It's for deleting a relationship from the chain, to reroute the pointers of the previous records to the next records. Inserted relationships are added at the start of the chain because that´s a pointer we always know about, when looking at the node.
On Sun, Oct 26, 2014 at 4:20 PM, jer <[email protected]> wrote: > Hi all, > > I am going through the way to do traversals in Neo4j, and just come up > with a question in relationship store. They are: 33 bytes:1in-use-byte,4 > bytes: fistNode,4secondNode, 4 bytes: relationshipType, 4 bytes: start > node previous relationship(SP) , 4 bytes: start node next > relationship(SN), 4 bytes: end node previous relationship(EP), 4 bytes: end > node next relationship(EN). My problem is, why are two fields for previous > relationship(SP and EP) useful at all? > > The way node store file is organized such that only the index of first > property and first relationship is stored. When I do paper simulations for > traversal, I could see some pattern happen, eg. Node -> First > Relaitonship.SN(start next) -> Second Relationship.SN(start next) -> Third > Relationship.EN(end next) ... and so forth. The point I'm trying to make > here is that, our traversal target can either be organized as the start or > end node of the relationship, thus when going through the traversal, it > make sense to see either SN or EN pointer in relationship being fetched. > However, it seems to me that existence of previous pointers (SP and EP) > does not seem to be useful in traversal? There doesn't seem to be a > situation where you need to go back in the traversal? Even if we want to > insert a new relationship between the start and the end node, we can just > append another relationship to the end of chain. Would you mind providing > me with one case that SP and EP indexes are useful? > > Yours Sincerely, > Jer > > -- > You received this message because you are subscribed to the Google Groups > "Neo4j" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to [email protected]. > For more options, visit https://groups.google.com/d/optout. > -- Mattias Persson Neo4j Hacker at Neo Technology -- You received this message because you are subscribed to the Google Groups "Neo4j" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. For more options, visit https://groups.google.com/d/optout.
