George Hartzell scribbled on 12/4/06 3:55 PM:
> Peter Karman writes:
> >
> >
> > Peter Leonard scribbled on 12/4/06 1:57 PM:
> >
> > >
> > > I understand the complexity involved, and what we're looking for is the
> > > quick storage & retrieval of the underlying data - I'll look at the
> > > json/yaml serialization implementations.
> > >
> >
> >
> > fwiw, I'm using Swish-e to store my serialized RDBO objects, with the
> > added benefit of being able to do full-text search and sorting on them
> > as well. I'm using
> >
> http://search.cpan.org/~karman/SWISH-API-Object-0.05/lib/SWISH/API/Object.pm
>
> Did you stumble upon and any pitfalls when you set this up? Is it
> magical enough that an example might be useful? (skimming the link
> above doesn't give me a feel if it's enough just to say that the
> result is an RDBO::... or if something more special needs to happen.
Sorry it's taken me awhile to get back to this; have been busy with new kid at
our house.
I just uploaded to CPAN new versions of both SWISH::API::Object and
SWISH::Prog::Object, which can be used together to de/serialize data structures
in Swish-e indexes.
I was inspired by Rose's use of YAML::Syck and JSON::Syck to serialize objects,
so the SWISH::* modules now offer those serialization formats (as well as the
ability to subclass and define your own de/serializer).
Here's how I use them.
We have a group of <1000 RDBO objects that change only once a month or so. They
are composed of several JOINs against many-to-many table relationships. We need
to select and sort subgroups very often. The sorting in particular proved
challenging and time-consuming. So we index them with the SWISH::* modules to
pre-sort all the simple scalar values, which helps immensely. All those JOINs
and the sorting is done once and cached in the index. And the Swish-e syntax is
much simpler for searching than using SQL (though of course not as powerful).
Example:
my @rdbobjects = get_my_objects();
# these are flattened to remove all the Rose::* stuff and
# leave only the column values. The values can be whatever can
# be serialized with YAML or JSON.
my $indexer = SWISH::Prog::Object->new(
methods => [ @list_of_accessors_in_rdbobject ]
);
$indexer->create( [EMAIL PROTECTED] );
# create the index
Then to search/thaw the objects, I use SWISH::API::Object to fetch each result
as a hashref blessed into the class of my choice. Could be a RDBO, to get a
'thaw-like' effect, though in my case, I don't need to hit the DB anymore, so I
don't bother. But see the SWISH::API::Object doc, especially the 'stash' and
'class' params to new().
Let me know offlist if you have more questions, since we've ventured out of
Rose
territory here. :)
Thanks again to John for the good Rose mojo and for the idea of using libsyck.
cheers,
pek
--
Peter Karman . http://peknet.com/ . [EMAIL PROTECTED]
-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys - and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
_______________________________________________
Rose-db-object mailing list
Rose-db-object@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/rose-db-object
