stardog-union / pinto Goto Github PK

I was wondering why, instead of updating to RDF4J, Pinto was only updated to Sesame 4.x? Was this a conscious decision based on certain things or just a preference?

http://docs.rdf4j.org/migration/

so identifiers can be generated into a specific namespace

There's an "i" missing in the class name "IdentifableImpl".

The top level readme.md says "If the object extends Identifiable", but this should say "If the object implements Identifiable".

similar to jackon's @class field, add a property pinto:javaClass which is the FQCN of the Java class corresponding to the RDF resource

My input is a TTL file.
The desired output are POJO's.
Similar to how a json file is being parsed to pojo's.

right now we assume strict adherence to the beans standard, which means the beans should be mutable w/ their set methods. but it would be great if you can use the builder pattern and have immutable beans.

iirc, JAXB has some support for this that would be worth reviewing

you can use Graphs#withContext to set the named graph for RDF generated by Pinto after the fact, but it would be convenient of the mapper did this for you. either by mapper configuration, annotations, or both.

Hello,

Is there any will to add OWL annotations? with OWL API?

Thank you,

Hello,

Is there any comparison between these 4 libraries? That anybody already written or is willing to share?

Thank you,
Charbel

running gradle build:

com.complexible.pinto.RDFMapperTests > testWriteMap FAILED
    java.lang.AssertionError at RDFMapperTests.java:740

Turns out this is because Models.isomorphic doesn't consider the same instant at different timezones to be equal:
My local TZ is UTC+10:00.
The assertTrue isomorphism comparison at RDFMapperTests.java#L740 fails because the date as constructed from the Long at RDFMapperTests.java#L732 becomes:
"2015-03-15T05:24:42.470+10:00"^^<http://www.w3.org/2001/XMLSchema#dateTime>
which is different than the reference value from data/map.nt of:
"2015-03-14T15:24:42.470-04:00"^^<http://www.w3.org/2001/XMLSchema#dateTime>

If I change the date in data/map.nt to be "2015-03-15T05:24:42.470+10:00"^^<http://www.w3.org/2001/XMLSchema#dateTime>, the test passes.

the round tripping correctly works for the id, but it should be updating Identifiable as it goes

specifically, supporting JAXB like Jackson makes sense so the same beans could be used for XML, JSON, and RDF serialization.

Return a Value or a ResourceBuilder for simple cases where there's a known IRI, like w/ an Enum, or simple string mapping for a class

if multiple values are found for a property which is only cardinality 1, ie public Foo getFoo(), add something to MappingOptions to make this non-fatal and log a warning.

maps: NavigableMap, SortedMap, WeakHashMap, Hashtable, Concurrent*

enum: EnumSet, EnumMap

Guava immutables?

By default, we'll traverse the entire object graph serializing it into RDF, and serialize all of the RDF into the full object graph. There could be cases where you only want to serialize the top-level of the object graph and not cascade into its child objects. More often, during serialization to RDF rather than deserialization; during deserialization you can control this on your own by not providing any more than the top level of the object graph

public Builder namespace(final Namespace theNamespace) {
  return namespace(theNamespace.getPrefix(), theNamespace.getPrefix());
}

should be

public Builder namespace(final Namespace theNamespace) {
  return namespace(theNamespace.getPrefix(), theNamespace.getName());
}

in addition to the code, dependencies will be updated, primarily sesame

This method seems to erroneously use the namespace prefix in place of the name:

public Builder namespace(final Namespace theNamespace) {
	return namespace(theNamespace.getPrefix(), theNamespace.getPrefix());
}

https://github.com/stardog-union/pinto/blob/develop/main/src/com/complexible/pinto/RDFMapper.java#L1071

auto-extend them from Identifiable and SourcedObject so via some casts we have access to 1) the original rdf:ID of the object and the 2) RDF used to generate the bean. This would be ideal for using Pinto as the base for the next-generation of Empire