This is a public chat log generated from the #semsol IRC channel.
09:10:36
I discovered a problem while working with a LOAD query, but I'm not sure if it's actually a problem or if it's wanted
09:11:21
what is it elbart ?
09:11:30
I want to load an URI that's constructed how it's recommended by the W3C, they call it Hash URI with content negotiation
09:11:37
i get the following error:
09:11:53
array(3) { [0]=> string(93) "Socket error: No supported URI scheme detected. in ARC2_Reader via ARC2_StoreLoadQueryHandler" [1]=> string(76) "missing stream in "getFormat" via ARC2_Reader via ARC2_StoreLoadQueryHandler" [2]=> string(99) "No loader available for "xxxxxxx": in ARC2_StoreLoadQueryHandler" }
09:12:06
I'm not sure if the problem is the 303 redirection or s.th. else
09:12:25
what is the actual uri ?
09:12:27
do I have to resolve the redirection myself or is ARC2 able to do that?
09:12:49
arc will do that
09:12:52
http://work.webholics.de/origo/profile#mv
09:13:42
of course in order to get the RDF xml file the loader has to set the appropriate header
09:13:49
not sure if this is the case
09:14:22
it should do
09:14:42
hm then where's the problem?
09:15:53
you can find the actual rdf xml data under http://work.webholics.de/origo/profile.rdf
09:16:03
not sure yet
09:16:08
but this rdf should be ok, it's directly generated with ARC2
09:16:24
elbart: and does a load of that uri directly work ?
09:16:58
hm can't test this so easy
09:17:17
my system is already very complex
09:17:57
don't you have an open sparql endpoint to try out?
09:18:49
ah wait I can try it
09:18:51
elbart: ARC is developed by bengee who unfortunately isn't here right now
09:21:51
directly loading the rdf uri works
09:22:38
it has to do s.th. with the 303 redirection i think
09:28:13
I currently try to understand the code of ARC2_Reader, that does the http request stuff, maybe I can find s.th.
09:28:56
ok: the function getHTTPSocket() (line 136) gets called twice
09:29:06
at first everything is ok
09:29:31
but the second time $url is only the path not the url, so if (!isset($parts['scheme'])) { fails
09:29:41
not sure why this method is called twice though
09:31:15
ok here is s.th wrong!! (line 210) return $this->getHTTPSocket($new_url, $redirs+1);
09:31:23
$new_url is not correct
09:32:41
i think i know where the actual problem is:
09:34:09
If the Location header within the 303 redirection isn't absolute, ARC2 has to build an absolute URI
09:34:14
but it doesn't
09:34:31
Location doesn't have to be absolute
09:41:48
I've checked the http standards, and I have to say ARC2 has a correct behaviour here, a Location url has to be absolute
09:42:02
so there's no bug in arc
09:43:14
and the birds are singing and all is right with the world ;)
09:43:37
;)
09:44:03
it's really hard to find such bugs if you're using a server side framework that manages all this url stuff
11:34:06
Hi there. Any SPARQL and RSS2.0 guru in here ?
11:34:28
what do you need to know?
11:34:52
rss 2.0 isn't RDF though ... ?
11:35:06
I'm trying to query a rss2.0 feed in SPARQL and .. I don"t find the right URI for the rss2.0 namespace
11:35:30
oh .. right ... it's not rdf ?
11:36:04
i think ARC parses it into rdf though
11:36:26
in fact I have issue with pubDate
11:36:32
looking at the code, it just migrates it to RSS 1.0
11:36:35
ARC sees it as dc:date
11:36:40
yes
11:37:02
if you look at the extractRDF function in ARC2_RSSParser
11:37:09
(if you haven't already)
11:37:15
you can see what the mappings are
11:37:32
arnaudfontaine go look at the arc28rssparser
11:37:43
thx a lot kwijibo
11:37:51
no probs