On 16 Jan 2004, at 12:30, Oliver Zeigermann wrote:


Stefano Mazzocchi wrote:

On 16 Jan 2004, at 09:59, Oliver Zeigermann wrote:
Christophe wrote:

Here is proposal for an index engine intregration like Lucene. This is a starting point and need to be review.
Let me know if it is not really what you expected.


Christophe,

thanks for letting us participate :)
hey, that's unfair. Christophe is doing a 'proposal' and the best way to put something going is to get the code done and let people patch it.... but without a code to start people won't start. so I think a seed is great, any seed.

?????????


I asked Christophe to send his proposal, he did, I said thanks, we start a discussion. What is unfair?

Uh, sorry, I thought you were sarcastic with that "thanks for letting us participate". I deeply apologize if this was not the case. [email communication tends to be difficult to understand sometimes]


I had a look at it, but am no expert neither in searching nor in Lucene and I would be interested what Stefano or Erik have to say.
I don't have time to look seriously into it, but I say that we add it to the loop (in the scratchpad) and see what happens, we can patch it later.

It contains patches of core classes, how do we do this then?

uh, that changes the picture. What do others think?


Anyway, these are my comments:

* Is it necessary to index the document properties or only the content ? I expect to have faster queries if we index the content properties. Maybe we can provide both methods.


To have this query stuff really useful we need fast and scalable searching of properties. I thought Stefano proposed to unconditionally use an RDBMS for this purpose. I'd say this is one way to go, but it still would be nice to have a Lucene solution. There was this Lucene thread in this list mainly by Stefano and Erik. Have a look at it...
I would not index the metadata with lucene, otherwise, it would generate a mess and semantic collision between metadata and search. for example, say you want to do a full-text search of 'author' and you have your own properties that have author in it... all documents come up in the search, but that's not what you wanted.

You would need to store the properties in different fields.

what do you mean?


--
Stefano.


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



Reply via email to