feat: graph traversals

[tpluscode]

I propose this to allow moving pointer(s) between (named) graphs in a dataset

import cf from 'clownface'
import rdf from 'rdf-ext'

const pointer = cf({ dataset: rdf.dataset(), term: rdf.blankNode() })

const defaultGraphPointer = unionGraphPointer.fromGraph(pointer.defaultGraph)

const namedGraphPointer = defaultGraphPointer.fromGraph(rdf.namedNode())

const againUnionGraph = namedGraphPointer.fromGraph(null)

Now that I look at this, I'm a little concerned about literals and blank nodes. While the former would be harmless albeit strange, blank nodes should maybe be excluded from this feature?

[codecov-commenter]

Codecov Report

Merging #45 into master will increase coverage by 0.38%.
The diff coverage is 100.00%.

@@            Coverage Diff             @@
##           master      #45      +/-   ##
==========================================
+ Coverage   98.18%   98.57%   +0.38%     
==========================================
  Files           9        9              
  Lines         276      281       +5     
==========================================
+ Hits          271      277       +6     
+ Misses          5        4       -1     

Impacted Files	Coverage Δ
lib/Clownface.js	99.15% <100.00%> (+0.03%)	⬆️
lib/Context.js	100.00% <0.00%> (+1.42%)	⬆️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update b1a4500...3b33b88. Read the comment docs.

[martinwebrant]

I struggle to use the implicit/explicit graphs in an easy way. Normally I do not care which graph something belongs to but there are cases like when I want to get a label of the type of a node - then I only want to get the exact type from the explicit graph.

For my use case it would be great to add the graph when matching things like:
myNamedNode.out(ns.rdf.type, ns.onto.explicit).out(ns.rdfs.label).value;
where ns.onto.explicit is the explicit graph.
Could also use the options if one does not want to add more arguments.
myNamedNode.out(ns.rdf.type, {graph: ns.onto.explicit}).out(ns.rdfs.label).value;

What do you think?

edit:
I found an ok solution for my problem. The sourcecode revealed that one can create two context's and if I create two datasets (one of explicit and one for implicit) and order the explicit first then I will get the type result from that first which is exactly what I want.
Looks like you are working on making the _context a bit more public which would suit me fine.
Not sure who helps out with the typescript definitions, but I can fix them up when you are done.

edit2:
Creating context does not help, after traversing you loose the context that did not match and subsequent queries will not get the complete result.

[tpluscode]

Hello @martinwebrant. Sorry for late reply.

In code where I needed a specific graph, in the current version I would create a new pointer object like:

let pointer: GraphPointer
let graph: Term

const pointerAtGraph = cf({ dataset: pointer.dataset, term: pointer.term, graph })

I maintain the typings by the way

[martinwebrant]

const pointerAtGraph = cf({ dataset: pointer.dataset, term: pointer.term, graph })

Thanks, I'll do it like that for now!