy-file" />
234    <arg value="-Durl=${maven-repository-url}" />
235    <arg value="-DrepositoryId=${maven-repository-id}" />
236    <arg value="-DpomFile=pom.xml" />
237    <arg value="-Dfile=${maven-jar}" />
238    <arg value="-Papache-release" />
239   </artifact:mvn>
240 
241   <!-- sign and deploy the sources artifact -->
242   <artifact:mvn>
243    <arg value="org.apache.maven.plugins:maven-gpg-plugin:1.4:sign-and-deploy-file" />
244    <arg value="-Durl=${maven-repository-url}" />
245    <arg value="-DrepositoryId=${maven-repository-id}" />
246    <arg value="-DpomFile=pom.xml" />
247    <arg value="-Dfile=${maven-sources-jar}" />
248    <arg value="-Dclassifier=sources" />
249    <arg value="-Papache-release" />
250   </artifact:mvn>
251 
252   <!-- sign and deploy the javadoc artifact -->
253   <artifact:mvn>
254    <arg value="org.apache.maven.plugins:maven-gpg-plugin:1.4:sign-and-deploy-file" />
255    <arg value="-Durl=${maven-repository-url}" />
256    <arg value="-DrepositoryId=${maven-repository-id}" />
257    <arg value="-DpomFile=pom.xml" />
258    <arg value="-Dfile=${maven-javadoc-jar}" />
259    <arg value="-Dclassifier=javadoc" />
260    <arg value="-Papache-release" />
261   </artifact:mvn>
262  </target>
263 
264  <!-- ================================================================== -->
265  <!-- Make job jar -->
266  <!-- ================================================================== -->
267  <!-- -->
268  <!-- ================================================================== -->
269  <target name="job" depends="compile" description="--> make nutch.job jar">
270   <jar jarfile="${build.dir}/${final.name}.job">
271    <!-- 
272    If the build.classes has the nutch config files because the jar command 
273    command has run, exclude them. The conf directory has them. 
274    -->
275    <zipfileset dir="${build.classes}" excludes="nutch-default.xml,nutch-site.xml" />
276    <zipfileset dir="${conf.dir}" excludes="*.template,hadoop*.*" />
277    <!-- 
278    need to exclude hsqldb.jar due to a conflicting version already present 
279    in Hadoop/lib. 
280    -->
281    <zipfileset dir="${build.lib.dir}" prefix="lib" includes="**/*.jar"
282     excludes="jasper*.jar,jsp-*.jar,hadoop-*.jar,hbase*test*.jar,ant*jar,hsqldb*.jar" />
283    <zipfileset dir="${build.plugins}" prefix="classes/plugins" />
284   </jar>
285  </target>
286 
287  <target name="runtime" depends="jar, job" description="--> default target for running Nutch">
288   <mkdir dir="${runtime.dir}" />
289   <mkdir dir="${runtime.local}" />
290   <mkdir dir="${runtime.deploy}" />
291   <!-- deploy area -->
292   <copy file="${build.dir}/${final.name}.job" todir="${runtime.deploy}" />
293   <copy todir="${runtime.deploy}/bin">
294    <fileset dir="src/bin" />
295   </copy>
296   <chmod perm="ugo+x" type="file">
297    <fileset dir="${runtime.deploy}/bin" />
298   </chmod>
299   <!-- local area -->
300   <copy file="${build.dir}/${final.name}.jar" todir="${runtime.local}/lib" />
301   <copy todir="${runtime.local}/lib/native">
302    <fileset dir="lib/native" />
303   </copy>
304   <copy todir="${runtime.local}/conf">
305    <fileset dir="${conf.dir}" excludes="*.template" />
306   </copy>
307   <copy todir="${runtime.local}/bin">
308    <fileset dir="src/bin" />
309   </copy>
310   <chmod perm="ugo+x" type="file">
311    <fileset dir="${runtime.local}/bin" />
312   </chmod>
313   <copy todir="${runtime.local}/lib">
314    <fileset dir="${build.dir}/lib"
315     excludes="ant*.jar,jasper*.jar,jsp-*.jar,hadoop*test*.jar,hbase*test*.jar" />
316   </copy>
317   <copy todir="${runtime.local}/plugins">
318    <fileset dir="${build.dir}/plugins" />
319   </copy>
320   <copy todir="${runtime.local}/test">
321    <fileset dir="${build.dir}/test" />
322   </copy>
323  </target>
324 
325  <!-- ================================================================== -->
326  <!-- Compile test code -->
327  <!-- ================================================================== -->
328  <target name="compile-core-test" depends="compile-core, resolve-test" description="--> compile test code">
329   <javac 
330       encoding="${build.encoding}" 
331       srcdir="${test.src.dir}"
332     includes="org/apache/nutch*/**/*.java" 
333       destdir="${test.build.classes}"
334     debug="${javac.debug}" 
335       optimize="${javac.optimize}" 
336       target="${javac.version}"
337     source="${javac.version}" 
338       deprecation="${javac.deprecation}">
339    <compilerarg value="-Xlint:-path"/>
340    <classpath refid="test.classpath" />
341   </javac>
342  </target>
343 
344  <!-- ================================================================== -->
345  <!-- Run Nutch proxy -->
346  <!-- ================================================================== -->
347 
348  <target name="proxy" depends="job, compile-core-test" description="--> run nutch proxy">
349   <java classname="org.apache.nutch.tools.proxy.TestbedProxy" fork="true">
350    <classpath refid="test.classpath" />
351    <arg value="-fake" />
352    <!-- 
353    <arg value="-delay"/> 
354    <arg value="-200"/> 
355    -->
356    <jvmarg line="-Djavax.xml.parsers.DocumentBuilderFactory=com.sun.org.apache.xerces.internal.jaxp.DocumentBuilderFactoryImpl" />
357   </java>
358  </target>
359 
360  <!-- ================================================================== -->
361  <!-- Run Nutch benchmarking analysis -->
362  <!-- ================================================================== -->
363 
364  <target name="benchmark" description="--> run nutch benchmarking analysis">
365   <java classname="org.apache.nutch.tools.Benchmark" fork="true">
366    <classpath refid="test.classpath" />
367    <jvmarg line="-Xmx512m -Djavax.xml.parsers.DocumentBuilderFactory=com.sun.org.apache.xerces.internal.jaxp.DocumentBuilderFactoryImpl" />
