Skip to main content


Eclipse Community Forums
Forum Search:

Search      Help    Register    Login    Home
Home » Modeling » Sphinx » Sphinx support
Sphinx support [message #652314] Thu, 03 February 2011 14:54 Go to next message
saurav sarkar is currently offline saurav sarkarFriend
Messages: 428
Registered: July 2009
Senior Member
Hi,

We are from Query2 and learnt that Sphinx intends to use Query2 for its Querying and indexing facilities.

So would like to know if we could provide any help from our side to enable this functionality in Sphinx.

Thanks and Regards,
Saurav


Re: Sphinx support [message #654997 is a reply to message #652314] Thu, 17 February 2011 22:27 Go to previous message
Stephan Eberle is currently offline Stephan EberleFriend
Messages: 40
Registered: July 2009
Member

Hi,

> We are from Query2 and learnt that Sphinx intends to use Query2 for its
> Querying and indexing facilities.

Yes, that's true. In Sphinx we want to enable efficient handling of
large file based models, i.e., models consisting of 100MB+ XML data
which may be spread over 10,000+ files. One somewhat critical element
for achieving that is to have a generic, i.e, metamodel-independent
indexing service in place.

However, we do not want to reinvent the wheel in Sphinx and come up with
a new implementation of such an indexing service. Instead we'd heavily
like to reuse an existing framework. And we actually thought that Query
2 could be an interesting candidate for that.

> So would like to know if we could provide any help from our side to
> enable this functionality in Sphinx.

Thank you very much for your kind offer, this is highly appreciated!

The first thing that we'd like to do is to use Query 2 for doing lazy
proxy resolution in models with fragment-based proxy URIs (aka agnostic
URIs). In contrast to conventional proxy URIs only the path to the
target object in the target resource is known but the target resource
itself is unknown. So, if you want to get around attempting to resolve
such agnostic URIs against all of the 10,000+ resources in your big
model you can almost not do without indexing.

We already have made some preliminary attempts of realizing such an
indexing service by reusing that from Xtext. Thereby we have learned
that the size of the index data and the additional memory consumption
implied by that are a very critical factors. So I'd be interested in
finding out if Query 2 would allow to a) realize an index that fits our
use case (lazy resolution of agnostic proxy URIs) and b) how it manages
its index data and which extra amount of memory space it requires.

A small update regarding the status of Sphinx: the preliminary IP review
has been successfully completed and all the code has been checked in to
Eclipse SVN. We are currently in the process of merging in the latests
evolutions and bug fixes that have been done in Artop since the
beginning of the preliminary IP review. We are optimistic to complete
this task up to end of Feburary 2011 and then will work on providing
Sphinx integration builds and download.

Stephan
Previous Topic:[Announce] Sphinx Webinar - Jan. 26
Next Topic:skip creating/unloading unavailable resources
Goto Forum:
  


Current Time: Fri Mar 29 12:39:19 GMT 2024

Powered by FUDForum. Page generated in 0.03048 seconds
.:: Contact :: Home ::.

Powered by: FUDforum 3.0.2.
Copyright ©2001-2010 FUDforum Bulletin Board Software

Back to the top