[opencms-dev] Help: Lucene Search CronIndexManager
Arthur Visser
arthur.visser at student.groept.be
Mon May 2 14:24:42 CEST 2005
After a (very!) long search I finally found the problem why
the Indexer would not run and produce a
NullPointerException.
I'll post the solution so that others can use the same if
they are confronted with this problem.
The solution is to edit tomcat/bin/catalina.bat and change
the line which sets the
temp dir. If you don't set it to an absolute path, Tomcat
resolves filenames as relative
file names which then results in a directory without files
and hence the NullPointerException.
catalina.bat:
#set CATALINA_TMPDIR=%CATALINA_BASE%\temp (OLD)
set
CATALINA_TMPDIR=C:\Java\Libraries\jakarta-tomcat-5.0.24\temp
(SET TO YOUR OWN DIRECTORY)
I hope this may help for other implementation of the Lucene
Search module which works fine now!
I am using : OpenCMS 5.0.1 - Tomcat 5.0.18 - Java 1.4.2 - XP
Home SP2 - MySQL 4.0.22
Arthur Visser
-----Oorspronkelijk bericht-----
Van: opencms-dev-bounces at opencms.org
[mailto:opencms-dev-bounces at opencms.org] Namens Arthur
Visser
Verzonden: 24 April 2005 23:52
Aan: 'The OpenCms mailing list'
Onderwerp: [opencms-dev] Help: Lucene Search
CronIndexManager
Hello,
I hope anybody can help me with the following problem. I
have installed the Lucene Search Module in OpenCms, but do
not succeed getting the Index job done.
I have checked everything several times and even
re-installed the module etc etc. Also checked and changed
registry file many times, tried different directories.
Everything seems to fail. I have just about run out of ideas
what else to check. The only thing I can imagine would be a
problem with a failure to write files in the indexdir on the
XP Home system. Does anyone know how to set the file write
permission for XP Home for a specific folder? Just
right-click on a folder doesn't give too many options (just
sharing, but no security or 'level 5' options as in XP
professional.
I am using : OpenCMS 5.0.1 - Tomcat 5.0.18 - Java 1.4.2 - XP
Home SP2 - MySQL 4.0.22
Please find the error message and registry below.
Part of the message in my fileviewer (always the same):
13:42:10] <opencms_cronscheduler> Error running job for
com.opencms.core.CmsCronEntry{42 13 * * * admin
Administrators
net.grcomputing.opencms.search.lucene.CronIndexManager
createIndex=true} Error: java.lang.NullPointerException
at
org.apache.lucene.store.FSDirectory.create(FSDirectory.java:
172)
at
org.apache.lucene.store.FSDirectory.<init>(FSDirectory.java:
151)
at
org.apache.lucene.store.FSDirectory.getDirectory(FSDirectory
.java:132)
at
org.apache.lucene.store.FSDirectory.getDirectory(FSDirectory
.java:113)
at
org.apache.lucene.index.IndexWriter.<init>(IndexWriter.java:
151)
at
net.grcomputing.opencms.search.lucene.IndexManager.doIndex(U
nknown Source)
at
net.grcomputing.opencms.search.lucene.CronIndexManager.launc
h(Unknown Source)
at
com.opencms.core.CmsCronScheduleJob.run(CmsCronScheduleJob.j
ava:68)
My registry (many times changed - this is just one version):
<?xml version="1.0" encoding="ISO-8859-1"?>
<registry>
<system>
<luceneSearch>
<!--
- mergeFactor and permCheck are currently
ignored.
-->
<mergeFactor>100000</mergeFactor>
<permCheck>true</permCheck>
<!--
- directory in which lucene will store its
indexes. Note: this is real
- fs, not VFS.
-->
<indexDir>/opt/luceneindex</indexDir>
<!-- <indexDir>F:\luceneindex\</indexDir> -->
<!--
- The analyzer is used for parsing documents.
Choose one for your
- language. If language is English, use the
StandardAnalyzer.
- There are additional analyzers at
http://jakarta.apache.org/lucene
-->
<analyzer>org.apache.lucene.analysis.standard.StandardAnalyz
er</analyzer>
<!--
<analyzer>org.apache.lucene.analysis.de.GermanAnalyzer</anal
yzer> -->
<!--
- If subsearch is true, subfolders will be
searched by default.
- This can be turned on/off per directory.