368    <arg value="-maxPerHost" />
369    <arg value="10" />
370    <arg value="-seeds" />
371    <arg value="1" />
372    <arg value="-depth" />
373    <arg value="5" />
374   </java>
375  </target>
376 
377 
378  <!-- ================================================================== -->
379  <!-- Run unit tests -->
380  <!-- ================================================================== -->
381  <target name="test" depends="test-core, test-plugins" description="--> run JUnit tests"/>
382 
383  <target name="test-core" depends="job, compile-core-test" description="--> run core JUnit tests only">
384 
385   <delete dir="${test.build.data}" />
386   <mkdir dir="${test.build.data}" />
387   <!-- 
388   copy resources needed in junit tests 
389   -->
390   <copy todir="${test.build.data}">
391    <fileset dir="src/testresources" includes="**/*" />
392   </copy>
393   <copy file="${test.src.dir}/nutch-site.xml" 
394       todir="${test.build.classes}" />
395 
396   <copy file="${test.src.dir}/log4j.properties" 
397       todir="${test.build.classes}" />
398       
399   <copy file="${test.src.dir}/gora.properties" 
400       todir="${test.build.classes}" />
401 
402   <copy file="${test.src.dir}/crawl-tests.xml"
403         todir="${test.build.classes}"/>
404         
405   <copy file="${test.src.dir}/domain-urlfilter.txt"
406         todir="${test.build.classes}"/>
407 
408   <copy file="${test.src.dir}/filter-all.txt"
409         todir="${test.build.classes}"/>
410 
411   <junit printsummary="yes" haltonfailure="no" fork="yes" dir="${basedir}"
412    errorProperty="tests.failed" failureProperty="tests.failed" maxmemory="1000m">
413    <sysproperty key="test.build.data" value="${test.build.data}" />
414    <sysproperty key="test.src.dir" value="${test.src.dir}" />
415    <sysproperty key="javax.xml.parsers.DocumentBuilderFactory" value="com.sun.org.apache.xerces.internal.jaxp.DocumentBuilderFactoryImpl" />
416    <classpath refid="test.classpath" />
417    <formatter type="${test.junit.output.format}" />
418    <batchtest todir="${test.build.dir}" unless="testcase">
419     <fileset dir="${test.src.dir}" 
420         includes="**/Test*.java" excludes="**/${test.exclude}.java" />
421    </batchtest>
422    <batchtest todir="${test.build.dir}" if="testcase">
423     <fileset dir="${test.src.dir}" includes="**/${testcase}.java" />
424    </batchtest>
425   </junit>
426 
427   <fail if="tests.failed">Tests failed!</fail>
428 
429  </target>
430 
431  <target name="test-plugins" depends="compile" description="--> run plugin JUnit tests only">
432   <ant dir="src/plugin" target="test" inheritAll="false" />
433  </target>
434 
435  <target name="nightly" depends="test, tar-src, zip-src, javadoc" description="--> run the nightly target build">
436  </target>
437 
438  <!-- ================================================================== -->
439  <!-- Ivy targets -->
440  <!-- ================================================================== -->
441 
442  <!-- target: resolve ================================================= -->
443  <target name="resolve-default" depends="clean-lib, init" description="--> resolve and retrieve dependencies with ivy">
444   <ivy:resolve file="${ivy.file}" conf="default" log="download-only" />
445   <ivy:retrieve pattern="${build.lib.dir}/[artifact]-[revision].[ext]" symlink="false" log="quiet" />
446   <antcall target="copy-libs" />
447  </target>
448 
449  <target name="resolve-test" depends="clean-lib, init" description="--> resolve and retrieve dependencies with ivy">
450   <ivy:resolve file="${ivy.file}" conf="test" log="download-only" />
451   <ivy:retrieve pattern="${build.lib.dir}/[artifact]-[revision].[ext]" symlink="false" log="quiet" />
452   <antcall target="copy-libs" />
453  </target>
454 
455  <target name="copy-libs" description="--> copy the libs in lib, which are not ivy enabled">
456   <!-- copy the libs in lib, which are not ivy enabled -->
457   <copy todir="${build.lib.dir}/" failonerror="false">
458    <fileset dir="${lib.dir}" includes="**/*.jar" />
459   </copy>
460  </target>
461 
462  <!-- target: publish-local =========================================== -->
463  <target name="publish-local" depends="jar" description="--> publish this project in the local ivy repository">
464   <ivy:publish artifactspattern="${build.dir}/[artifact]-${version}.[ext]"
465     resolver="local" 
466       pubrevision="${version}" 
467       pubdate="${now}" 
468       status="integration"
469     forcedeliver="true" overwrite="true" />
470   <echo message="project ${ant.project.name} published locally with version ${version}" />
471  </target>
472 
473  <!-- target: report ================================================== -->
474  <target name="report" depends="resolve-test" description="--> generates a report of dependencies">
475   <ivy:report todir="${build.dir}" />
476  </target>
477 
478  <!-- target: ivy-init ================================================ -->
479  <target name="ivy-init" depends="ivy-probe-antlib, ivy-init-antlib" description="--> initialise Ivy settings">
480   <ivy:settings file="${ivy.dir}/ivysettings.xml" />
481  </target>
482 
483  <!-- target: ivy-probe-antlib ======================================== -->
484  <target name="ivy-probe-antlib" description="--> probe the antlib library">
485   <condition property="ivy.found">
486    <typefound uri="antlib:org.apache.ivy.ant" name="cleancache" />
487   </condition>
488  </target>
489 
490  <!-- target: ivy-download ============================================ -->
491  <target name="ivy-download" description="--> download ivy">
492   <available file="${ivy.jar}" property="ivy.jar.found" />
493   <antcall target="ivy-download-unchecked" />
494  </target>
495 
496  <!-- target: ivy-download-unchecked ================================== -->
