[Persistent] Neo4j backend

124 views
Skip to first unread message

Cesar Barata

unread,
Nov 3, 2011, 2:02:55 PM11/3/11
to yeso...@googlegroups.com
Hi all,

Are there any plans for a Neo4j (graph database) backend for Persistent? If not, where can I find documentation for implementing a Persistent backend?

If you want more information about Neo4j, please visit the official website: neo4j.org

Thanks.

--
Cesar Barata

Greg Weber

unread,
Nov 3, 2011, 2:37:52 PM11/3/11
to yeso...@googlegroups.com
Hi Cesar,

To find out information about implementing a Persistent backend, you are taking the right approach by mailing this list. persistent's type class define the API and the current backends serve as guidance for implementing new ones, and you can ask us any questions you have.

I haven't used neo4j, but I am not sure if it will be a good fit for Persistent. Certainly Persistent's concept of automatically serializing data could be useful, but I imagine that a graph database's query language has a lot of differences between Persistent's query language.

Either way the first step would be to create low level bindings for neo4j in Haskell.

Greg Weber 

Cesar Barata

unread,
Nov 3, 2011, 9:26:34 PM11/3/11
to yeso...@googlegroups.com
Thanks for the heads up Greg.

neo4j is implemented in Java and unfortunately, at the moment, I have no idea how to make GHC and the JVM talk to each other. However neo4j has a REST API and I was thinking about writing a low level wrapper for this API.

2011/11/3 Greg Weber <gr...@gregweber.info>

Joshua Cason

unread,
Jun 8, 2012, 6:28:35 PM6/8/12
to yeso...@googlegroups.com
I forgot to mention that it also has a REST API.

On Tuesday, June 5, 2012 12:31:21 AM UTC-7, Joshua Cason wrote:
Hi, I was also interested in this idea. Thought I would make another suggestion though. OrientDB, while perhaps not yet as well known as Neo4j, seems like a better choice for Yesod. OrientDB is a graph database that also has the features of a document database - data is imported/exported through JSON. Moreover, in addition to graph oriented queries, it also supports a subset of SQL. It claims to have a super performance too (benchmarks are published on its wiki). Seems like it could possibly be an easier db to integrate given the SQL and JSON support. It's worth a look into anyway :)

Josh Cason
Reply all
Reply to author
Forward
0 new messages