-->
<subsearch>true</subsearch>
<!--
- Name of the project to index. Online is
recommended.
-->
<project>online</project>
<!--
- docFactories determine how documents are
processed. Generally, one
- docFactory exists for each type of content
(viz. JSP, Page, Plain)
- that you want to index.
-->
<docFactories>
<!--
- This docFactory indexes documents with
type page (e.g. HTML
- files edited with the WYSIWYG
editor).
-->
<docFactory enabled="true" type="page">
<class>net.grcomputing.opencms.search.lucene.PageDocument</c
lass>
</docFactory>
<!--
- This docFactory is a little more complex.
It takes documents of
- type "plain" and determines, by
extension, what class should be
- used to index each particular
file. In this example, we want to
- index plain text files exactly as
they are, but any files that
- contain tags need the tags
stripped out before they are indexed.
-
- Note that the name="" attribute
is simply for pretty output, and
- can contain any allowable PCDATA
text.
-->
<docFactory enabled="true" type="plain">
<fileType name="plaintext">
<extension>.txt</extension>
<class>net.grcomputing.opencms.search.lucene.PlainDocument</
class>
</fileType>
<fileType name="taggedtext">
<extension>.html</extension>
<extension>.htm</extension>
<extension>.xml</extension>
<!-- This will strip tags before processing -->
<class>net.grcomputing.opencms.search.lucene.TaggedPlainDocu
ment</class>
</fileType>
</docFactory>
<!--
- This will strip JSP tags and all
scriptlets. IT WILL NOT RENDER THE
- JSP FIRST, as JSPs are, by
nature, dynamic.
-
- Usually, this is off by default.
-->
<docFactory enabled="false" type="jsp">
<class>net.grcomputing.opencms.search.lucene.JspDocument</cl
ass>
</docFactory>
<!-- For the news module. Enable if you use
news -->
<docFactory enabled="false" type="news">
<class>net.grcomputing.opencms.search.lucene.NewsDocument</c
lass>
</docFactory>
<!-- For the forum module. Enable if you use
forums. -->
<docFactory enabled="false" type="forum">
<class>de.wfnetz.opencms.modules.forum.ContributionDocument<
/class>
</docFactory>
<!-- If you need to index XML Template files
(bad idea) use this: -->
<docFactory enabled="false" type="XML Template"/>
</docFactories>
<!--
- <directories/> determines which directories
are indexed. By default,
- the /system directory is never indexed, so
it is safe to index root.
-
- If you want to specify only certain
directories for indexing, create
- one <directory/> entry per directory. Again,
you may use subsearch to
- override the default subsearch setting
discussed above.
-->
<directories>
<directory location="/">
<section>Root</section>
<subsearch>true</subsearch>
</directory>
</directories>
<!--
- Use this section to define specific
contentDefinitions. Provided below
- are entries for the news and forum modules.
-->
<contentDefinitions>
<!-- <contentDefinition type="news">
<class>com.opencms.modules.homepage.news.NewsContentDefiniti
on</class>
<initClass>net.grcomputing.opencms.search.lucene.NewsInitial
ization</initClass>
<listMethod name="getNewsList">
<param type="java.lang.Integer">1</param>
<param type="java.lang.String">-1</param>
</listMethod>
<page uri="/news.html?__element=entry">
<param method="getIntId" name="newsid"/>
</page>
</contentDefinition>
<contentDefinition type="forum">
<class>de.wfnetz.opencms.modules.forum.ContributionContentDe
finition</class>
<listMethod name="getSortedList">
<param type="java.lang.String"/>
</listMethod>
<page
uri="/forum.html?forumtemplate=viewcontributionentry">
<param method="getId" name="conid"/>
</page>
</contentDefinition> -->
</contentDefinitions>
</luceneSearch>
Arthur Visser
Managing Director
I.C.E. Europe bvba
Jachthoornlaan 42
B-1170 Brussels, Belgium
Mobile : +32-(0)477-341850
Fax. : +32-(0)2-675.8374
<http://www.ice-eu.com> www.ice-eu.com
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://webmail.opencms.org/pipermail/opencms-dev/attachments/20050502/5946b15b/attachment.htm>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: image001.jpg
Type: image/jpeg
Size: 7489 bytes
Desc: not available
URL: <https://webmail.opencms.org/pipermail/opencms-dev/attachments/20050502/5946b15b/attachment.jpg>
More information about the opencms-dev
mailing list