You are viewing a plain text version of this content. The canonical link for it is here.
Posted to users@jackrabbit.apache.org by Ista Pouss <is...@gmail.com> on 2011/09/27 14:06:39 UTC
Search from multiple things
Hi,
I have a jackrabbit application with a repository and I use the jackrabbit
search system with it.
But I have also other documents, in file systems, or in database, or other.
I should want search system searches also in these others documents.
How is it possible ?
Thanks.
Re: Search from multiple things
Posted by Ista Pouss <is...@gmail.com>.
2011/9/27 Alexander Klimetschek <ak...@adobe.com>
> On 27.09.11 17:29, "David Buchmann" <da...@liip.ch> wrote:
> >that said, i also wished i could replace the built-in lucene with solr
> >and expose a full solr to my client application.
>
> The problem is that the built-in search must comply to the JCR query
> specifications, hence any attempt in Jackrabbit to replace the search
> implementation with something else or make it "extensible" will always be
> difficult.
>
> I think very often you want to rely on built-in search features (e.g. all
> properties indexed), since you use a jcr-based library that requires it,
> but somewhere else in your app you want your custom search that might
> conflict with that.
>
> Hence I think it makes sense to build such custom searches on the
> application level, using observation listeners. Of course that makes it
> asynchronous.
>
>
>
I think it's possible for me, because, I shall use only my build custom
search, no jackrabbit built in search , so no conflict (I hope).
Is there some "How-to" Solr / jackrabbit ?
Thanks again.
Re: Search from multiple things
Posted by Alexander Klimetschek <ak...@adobe.com>.
On 27.09.11 17:29, "David Buchmann" <da...@liip.ch> wrote:
>that said, i also wished i could replace the built-in lucene with solr
>and expose a full solr to my client application.
The problem is that the built-in search must comply to the JCR query
specifications, hence any attempt in Jackrabbit to replace the search
implementation with something else or make it "extensible" will always be
difficult.
I think very often you want to rely on built-in search features (e.g. all
properties indexed), since you use a jcr-based library that requires it,
but somewhere else in your app you want your custom search that might
conflict with that.
Hence I think it makes sense to build such custom searches on the
application level, using observation listeners. Of course that makes it
asynchronous.
Cheers,
Alex
--
Alexander Klimetschek
Developer // Adobe (Day) // Berlin - Basel
Re: Search from multiple things
Posted by David Buchmann <da...@liip.ch>.
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
>> Is it possible to store a link to it in jcr, and then the jackrabbit
>> search
>> system use it ?
>>
>
> Using Jackrabbit out of the box, no. It would probably be easiest to store
> the document in the repository, and link to it from your other location
> (going through Jackrabbit's webdav interface).
or, if you are afraid of changing that, you could just copy your data
into jackrabbit and add the original paths to that, so that is found in
the search. and still deliver the original file.
that said, i also wished i could replace the built-in lucene with solr
and expose a full solr to my client application.
cheers,david
- --
Liip AG // Agile Web Development // T +41 26 422 25 11
CH-1700 Fribourg // PGP 0xA581808B // www.liip.ch
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.10 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/
iEYEARECAAYFAk6B6+oACgkQqBnXnqWBgItPmACfVGfLo0ztAdatjtm90Ij70s1+
4MoAoL5VYcHuBYPNDGmJtUEmrL06UtIR
=cYRP
-----END PGP SIGNATURE-----
Re: Search from multiple things
Posted by Mark Herman <MH...@NBME.org>.
Ista Pouss wrote:
>
> I have a word doc somewhere. I don't want store it in jcr, because I don't
> want clone it.
>
> Is it possible to store a link to it in jcr, and then the jackrabbit
> search
> system use it ?
>
Using Jackrabbit out of the box, no. It would probably be easiest to store
the document in the repository, and link to it from your other location
(going through Jackrabbit's webdav interface).
--
View this message in context: http://jackrabbit.510166.n4.nabble.com/Search-from-multiple-things-tp3846992p3847383.html
Sent from the Jackrabbit - Users mailing list archive at Nabble.com.
Re: Search from multiple things
Posted by Ista Pouss <is...@gmail.com>.
2011/9/27 Alexander Klimetschek <ak...@adobe.com>
>
> The nice thing about JCR and Jackrabbit is actually that you can store
> everything in it (fine-granular data plus binary files), and then you have
> a unified API and a search across everything.
>
>
Ok, for exemple :
I have a word doc somewhere. I don't want store it in jcr, because I don't
want clone it.
Is it possible to store a link to it in jcr, and then the jackrabbit search
system use it ?
Thanks.
Re: Search from multiple things
Posted by Alexander Klimetschek <ak...@adobe.com>.
On 27.09.11 14:06, "Ista Pouss" <is...@gmail.com> wrote:
>I have a jackrabbit application with a repository and I use the jackrabbit
>search system with it.
>
>But I have also other documents, in file systems, or in database, or
>other.
>
>I should want search system searches also in these others documents.
>
>How is it possible ?
Not with Jackrabbit alone.
Please note that JCR and Jackrabbit are more about a great repository for
all kinds of content, rather than being an abstraction API across multiple
different existing storages (databases, file systems, etc.). While the
latter is definitely possible (see mode shape), it's not really the sweet
spot of JCR.
The nice thing about JCR and Jackrabbit is actually that you can store
everything in it (fine-granular data plus binary files), and then you have
a unified API and a search across everything.
Cheers,
Alex
--
Alexander Klimetschek
Developer // Adobe (Day) // Berlin - Basel