497  <target name="ivy-download-unchecked" unless="ivy.jar.found" description="--> fetch any ivy file">
498   <get src="${ivy.repo.url}" dest="${ivy.jar}" usetimestamp="true" />
499  </target>
500 
501  <!-- target: ivy-init-antlib ========================================= -->
502  <target name="ivy-init-antlib" depends="ivy-download" unless="ivy.found" description="--> attempt to use Ivy with Antlib">
503   <typedef uri="antlib:org.apache.ivy.ant" onerror="fail"
504    loaderRef="ivyLoader">
505    <classpath>
506     <pathelement location="${ivy.jar}" />
507    </classpath>
508   </typedef>
509   <fail>
510    <condition>
511     <not>
512      <typefound uri="antlib:org.apache.ivy.ant" name="cleancache" />
513     </not>
514    </condition>
515    You need Apache Ivy 2.0 or later from http://ant.apache.org/
516    It could not be loaded from ${ivy.repo.url}
517   </fail>
518  </target>
519     
520   <target name="compile-avro-schema" depends="resolve-default" description="--> compile the avro schema(s) in src/gora/*.avsc">
521     <typedef name="schema" 
522                classname="org.apache.avro.specific.SchemaTask"
523                classpathref="classpath" />
524     
525       <mkdir dir="${build.gora}" />
526       <schema destdir="${build.gora}">
527         <fileset dir="./src/gora">
528           <include name="**/*.avsc"/>
529         </fileset>
530       </schema>
531       
532     </target>
533 
534    <!-- ====================================================== -->
535   <!-- Generate the Java files from the GORA schemas          -->
536   <!-- Will call this automatically later                     -->
537   <!-- ====================================================== -->
538   <target name="generate-gora-src" depends="init" description="--> compile the avro schema(s) in src/gora/*.avsc">
539     <java classname="org.apache.gora.compiler.GoraCompiler">
540      <classpath refid="classpath"/>
541      <arg value="src/gora/"/>
542      <arg value="${src.dir}"/>
543     </java>
544  </target>
545 
546  <!-- ================================================================== -->
547  <!-- Documentation -->
548  <!-- ================================================================== -->
549  <target name="javadoc" depends="compile" description="--> generate Javadoc">
550   <mkdir dir="${build.javadoc}" />
551   <javadoc 
552       overview="${src.dir}/overview.html" 
553       destdir="${build.javadoc}"
554     author="true" 
555       version="true" 
556       use="true" 
557       windowtitle="${name} ${version} API"
558     doctitle="${name} ${version} API"
559     bottom="Copyright &amp;copy; ${year} The Apache Software Foundation">
560    <arg value="${javadoc.proxy.host}" />
561    <arg value="${javadoc.proxy.port}" />
562 
563    <packageset dir="${src.dir}"/>
564       <packageset dir="${plugins.dir}/creativecommons/src/java"/>
565       <packageset dir="${plugins.dir}/feed/src/java"/>
566       <packageset dir="${plugins.dir}/index-anchor/src/java"/>
567       <packageset dir="${plugins.dir}/index-basic/src/java"/>
568       <packageset dir="${plugins.dir}/index-more/src/java"/>
569       <packageset dir="${plugins.dir}/language-identifier/src/java"/>
570       <packageset dir="${plugins.dir}/lib-http/src/java"/>
571       <packageset dir="${plugins.dir}/lib-regex-filter/src/java"/>
572       <packageset dir="${plugins.dir}/microformats-reltag/src/java"/>
573       <packageset dir="${plugins.dir}/parse-ext/src/java"/>
574       <packageset dir="${plugins.dir}/parse-html/src/java"/>
575       <packageset dir="${plugins.dir}/parse-js/src/java"/>
576       <packageset dir="${plugins.dir}/parse-swf/src/java"/>
577       <packageset dir="${plugins.dir}/parse-tika/src/java"/>
578       <packageset dir="${plugins.dir}/parse-zip/src/java"/>
579       <packageset dir="${plugins.dir}/protocol-file/src/java"/>
580       <packageset dir="${plugins.dir}/protocol-ftp/src/java"/>
581       <packageset dir="${plugins.dir}/protocol-http/src/java"/>
582       <packageset dir="${plugins.dir}/protocol-httpclient/src/java"/>
583       <packageset dir="${plugins.dir}/protocol-sftp/src/java"/>
584       <packageset dir="${plugins.dir}/scoring-link/src/java"/>
585       <packageset dir="${plugins.dir}/scoring-opic/src/java"/>
586       <packageset dir="${plugins.dir}/subcollection/src/java"/>
587       <packageset dir="${plugins.dir}/tld/src/java"/>
588       <packageset dir="${plugins.dir}/urlfilter-automaton/src/java"/>
589       <packageset dir="${plugins.dir}/urlfilter-domain/src/java"/>
590       <packageset dir="${plugins.dir}/urlfilter-prefix/src/java"/>
591       <packageset dir="${plugins.dir}/urlfilter-regex/src/java"/>
592       <packageset dir="${plugins.dir}/urlfilter-suffix/src/java"/>
593       <packageset dir="${plugins.dir}/urlfilter-validator/src/java"/>
594       <packageset dir="${plugins.dir}/urlnormalizer-basic/src/java"/>
595       <packageset dir="${plugins.dir}/urlnormalizer-pass/src/java"/>
596       <packageset dir="${plugins.dir}/urlnormalizer-regex/src/java"/>
597 
598    <link href="${javadoc.link.java}" />
599    <link href="${javadoc.link.lucene}" />
600    <link href="${javadoc.link.hadoop}" />
601 
602    <classpath refid="classpath" />
603    <classpath>
604     <fileset dir="${plugins.dir}">
605      <include name="**/*.jar" />
606     </fileset>
607    </classpath>
608 
609    <group title="Core" packages="org.apache.nutch.*" />
610    <group title="Plugins API" packages="${plugins.api}" />
611    <group title="Protocol Plugins" packages="${plugins.protocol}" />
612    <group title="URL Filter Plugins" packages="${plugins.urlfilter}" />
613    <group title="Scoring Plugins" packages="${plugins.scoring}" />
614    <group title="Parse Plugins" packages="${plugins.parse}" />
615    <group title="Indexing Filter Plugins" packages="${plugins.index}" />
