erik at ehatchersolutions
Jul 22, 2005, 4:55 AM
Post #7 of 7
(4015 views)
Permalink
|
On Jul 22, 2005, at 4:37 AM, Namrata Kumari wrote:
> - Well, the application I want to develop is more like storing xml
> files and
> with each of them having different structure. And then performing
> search on
> them that in turn can depend on the structure of the xml doc and
> user's
> requirement.
That's still a pretty generic requirement. What type of queries?
XPath?
> - Moreover, I did not exactly understood as to how I can store the xml
> document. I mean, I went through the java doc and couldnot figure
> out the
> api's that could be used for this purpose. Can you guide me in this?
Look at the various types of fields. There is a "stored" attribute
on Field that allows the field to be stored.
> - But the biggest question is: Is Lucene a good option [.which now I
> doubt on
> the basis of what I have read till now :-(]
It really all depends. I built a search engine for the Rossetti
Archive (http://www.rossettiarchive.org/rose/) which indexes XML
files like this:
http://www.rossettiarchive.org/docs/1-1847.s244.raw.xml
XPath queries are not possible into the XML, but that is also not a
use case for the system. Highly structured queries such as this one
are supported because the indexing process extracted detailed
information from the XML files:
http://www.rossettiarchive.org/rose/?query=%2Bgenre%3Asonnet+%2B%
28author%3Arossetti+OR+author%3Adgr%29+%2Byear%3A%5B1850+TO+1870%5D
I still do not have a clear cut understanding of your needs and thus
still not sure if Lucene is suitable or not. Certainly for full-text
searches it is a fine choice, but the structured queries are a
different story.
Erik
>
> Regards,
> Namrata
>
>
> -----Original Message-----
> From: Erik Hatcher [mailto:erik [at] ehatchersolutions]
> Sent: Friday, July 22, 2005 2:11 PM
> To: general [at] lucene
> Subject: Re: Lucene as xml store
>
>
> On Jul 22, 2005, at 1:07 AM, Namrata Kumari wrote:
>
>
>>
>> hi,
>>
>> I am a beginner to lucene , So kindly excuse me if the questions
>> mentioned a bit naive.
>> - Can I use lucene as an xml store + search engine?
>> - What I understood is that if we want to perform search on xml doc.
>> we need to parse xml document, form indexes and on the basis of
>> fields
>> perform search.
>> - So, does this mean, that even if we use lucene as xml store (IF WE
>> CAN!!), we need to parse it to form indexes?
>>
>
> Lucene is a search engine and only deals with text (Strings
> essentially).
> Lucene is also a flat document space and doing queries for things
> hierarchical is not how it was designed, but it can be done to a
> limited
> degree depending on how data is indexed.
>
> Yes, Lucene can store text as well as make it searchable - so you
> could
> store an XML document in it as well.
>
> You have not provided any information on the types of queries you
> need to
> support or what the user experience will be like. There are many
> ways to
> use Lucene and whether it is suitable solution to your
> application depends on that information. Tell us more about what
> you're wanting to do and we can guide you further.
>
>
>> Please reply to this as soon as possible
>>
>
> That's what they all say! :) No need to say such a thing - if you
> have well articulated questions that are straightforward enough to
> answer,
> you'll get responses quickly here.
>
> Erik
>
|
