On 6/3/05, Rolando Isidoro <[EMAIL PROTECTED]> wrote:
> Hi, I've been looking the online documentation for a way to configure
> the Lucene Search in a new publication, all I found was relative to
> installations of Lenya using Jetty. I'm using Lenya 1.2.3 backed by
> Tomcat 5.0.28 and thought of following the same steps as the ones
> described in the documentation with the necessary adjustments, but I
> wish to clarify some doubts that arose before trying that solution.
> All the instructions I'm using here are taken from
> http://lenya.apache.org/1_2_x/how-to/search.html:

I wrote that.  The original is at http://solprovider.com/lenya/search 
The only change (so far) is adding information about using on Linux.  
The instructions were written against Lenya 1.2.2; I have been told
there were no changes affecting this yet.  They worked under Tomcat
5.0.28 without changing anything except filepaths.

- configuration2xslt.xsl: Bug fix.
- lucene-live.xconf: Change to ConfigurableIndexer
- lenyadocs.xconf: Same directory as lucene-live.xconf
- ant needs the location of crawl_and_index.xml
Everything is good so far.  log4j.properties is optional, but a good
log is very useful for debugging.

> should it work by making this small changes?

With what you did, the indexer should create the index, and search
will display results, but the results will not be usable.  You need
searchfixer.xsl.

> Do I need to perform the instructions in "Fix the XML results to be
> usable" in Lenya 1.2.3?
> Finally, am I missing anything else or are these steps enough in order
> to make the Lucene search work in the publication?
> Thanks in advance, regards
> Rolando

OPTIONAL: sort.xsl may not need to be copied if you skip the other steps.
OPTIONAL: search.xsl is needed for language filtering.
OPTIONAL: search-and-results.xsp adds filtering.  I have not tested
without changing this file.  I remember several bug fixes, but I do
not remember if they were crucial.  (I hope those fixes are integrated
into Lenya, but I am not set up to create a diff yet.)
REQUIRED: searchfixer.xsl fixes the URLs in the search results.
OPTIONAL? usecase-search.xmap is needed for language filtering.

If you do not use usecase-search.xmap, you will need to add the
searchfixer.xsl transform to your search pipeline.  This does work;
development started in the normal pipeline, then moved to a Usecase
because the standard search URL did not include visitor's language. 
Look at usecase-search.xmap for the modified pipeline.

OPTIONAL: "Blocking Default Search" is only needed if you are
implementing security.  (I need to document the rest of the process
for configuring that.)

The process was developed as one piece, and the specs were:
- Add language support.
- Add security for one section.
These required indexing Lenya's XML rather than crawling.  The rest
followed from that.  I would be very interested in knowing if any
steps are optional, and any other ideas to improve the documentation.

solprovider

---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Reply via email to