616    <group title="Misc. Plugins" packages="${plugins.misc}" />
617   </javadoc>
618   <!-- Copy the plugin.dtd file to the plugin doc-files dir -->
619   <copy file="${plugins.dir}/plugin.dtd" todir="${build.javadoc}/org/apache/nutch/plugin/doc-files" />
620  </target>
621 
622  <target name="default-doc" description="--> generate default Nutch documentation">
623   <style basedir="${conf.dir}" destdir="${docs.dir}" 
624       includes="nutch-default.xml" style="conf/nutch-conf.xsl" />
625  </target>
626 
627  <!-- ================================================================== -->
628   <!-- D I S T R I B U T I O N                                            -->
629   <!-- ================================================================== -->
630   <!--                                                                    -->
631   <!-- ================================================================== -->
632   <target name="package-src" depends="runtime, javadoc" description="--> generate source distribution package">
633     <mkdir dir="${dist.dir}"/>
634     <mkdir dir="${src.dist.version.dir}"/>
635     <mkdir dir="${src.dist.version.dir}/lib"/>
636     <mkdir dir="${src.dist.version.dir}/docs"/>
637     <mkdir dir="${src.dist.version.dir}/docs/api"/>
638     <mkdir dir="${src.dist.version.dir}/ivy"/>
639 
640     <copy todir="${src.dist.version.dir}/lib" includeEmptyDirs="false">
641       <fileset dir="lib"/>
642     </copy>
643 
644     <copy todir="${src.dist.version.dir}/conf">
645       <fileset dir="${conf.dir}" excludes="**/*.template"/>
646     </copy>
647 
648     <copy todir="${src.dist.version.dir}/docs/api">
649       <fileset dir="${build.javadoc}"/>
650     </copy>
651 
652     <copy todir="${src.dist.version.dir}">
653       <fileset dir=".">
654         <include name="*.txt" />
655         <!--<include name="KEYS" />-->
656       </fileset>
657     </copy>
658 
659     <copy todir="${src.dist.version.dir}/src" includeEmptyDirs="true">
660       <fileset dir="src"/>
661     </copy>
662 
663     <copy todir="${src.dist.version.dir}/ivy" includeEmptyDirs="true">
664       <fileset dir="ivy"/>
665     </copy>
666 
667     <copy todir="${src.dist.version.dir}/" file="build.xml"/>
668     <copy todir="${src.dist.version.dir}/" file="default.properties"/>
669 
670   </target>
671 
672  <target name="package-bin" depends="runtime, javadoc" description="--> generate binary distribution package">
673     <mkdir dir="${dist.dir}"/>
674     <mkdir dir="${bin.dist.version.dir}"/>
675     <mkdir dir="${bin.dist.version.dir}/lib"/>
676     <mkdir dir="${bin.dist.version.dir}/bin"/>
677     <mkdir dir="${bin.dist.version.dir}/conf"/>
678     <mkdir dir="${bin.dist.version.dir}/docs"/>
679     <mkdir dir="${bin.dist.version.dir}/docs/api"/>
680     <mkdir dir="${bin.dist.version.dir}/plugins"/>
681 
682     <copy todir="${bin.dist.version.dir}/lib" includeEmptyDirs="false">
683       <fileset dir="runtime/local/lib"/>
684     </copy>
685    
686     <copy todir="${bin.dist.version.dir}/bin">
687       <fileset dir="runtime/local/bin"/>
688     </copy>
689 
690     <chmod perm="ugo+x" type="file">
691         <fileset dir="${bin.dist.version.dir}/bin"/>
692     </chmod>
693 
694     <copy todir="${bin.dist.version.dir}/conf">
695       <fileset dir="runtime/local/conf" excludes="**/*.template"/>
696     </copy>
697 
698     <copy todir="${bin.dist.version.dir}/docs/api">
699       <fileset dir="${build.javadoc}"/>
700     </copy>
701 
702     <copy todir="${bin.dist.version.dir}">
703       <fileset dir=".">
704         <include name="*.txt" />
705       </fileset>
706     </copy>
707 
708     <copy todir="${bin.dist.version.dir}/plugins" includeEmptyDirs="true">
709       <fileset dir="runtime/local/plugins"/>
710     </copy>
711 
712   </target>
713 
714   <!-- ================================================================== -->
715   <!-- Make src-release tarball                                               -->
716   <!-- ================================================================== -->
717   <target name="tar-src" depends="package-src" description="--> generate src.tar.gz distribution package">
718     <tar compression="gzip" longfile="gnu"
719       destfile="${src.dist.version.dir}.tar.gz">
720       <tarfileset dir="${src.dist.version.dir}" mode="664" prefix="${final.name}">
721         <exclude name="src/bin/*" />
722         <include name="**" />
723       </tarfileset>
724       <tarfileset dir="${src.dist.version.dir}" mode="755" prefix="${final.name}">
725         <include name="src/bin/*" />
726       </tarfileset>
727     </tar>
728   </target>
729   
730   <!-- ================================================================== -->
731   <!-- Make bin release tarball                                               -->
732   <!-- ================================================================== -->
733   <target name="tar-bin" depends="package-bin" description="--> generate bin.tar.gz distribution package">
734     <tar compression="gzip" longfile="gnu"
735       destfile="${bin.dist.version.dir}.tar.gz">
736       <tarfileset dir="${bin.dist.version.dir}" mode="664" prefix="${final.name}">
737         <exclude name="bin/*" />
738         <include name="**" />
739       </tarfileset>
740       <tarfileset dir="${bin.dist.version.dir}" mode="755" prefix="${final.name}">
741         <include name="bin/*" />
742       </tarfileset>
743     </tar>
744   </target>
745 
746   <!-- ================================================================== -->
747   <!-- Make src release zip                                               -->
748   <!-- ================================================================== -->
749   <target name="zip-src" depends="package-src" description="--> generate src.zip distribution package">
750    <zip compress="true" casesensitive="yes" 
