Github user okram commented on the pull request:
https://github.com/apache/incubator-tinkerpop/pull/175#issuecomment-167622062
From what I can tell from the code in this PR, there is nothing "too crazy"
here save that you assume incoming edges and not outgoing edges, where outgoing
edges are the typical "lowest common denominator" for a format. That is, most
people, if they don't have the full "star graph" will have at least outgoing
edges for each vertex (NOT incoming edges). I don't know why you require
incoming edges, but okay... I assume you have your reasons. If you can make it
work without that assumption, that would be best (I suppose).
Next, I don't know how `StarGraph` plays into this. Is `StarGraph` failing
with self-loops? `StarGraph` is serialized via `Gryo` (see `VertexWritable` as
well). If GryoReader/Writer is happy, then I suspect things will work with all
the Hadoop/Spark/etc. work. However, it would be best if we had some self-loop
tests. What sucks is we don't have a `GraphLoader.SELF_LOOPS` style toy dataset
and with `HadoopGraph`, you can't mutate the graph.............................
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at infrastruct...@apache.org or file a JIRA ticket
with INFRA.
---
