Rely on BlueBrainNexusStore source (to avoid any store side JSON transformation) endpoint when calling forge.retrieve or forge.search
Search (sparql) resource and get them at the rev they are stored in the configured searchendpoint and not at the latest revision. In consequence search results are no longer _sychronized
Added a 'array_as_set:bool' argument to keep (when True) unchanged one item long JSON arrays when calling forge.as_jsonld() (default to False). This is because RDF consider JSON arrays as set by default * Set _last_action to resources obtained from forge.retrieve
Added support for setting GET params when calling forge.retrieve retrieving
Added the property distribution.atLocation.store.type and distribution.atLocation.store.type._rev to file metadata