I think loading quads into Fuseki/TDB actually does make sense in an integration context-- for example, base ontologies or data sets for an application's startup.
But I will leave that for the future. I will investigate the general in-memory set and work out something similar to 313 there, if it needs it. ajs6f > On Nov 20, 2017, at 12:10 PM, Andy Seaborne <[email protected]> wrote: > > > > On 20/11/17 14:35, ajs6f wrote: >> I wasn't looking at the comment-- I wrote that code myself. It's documented >> at: >> https://jena.apache.org/documentation/rdf/datasets.html >> (which I'm pretty sure _you_ wrote!) >> The fact that it loads quads was originally unintentional, as the comments >> clearly indicate, but it is tested and documented. That's not good! >> 313 is fine for TIM, but let's make sure that the feature is uniform. >> (That's why I assigned myself the ticket.) Does TDB[1|2] load quads in the >> same way? Does the general in-memory dataset? > > TDB(1|2) are persistent so loading data at Fuseki start-up is not relevant. > > What they do have is ja:unionDefaultGraph. > > Andy > >> ajs6f >>> On Nov 20, 2017, at 6:18 AM, Andy Seaborne <[email protected]> wrote: >>> >>> >>> >>> On 19/11/17 21:06, ajs6f wrote: >>>> And TIM's assembler does _not_ support quads, >>> >>> You shouldn't trust the comment! >>> (is this documented anywhere? The class is in assembler/assembler.ttl, not >>> ja:data, and thetre is no HTML aside from deep in javadoc I could find.) >>> >>> If you read the code and work out the magic: it loads quads for the first >>> use of ja:data but it has to be a URI unlike ja:content. >>> >>> PR#313 is a step along the way. >>> >>> Andy >>> >>>> and there is no reason it shouldn't. We could work via: >>>> https://github.com/apache/jena/blob/master/jena-core/src/main/java/org/apache/jena/assembler/assemblers/AssemblerBase.java >>>> JENA-1430 is filed. >>>> ajs6f >>>>> On Nov 19, 2017, at 4:02 PM, ajs6f <[email protected]> wrote: >>>>> >>>>> Is this not: >>>>> >>>>> https://github.com/apache/jena/blob/master/jena-arq/src/main/java/org/apache/jena/sparql/core/assembler/InMemDatasetAssembler.java >>>>> >>>>> TIM's assembler? >>>>> >>>>> >>>>> ajs6f >>>>> >>>>>> On Nov 19, 2017, at 3:58 PM, Andy Seaborne <[email protected]> wrote: >>>>>> >>>>>> "fuseki2 --file ~/tmp/D.nq /ds" starts Fuseki with a quads file. It's >>>>>> done in the command line code, not by assembler. >>>>>> >>>>>> ja:RDFDataset is the general "link to any sort of graph" dataset. Quads >>>>>> would be convenient sometimes and the lack of loading a quads files is a >>>>>> bit of an omission. >>>>>> >>>>>> More importantly, for the TIM (Transaction in Memory) dataset >>>>>> implementation, loading a trig file makes a lot sense and now I come to >>>>>> look, TIM does not have an assembler yet we ought to favor that >>>>>> especially for Fuseki. I sense an AbstractDatasetAssembler is in our >>>>>> future! >>>>>> >>>>>> Could someone please create a ticket? >>>>>> >>>>>> Andy >>>>>> >>>>>> On 19/11/17 20:43, Mohammad Noorani Bakerally wrote: >>>>>>> no, there is no specific reason, was just asking if it is possible, >>>>>>> >>>>>>> On Sun, Nov 19, 2017 at 8:25 PM, ajs6f <[email protected]> wrote: >>>>>>>> Did you have a reason to get Fuseki to serve directly from a file? If >>>>>>>> not, >>>>>>>> you might do better (depending on your use) to load your file into >>>>>>>> either >>>>>>>> an in-memory or TDB (persistent) store and let Fuseki work from that. >>>>>>>> >>>>>>>> >>>>>>>> ajs6f >>>>>>>> >>>>>>>>> On Nov 19, 2017, at 1:10 PM, Mohammad Noorani Bakerally < >>>>>>>> [email protected]> wrote: >>>>>>>>> >>>>>>>>> I have a trig file containing a default graph and several named >>>>>>>>> graphs. >>>>>>>> Can >>>>>>>>> Fuseki directly consume the file, because in the documentation [1], I >>>>>>>>> see >>>>>>>>> that the default and named graph should be described. >>>>>>>>> >>>>>>>>> I have tried to load it in the two followings way but it doesn't work >>>>>>>> and I >>>>>>>>> get an exception >>>>>>>>> >>>>>>>>> Way1 >>>>>>>>> ====== >>>>>>>>> <#dataset1> rdf:type ja:RDFDataset ; >>>>>>>>> rdfs:label "Dataset 1: RDF LinkGeoData for Strasbourg" ; >>>>>>>>> ja:content [ja:externalContent <file:///path/to/trig.ttl> ] ; >>>>>>>>> >>>>>>>>> >>>>>>>>> How to write the configuration so that the dataset of a service is a >>>>>>>>> trig >>>>>>>>> file containing one default graph and several named graphs ? >>>>>>>>> >>>>>>>>> 1. https://jena.apache.org/documentation/serving_data/ >>>>>>>>> >>>>>>>>> >>>>>>>> >>>>>>>> >>>>>