751      destfile="${src.dist.version.dir}.zip">
752    <zipfileset dir="${src.dist.version.dir}" filemode="664" prefix="${final.name}">
753        <exclude name="src/bin/*" />
754        <include name="**" />
755    </zipfileset>
756    <zipfileset dir="${src.dist.version.dir}" filemode="755" prefix="${final.name}">
757        <include name="src/bin/*" />
758    </zipfileset>
759    </zip>
760   </target>
761 
762   <!-- ================================================================== -->
763   <!-- Make bin release zip                                               -->
764   <!-- ================================================================== -->
765   <target name="zip-bin" depends="package-bin" description="--> generate bin.zip distribution package">
766    <zip compress="true" casesensitive="yes" 
767      destfile="${bin.dist.version.dir}.zip">
768    <zipfileset dir="${bin.dist.version.dir}" filemode="664" prefix="${final.name}">
769        <exclude name="bin/*" />
770        <include name="**" />
771    </zipfileset>
772    <zipfileset dir="${bin.dist.version.dir}" filemode="755" prefix="${final.name}">
773        <include name="bin/*" />
774    </zipfileset>
775    </zip>
776   </target>
777 
778  <!-- ================================================================== -->
779  <!-- Clean. Delete the build files, and their directories -->
780  <!-- ================================================================== -->
781 
782  <!-- target: clean =================================================== -->
783  <target name="clean" depends="clean-build, clean-lib, clean-dist, clean-runtime" description="--> clean the project" />
784 
785  <!-- target: clean-local ============================================= -->
786  <target name="clean-local" depends="" description="--> cleans the local repository for the current module">
787   <delete dir="${ivy.local.default.root}/${ivy.organisation}/${ivy.module}" />
788  </target>
789 
790  <!-- target: clean-lib =============================================== -->
791  <target name="clean-lib" description="--> clean the project libraries directory (dependencies)">
792   <delete includeemptydirs="true" dir="${build.lib.dir}" />
793  </target>
794 
795  <!-- target: clean-build ============================================= -->
796  <target name="clean-build" description="--> clean the project built files">
797   <delete includeemptydirs="true" dir="${build.dir}" />
798  </target>
799 
800  <!-- target: clean-dist ============================================= -->
801  <target name="clean-dist" description="--> clean the project dist files">
802   <delete includeemptydirs="true" dir="${dist.dir}" />
803  </target>
804 
805  <!-- target: clean-cache ============================================= -->
806  <target name="clean-cache" depends="" description="--> delete ivy cache">
807   <ivy:cleancache />
808  </target>
809 
810  <target name="clean-runtime" description="--> clean the project runtime area">
811   <delete includeemptydirs="true" dir="${runtime.dir}" />
812  </target>
813 
814  <!-- ================================================================== -->
815  <!-- RAT targets -->
816  <!-- ================================================================== -->
817  <target name="rat-sources-typedef" description="--> run RAT antlib task">
818   <typedef resource="org/apache/rat/anttasks/antlib.xml">
819    <classpath>
820     <fileset dir="." includes="rat*.jar" />
821    </classpath>
822   </typedef>
823  </target>
824 
825  <target name="rat-sources" depends="rat-sources-typedef"
826   description="--> runs the tasks over src/java">
827   <rat:report xmlns:rat="antlib:org.apache.rat.anttasks">
828    <fileset dir="src">
829     <include name="java/**/*" />
830     <include name="plugin/**/src/**/*" />
831    </fileset>
832   </rat:report>
833  </target>
834 
835  <!-- ================================================================== -->
836  <!-- SONAR targets -->
837  <!-- ================================================================== -->
838 
839  <!-- Define the Sonar task if this hasn‘t been done in a common script -->
840  <taskdef uri="antlib:org.sonar.ant" resource="org/sonar/ant/antlib.xml">
841   <classpath path="${ant.library.dir}" />
842   <classpath path="${mysql.library.dir}" />
843   <classpath><fileset dir="." includes="sonar*.jar" /></classpath>
844  </taskdef>
845 
846  <!-- Add the target -->
847  <target name="sonar" description="--> run SONAR analysis">
848 
849   <!-- list of mandatory source directories (required) -->
850   <property name="sonar.sources" value="${src.dir}" />
851 
852   <!-- list of properties (optional) -->
853   <property name="sonar.projectName" value="Nutchgora Branch Sonar Analysis" />
854   <property name="sonar.binaries" value="${build.dir}/classes" />
855   <property name="sonar.binaries" value="${build.dir}/plugins" />
856   <property name="sonar.tests" value="${test.src.dir}" />
857 
858   <sonar:sonar workDir="${base.dir}" key="org.apache.nutch:branch"
859    version="2.0-SNAPSHOT" xmlns:sonar="antlib:org.sonar.ant" />
860  </target>
861  
862   <!-- ================================================================== -->
863   <!-- Eclipse targets                                                    -->
864   <!-- ================================================================== -->
865 
866   <!-- classpath for generating eclipse project -->
867   <path id="eclipse.classpath">
868     <fileset dir="${build.lib.dir}">
869       <include name="*.jar" />
870       <exclude name="ant-eclipse-1.0-jvm1.2.jar" />
871     </fileset>
872   </path>
873   
874   <!-- target: ant-eclipse-download   =================================== -->
875   <target name="ant-eclipse-download" description="--> Downloads the ant-eclipse binary.">
