You might want to re-examine your assumptions about Lucene.
One can view the index constructed by Lucene as being itself a database; and therefore it
would be possible (though probably not desireable) to construct an ODBC adapter and access
the content of a Lucene index as if it were a typical relational database using SQL via Microsoft
ADO.
Another thought to consider is that Lucene gets a large part of its search speed by the specific
data structures it uses. Revising Lucene to access a traditional database directly, to use
the database tables as its "index", would most likely reduce search performance below the
level where searching the data via Lucene provides any practical benefit.
I think it best to view Lucene as an attachment that you use in addition to (or in parallel
with) your database.
Hope this helps.
-- Neal
-----Original Message-----
From: Maha Khairy [mailto:black_opal_87@yahoo.com]
Sent: Wednesday, January 30, 2008 4:55 PM
To: lucene dotNet
Subject: question
Hello
I wanted to build a search engine using the net-lucene, the problem is I find it a little
ambiguous to understand , I know it doesn't do crawling and I understood it only do the indexing
and the searching, but it doesn't interact with any database, I assume it use data structures
instead of tables in database,
Anyway, now I need to know if I understood right, and how to make it interact with database
(if my assumption was true, what are the classes that I need to modify to do that)
And if the lucene have any flowcharts, activity diagrams or any diagrams that show the flow
of data how do I get them.
Another question
If I wanted to support a certain language what are the classes that I'll need to implement
Thanks in advance
|