Changeset 67

Timestamp:

May 21, 2009, 2:49:49 PM (15 years ago)

Author:

waue

Message:

v2

Location:

nutchez-0.1

Files:

• bin/nutchez (modified) (1 diff)
• bin/nutchez-func.sh (modified) (4 diffs)
• bin/test.sh (deleted)
• build.xml (modified) (1 diff)
• conf/.nutchez (added)
• conf/.nutchez/n.crawler.txt (added)
• conf/.nutchez/n.robot.txt (added)
• conf/.nutchez/n.tomcat.txt (added)
• conf/.nutchez/n.url.txt (added)
• conf/.nutchez/search (added)
• conf/.nutchez/urls (added)
• debian/nutchez.install (modified) (1 diff)
• debian/nutchez.postrm (modified) (1 diff)
• tomcat/webapps/ROOT/WEB-INF/classes/nutch-site.xml (modified) (1 diff)

nutchez-0.1/bin/nutchez

@@ -8 +8 @@
 
 . ./nutchez-func.sh
+
 # root ?
 #check_if_root
 
 # show url lists
+
+setup_nutchez
+
 CHECK=0
 while [ $CHECK -eq 0 ]; do

nutchez-0.1/bin/nutchez-func.sh

@@ -5 +5 @@
 # .
 
-. $NUTCHEZ_SCRIPT_PATH/conf/hadoop-env.sh
-
-VERB=1
+. /etc/nutch/hadoop-env.sh || . $NUTCHEZ_SCRIPT_PATH/conf/hadoop-env.sh
 
 : ${DIALOG=dialog}
 
-#NEZ_DIR=/home/waue/.nutchez
-#NEZ_DIR_URL
-#NEZ_DIR_OTHER
+VERB=0
 
 
@@ -53 +49 @@
 }
 
+setup_nutchez() {
+  if ! [ -e ~/.nutchez ] ;then
+    cp -rf /etc/nutch/.nutchez ~/
+    chown -R $LOGNAME:$LOGNAME ~/.nutchez
+    if [ -e /tmp/nutch ] ;then
+      rm -rf /tmp/nutch
+    fi
+    mkdir /tmp/nutch
+    ln -sf ~/.nutchez/search /tmp/nutch/
 
+  fi
+}
 
 show_urls (){
@@ -98 +105 @@
   echo " \n 2. The robot name is : \n" >> $tempfile
   cat /tmp/n.robot.tmp >> $tempfile
-  echo " \n 3. The crawled dipth is : \n " >> $tempfile
+  echo " \n 3. The crawled depth is : \n " >> $tempfile
   cat /tmp/n.crawler.tmp >> $tempfile
   echo " \n 4. The explorer port is : \n " >> $tempfile
@@ -113 +120 @@
 }
 
+# define paramaters
+ROBOT=`cat /tmp/n.url.txt`
+URLS=`cat /tmp/n.url.txt`
+DEPTH=`cat /tmp/n.url.txt`
+PORT=`cat /tmp/n.url.txt`
+
 start_crawl () {
   echo_vb "7. start_crawl"
+  setup_nutch
+  /opt/nutch/bin/nutch crawl ~/.nutch/urls -dir ~/.nutch/search -threads 2 -depth $DEPTH 
 }
 
 start_tomcat () {
   echo_vb "8. start_tomcat "
+  /opt/nutch/tomcat/bin/startup.sh
 }
 

nutchez-0.1/build.xml

@@ -14 +14 @@
  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
  See the License for the specific language governing permissions and
- limitations under the License.
+ limitations under this License.
 -->
 <project name="Nutch" default="job">

nutchez-0.1/debian/nutchez.install

@@ -1 +1 @@
 conf/*    etc/nutch
 bin   opt/nutch
+bin/nutchez*  usr/local/sbin
 lib   opt/nutch
 webapps   opt/nutch

nutchez-0.1/debian/nutchez.postrm

@@ -17 +17 @@
     rm -rf /tmp/hadoop*
     rm -rf /tmp/hsperfdata*
+    rm /usr/local/sbin/nutchez*
   fi
 }

nutchez-0.1/tomcat/webapps/ROOT/WEB-INF/classes/nutch-site.xml

@@ -6 +6 @@
     <property>
         <name>searcher.dir</name>
-        <value>/opt/nutch/search</value>
+        <value>/tmp/nutch/search</value>
     </property>
 </configuration>