876     <get src="http://downloads.sourceforge.net/project/ant-eclipse/ant-eclipse/1.0/ant-eclipse-1.0.bin.tar.bz2"
877          dest="${build.dir}/ant-eclipse-1.0.bin.tar.bz2" usetimestamp="false" />
878 
879     <untar src="${build.dir}/ant-eclipse-1.0.bin.tar.bz2"
880            dest="${build.dir}" compression="bzip2">
881       <patternset>
882         <include name="lib/ant-eclipse-1.0-jvm1.2.jar"/>
883       </patternset>
884     </untar>
885 
886     <delete file="${build.dir}/ant-eclipse-1.0.bin.tar.bz2" />
887   </target>
888   
889   <!-- target: eclipse   ================================================ -->
890   <target name="eclipse" 
891           depends="clean,init,job,ant-eclipse-download"
892           description="--> Create eclipse project files">
893 
894          <pathconvert property="eclipse.project">
895            <path path="${basedir}"/>
896            <regexpmapper from="^.*/([^/]+)$$" to="\1" handledirsep="yes"/>
897          </pathconvert>
898   
899     <taskdef name="eclipse"
900              classname="prantl.ant.eclipse.EclipseTask"
901              classpath="${build.dir}/lib/ant-eclipse-1.0-jvm1.2.jar" />
902     <eclipse updatealways="true">
903       <project name="${eclipse.project}" />
904       <classpath>
905         
906         <library path="${conf.dir}" exported="false" />
907         <library path="${basedir}/src/bin" exported="false" />
908         <library pathref="eclipse.classpath" exported="false" />
909         <library path="${basedir}/build/plugins/urlfilter-automaton/automaton-1.11-8.jar" 
910                  exported="false" />
911         <library path="${basedir}/src/plugin/parse-swf/lib/javaswf.jar" 
912                  exported="false" />
913         <library path="${basedir}/build/plugins/lib-nekohtml/nekohtml-0.9.5.jar" 
914                  exported="false" />
915         <library path="${basedir}/build/plugins/lib-nekohtml/nekohtml-0.9.5.jar" 
916                  exported="false" />
917         <library path="${basedir}/build/plugins/parse-html/tagsoup-1.2.jar" 
918                  exported="false" />
919         <library path="${basedir}/build/plugins/protocol-sftp/jsch-0.1.41.jar" 
920                  exported="false" />
921                  
922         <library path="${basedir}/build/plugins/parse-html/tagsoup-1.2.jar" 
923                  exported="false" />
924 
925         <source path="${basedir}/src/java/" />
926         <source path="${basedir}/src/test/" />
927         <source path="${basedir}/src/plugin/creativecommons/src/java/" />
928         <source path="${basedir}/src/plugin/creativecommons/src/test/" />
929         <!-- feed is currently disabled 
930         <source path="${basedir}/src/plugin/feed/src/java/" />
931         <source path="${basedir}/src/plugin/feed/src/test/" /> -->
932         <source path="${basedir}/src/plugin/index-anchor/src/java/" />
933         <source path="${basedir}/src/plugin/index-anchor/src/test/" />
934         <source path="${basedir}/src/plugin/index-basic/src/java/" />
935         <source path="${basedir}/src/plugin/index-basic/src/test/" />
936         <source path="${basedir}/src/plugin/index-more/src/java/" />
937         <source path="${basedir}/src/plugin/index-more/src/test/" />
938         <source path="${basedir}/src/plugin/language-identifier/src/java/" />
939         <source path="${basedir}/src/plugin/language-identifier/src/test/" />
940         <source path="${basedir}/src/plugin/lib-http/src/java/" />
941         <source path="${basedir}/src/plugin/lib-http/src/test/" />
942         <source path="${basedir}/src/plugin/lib-regex-filter/src/java/" />
943         <source path="${basedir}/src/plugin/lib-regex-filter/src/test/" />
944         <source path="${basedir}/src/plugin/microformats-reltag/src/java/" />
945         <source path="${basedir}/src/plugin/microformats-reltag/src/test/" />
946         <!-- parse-ext is currently disabled 
947         <source path="${basedir}/src/plugin/parse-ext/src/java/" />
948         <source path="${basedir}/src/plugin/parse-ext/src/test/" /> -->
949         <source path="${basedir}/src/plugin/parse-html/src/java/" />
950         <source path="${basedir}/src/plugin/parse-html/src/test/" />
951         <source path="${basedir}/src/plugin/parse-js/src/java/" />
952         <source path="${basedir}/src/plugin/parse-js/src/test/" />
953         <!-- parse-swf and parse-zip are currently disabled 
954         <source path="${basedir}/src/plugin/parse-swf/src/java/" />
955         <source path="${basedir}/src/plugin/parse-swf/src/test/" />
956         <source path="${basedir}/src/plugin/parse-zip/src/java/" />
957         <source path="${basedir}/src/plugin/parse-zip/src/test/" /> -->
958         <source path="${basedir}/src/plugin/parse-tika/src/java/" />
959         <source path="${basedir}/src/plugin/parse-tika/src/test/" />
960         <source path="${basedir}/src/plugin/protocol-file/src/java/" />
961         <source path="${basedir}/src/plugin/protocol-file/src/test/" />
962         <source path="${basedir}/src/plugin/protocol-ftp/src/java/" />
963         <source path="${basedir}/src/plugin/protocol-httpclient/src/java/" />
964         <source path="${basedir}/src/plugin/protocol-httpclient/src/test/" />
965         <source path="${basedir}/src/plugin/protocol-http/src/java/" />
966         <source path="${basedir}/src/plugin/protocol-sftp/src/java/" />
967         <source path="${basedir}/src/plugin/scoring-link/src/java/" />
968         <source path="${basedir}/src/plugin/scoring-opic/src/java/" />
969         <source path="${basedir}/src/plugin/subcollection/src/java/" />
970         <source path="${basedir}/src/plugin/subcollection/src/test/" />
971         <source path="${basedir}/src/plugin/tld/src/java/" />
972         <source path="${basedir}/src/plugin/urlfilter-automaton/src/java/" />
