This is a public chat log generated from the #semsol IRC channel.
14:59:15
bengee, is sparql+ a superset of sparul?
15:00:16
not really, although I think I called it one once
15:00:49
there is a subset of sparql+ that is compatible with sparul
15:23:28
kwijibo looks at sparul spec
15:24:14
oh yeah, it's got a few extra keywords
15:24:26
is MODIFY necessary?
15:24:40
yeah, that's what I wondered
15:24:57
MODIFY implements transactions
15:25:17
so it's pretty powerful
15:25:52
ah
15:26:37
bengee: maybe that's where changesets could come in handy
15:26:51
yeah
15:27:36
I dunno if I said, but what I'm doing at the moment is putting an ARC wrapper around the talis API
15:27:50
oh, that's cool
15:28:01
so, rewriting SPARQL+ INSERT/DELETE to changesets
15:28:09
etc
15:28:15
awesome
15:28:18
I can't do the aggregates obviously
15:29:02
but the code to generate the changesets could be used in the arc store too for transactions
15:29:07
and rollbacks
15:29:13
if you implement basic INSERT/DELETE, you could possibly edit rdf docs in a talis store via tabulator
15:29:31
yeah, I was thinking of that vaguely
15:29:39
since your wiki app
15:31:02
bengee: I see SPARUL lets you do DELETE ... INSERT ... in the same call - does arc do that too?
15:31:11
no
15:31:19
phew
15:31:21
heh
15:31:30
(less work)
15:31:40
I wasn't sure about query injection
15:32:15
I may add an UPDATE some day that allows you to do something like that in a single call
15:32:27
there was a proposal on the arc ml
15:32:41
wish I had more time :(
15:32:58
I'm hacking on something new: sparqlscript
15:33:13
yeah - i saw you mention it in your tweet :)
15:33:26
what does it do?
15:34:27
it combines sparql queries with basic operators like IF, =, +=, FOREACH
15:34:48
eg?
15:34:51
obvious use case is sparqlbot
15:35:29
IF ! ASK {check for graph existence}
15:35:37
LOAD <$g>
15:35:39
ENDIF
15:35:49
cool :)
15:36:44
INSERT INTO <#temp> { triples 1 }
15:36:48
INSERT INTO <#temp> { triples 2 }
15:37:15
INSERT INTO <#final> { some custom sparql code based on <#temp> }
15:37:25
DELETE FROM <#temp>
15:37:38
DELETE FROM <$g>
15:37:41
etc
15:37:59
what was it you were saying about endpoint selection?
15:38:03
e.g. to create proper rdf from stuff like del.icio.us
15:38:38
yeah, I'd like to allow the target endpoint to be specified in the query
15:38:42
like
15:39:00
SELECT VIA <http://dbpedia.org/sparql> ...
15:39:28
or rather rows = SELECT VIA
15:39:32
so the endpoint you send that query to simply proxies to dbpedia?
15:39:40
FOREACH rows as row
15:39:42
...
15:39:47
yes
15:40:36
not too impressive for a simple endpoint, but would be handy for something like sparqlscript
15:41:30
not the same thing at all, but I used your sparql parser to do a thing where it works out which endpoint to send the query to, based on the predicates and class types mentioned
15:41:43
ah, nice
15:43:44
crap, gtg, biab
15:44:07
cu
17:24:52
bengee is back
17:56:58
bengee: how do I just retrieve a document without parsing it using the Reader?
17:57:59
$reader = ARC2::getComponent('Reader', $config);
17:58:14
$reader->activate($url);
17:58:36
$data = '';
17:59:12
while ($d = $reader->readStream()) { $data .= $d; }
17:59:21
$reader->closeStream();
17:59:52
cheers bengee :)
17:59:56
np