973         <source path="${basedir}/src/plugin/urlfilter-automaton/src/test/" />
974         <source path="${basedir}/src/plugin/urlfilter-domain/src/java/" />
975         <source path="${basedir}/src/plugin/urlfilter-domain/src/test/" />
976         <source path="${basedir}/src/plugin/urlfilter-prefix/src/java/" />
977         <source path="${basedir}/src/plugin/urlfilter-regex/src/java/" />
978         <source path="${basedir}/src/plugin/urlfilter-regex/src/test/" />
979         <source path="${basedir}/src/plugin/urlfilter-suffix/src/java/" />
980         <source path="${basedir}/src/plugin/urlfilter-suffix/src/test/" />
981         <source path="${basedir}/src/plugin/urlfilter-validator/src/java/" />
982         <source path="${basedir}/src/plugin/urlnormalizer-basic/src/java/" />
983         <source path="${basedir}/src/plugin/urlnormalizer-basic/src/test/" />
984         <source path="${basedir}/src/plugin/urlnormalizer-pass/src/java/" />
985         <source path="${basedir}/src/plugin/urlnormalizer-pass/src/test/" />
986         <source path="${basedir}/src/plugin/urlnormalizer-regex/src/java/" />
987         <source path="${basedir}/src/plugin/urlnormalizer-regex/src/test/" />
988 
989         <output path="${basedir}/build/classes" />
990       </classpath>
991     </eclipse>
992   </target>
993 </project>

  1 <?xml version="1.0" ?>
  2 
  3 <!-- Licensed to the Apache Software Foundation (ASF) under one or more contributor 
  4   license agreements. See the NOTICE file distributed with this work for additional 
  5   information regarding copyright ownership. The ASF licenses this file to 
  6   You under the Apache License, Version 2.0 (the "License"); you may not use 
  7   this file except in compliance with the License. You may obtain a copy of 
  8   the License at http://www.apache.org/licenses/LICENSE-2.0 Unless required 
  9   by applicable law or agreed to in writing, software distributed under the 
 10   License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS 
 11   OF ANY KIND, either express or implied. See the License for the specific 
 12   language governing permissions and limitations under the License. -->
 13 
 14 <ivy-module version="1.0">
 15   <info organisation="org.apache.nutch" module="nutch">
 16     <license name="Apache 2.0"
 17       url="http://www.apache.org/licenses/LICENSE-2.0.txt/" />
 18     <ivyauthor name="Apache Nutch Team" url="http://nutch.apache.org" />
 19     <description homepage="http://nutch.apache.org">Nutch is an open source web-search
 20       software. It builds on Hadoop, Tika and Solr, adding web-specifics, such as a crawler, 
 21       a link-graph database etc.
 22     </description>
 23   </info>
 24 
 25   <configurations>
 26     <include file="${basedir}/ivy/ivy-configurations.xml" />
 27   </configurations>
 28 
 29   <publications>
 30     <!--get the artifact from our module name -->
 31     <artifact conf="master" />
 32   </publications>
 33 
 34   <dependencies>
 35     <dependency org="org.elasticsearch" name="elasticsearch" rev="0.19.4" 
 36                 conf="*->default"/>
 37   
 38     <dependency org="org.apache.solr" name="solr-solrj" rev="3.4.0"
 39       conf="*->default" />
 40     <dependency org="org.slf4j" name="slf4j-log4j12" rev="1.6.1"
 41       conf="*->master" />
 42 
 43     <dependency org="commons-lang" name="commons-lang" rev="2.4"
 44       conf="*->default" />
 45     <dependency org="commons-collections" name="commons-collections"
 46       rev="3.1" conf="*->default" />
 47     <dependency org="commons-httpclient" name="commons-httpclient"
 48       rev="3.1" conf="*->master" />
 49     <dependency org="commons-codec" name="commons-codec" rev="1.3"
 50       conf="*->default" />
 51 
 52     <dependency org="org.apache.hadoop" name="hadoop-core"
 53       rev="1.2.1" conf="*->default">
 54       <exclude org="net.sf.kosmosfs" name="kfs" />
 55       <exclude org="net.java.dev.jets3t" name="jets3t" />
 56       <exclude org="org.eclipse.jdt" name="core" />
 57       <exclude org="org.mortbay.jetty" name="jsp-*" />
 58     </dependency>
 59 
 60     <dependency org="com.ibm.icu" name="icu4j" rev="4.0.1" />
 61     <dependency org="org.apache.tika" name="tika-core" rev="1.3" />
 62     <dependency org="com.googlecode.juniversalchardet" name="juniversalchardet" rev="1.0.3"/>
 63 
 64     <dependency org="log4j" name="log4j" rev="1.2.15" conf="*->master" />
 65 
 66     <dependency org="xerces" name="xercesImpl" rev="2.9.1" />
 67     <dependency org="xerces" name="xmlParserAPIs" rev="2.6.2" />
 68     <dependency org="xalan" name="serializer" rev="2.7.1" />
 69     <dependency org="oro" name="oro" rev="2.0.8" />
 70 
 71     <dependency org="org.jdom" name="jdom" rev="1.1" conf="*->default" />
 72 
 73     <dependency org="com.google.guava" name="guava" rev="11.0.2" />
 74     <dependency org="com.google.code.crawler-commons" name="crawler-commons" rev="0.2" />
 75 
 76     <!--Configuration: test -->
 77 
 78     <!--artifacts needed for testing -->
 79     <dependency org="junit" name="junit" rev="4.11" conf="*->default" />
 80 
 81     <dependency org="org.apache.hadoop" name="hadoop-test" rev="1.2.1" conf="test->default">
 82       <exclude org="net.sf.kosmosfs" name="kfs" />
 83       <exclude org="net.java.dev.jets3t" name="jets3t" />
 84       <exclude org="org.eclipse.jdt" name="core" />
 85       <exclude org="org.mortbay.jetty" name="jsp-*" />
 86     </dependency>
 87 
 88     <dependency org="org.mortbay.jetty" name="jetty" rev="6.1.26" conf="test->default" />
 89     <dependency org="org.mortbay.jetty" name="jetty-util" rev="6.1.26" conf="test->default" />
 90     <dependency org="org.mortbay.jetty" name="jetty-client" rev="6.1.26" />
 91 
 92     <dependency org="org.hsqldb" name="hsqldb" rev="2.2.8" conf="*->default" />
 93     <dependency org="org.jdom" name="jdom" rev="1.1" conf="test->default"/>
 94 
 95     <dependency org="org.restlet.jse" name="org.restlet" rev="2.0.5" conf="*->default" />
 96     <dependency org="org.restlet.jse" name="org.restlet.ext.jackson" rev="2.0.5" 
 97       conf="*->default" />
 98 
 99     <!--================-->
100     <!-- Gora artifacts -->
101     <!--================-->
102     <dependency org="org.apache.gora" name="gora-core" rev="0.3" conf="*->default"/>
103     <!-- Uncomment this to use SQL as Gora backend. It should be noted that the 
104     gora-sql 0.1.1-incubating artifact is NOT compatable with gora-core 0.3. Users should 
105     downgrade to gora-core 0.2.1 in order to use SQL as a backend. -->
106     <!--
107     <dependency org="org.apache.gora" name="gora-sql" rev="0.1.1-incubating" conf="*->default" />
108     -->
109     <!-- Uncomment this to use MySQL as database with SQL as Gora store. -->
110     <!-- 
111     <dependency org="mysql" name="mysql-connector-java" rev="5.1.18" conf="*->default"/> 
112     -->
113     <!-- Uncomment this to use HBase as Gora backend. -->
114     
115     <dependency org="org.apache.gora" name="gora-hbase" rev="0.3" conf="*->default" />
116     
117     <!-- Uncomment this to use Accumulo as Gora backend. -->
118     <!-- 
119     <dependency org="org.apache.gora" name="gora-accumulo" rev="0.3" conf="*->default" />
120     -->
121     <!-- Uncomment this to use Cassandra as Gora backend. -->
122     <!-- 
123     <dependency org="org.apache.gora" name="gora-cassandra" rev="0.3" conf="*->default" />
124     -->
125 
126     <!--global exclusion -->
127     <exclude module="ant" />
128     <exclude module="slf4j-jdk14" />
129     <exclude module="slf4j-simple" />
130     <exclude org="hsqldb"/>
131     <exclude org="maven-plugins"/>
132     <exclude module="jmxtools" />
133     <exclude module="jms" />
134     <exclude module="jmxri" />
135     <exclude module="thrift" />
136   </dependencies>
137 
138 </ivy-module>

  1 <ivysettings>
  2 
  3  <!--
  4    Licensed to the Apache Software Foundation (ASF) under one or more
  5    contributor license agreements.  See the NOTICE file distributed with
  6    this work for additional information regarding copyright ownership.
  7    The ASF licenses this file to You under the Apache License, Version 2.0
  8    (the "License"); you may not use this file except in compliance with
  9    the License.  You may obtain a copy of the License at
 10 
 11        http://www.apache.org/licenses/LICENSE-2.0
 12 
 13    Unless required by applicable law or agreed to in writing, software
 14    distributed under the License is distributed on an "AS IS" BASIS,
 15    WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 16    See the License for the specific language governing permissions and
 17    limitations under the License.
 18 -->
 19 
 20  <!--
 21   see http://www.jayasoft.org/ivy/doc/configuration
 22   -->
 23   <!-- you can override this property to use mirrors
 24           http://repo1.maven.org/maven2/
 25           http://mirrors.dotsrc.org/maven2
 26           http://ftp.ggi-project.org/pub/packages/maven2
 27           http://mirrors.sunsite.dk/maven2
 28           http://public.planetmirror.com/pub/maven2
 29           http://ibiblio.lsu.edu/main/pub/packages/maven2
 30           http://www.ibiblio.net/pub/packages/maven2
 31   -->
 32   <property name="oss.sonatype.org" 
 33     value="http://oss.sonatype.org/content/repositories/releases/" 
 34     override="false"/>
 35   <property name="repo.maven.org"
 36     value="http://repo1.maven.org/maven2/"
 37     override="false"/>
 38   <property name="snapshot.apache.org"
 39     value="http://people.apache.org/repo/m2-snapshot-repository/"
 40     override="false"/>
 41   <property name="maven2.pattern"
 42     value="[organisation]/[module]/[revision]/[module]-[revision]"/>
 43   <property name="maven2.pattern.ext"
 44     value="${maven2.pattern}.[ext]"/>
 45   <!-- pull in the local repository -->
 46   <include url="${ivy.default.conf.dir}/ivyconf-local.xml"/>
 47   <settings defaultResolver="default"/>
 48   <resolvers>
 49     <ibiblio name="maven2"
 50       root="${repo.maven.org}"
 51       pattern="${maven2.pattern.ext}"
 52       m2compatible="true"
 53       />
 54     <ibiblio name="apache-snapshot"
 55       root="${snapshot.apache.org}"
 56       pattern="${maven2.pattern.ext}"
 57       m2compatible="true"
 58       />
 59     <ibiblio name="restlet"
 60       root="http://maven.restlet.org"
 61       pattern="${maven2.pattern.ext}"
 62       m2compatible="true"
 63       />
 64      <ibiblio name="sonatype"
 65       root="${oss.sonatype.org}"
 66       pattern="${maven2.pattern.ext}"
 67       m2compatible="true"
 68       />
 69      
 70     <chain name="default" dual="true">
 71       <resolver ref="local"/>
 72       <resolver ref="maven2"/>
 73       <resolver ref="sonatype"/>
 74     </chain>
 75     <chain name="internal">
 76       <resolver ref="local"/>
 77     </chain>
 78     <chain name="external">
 79       <resolver ref="maven2"/>
 80       <resolver ref="sonatype"/>
 81     </chain>
 82     <chain name="external-and-snapshots">
 83       <resolver ref="maven2"/>
 84       <resolver ref="apache-snapshot"/>
 85       <resolver ref="sonatype"/>
 86     </chain>
 87     <chain name="restletchain">
 88       <resolver ref="restlet"/>
 89     </chain>
 90   </resolvers>
 91   <modules>
 92 
 93     <!--
 94     This forces a requirement for other nutch-artifacts to be built locally
 95     rather than look for them online.
 96     -->
 97     <module organisation="org.apache.nutch" name=".*" resolver="internal"/>
 98     <module organisation="org.restlet" name=".*" resolver="restletchain"/>
 99     <module organisation="org.restlet.jse" name=".*" resolver="restletchain"/>
100   </modules>
101 </ivysettings>