Hi. Yes. You are right. Tree is a graph. But could you advise what is the
most preferable way to run through the whole tree and collect all nodes (I
want to build java object tree from database tree and convert to json)?
1. to recursively go from root node to the last child
protected Group buildTree(Vertex vertex){
final List<Group> children = new LinkedList<Group>();
for (Vertex child : vertex.getVertices(Direction.OUT)){
children.add(buildTree(child));
}
return new Group(vertex, children);
}
2. to use java traverse. I do not know if it is even possible. As I
understand traverse gives me just a (unordered?) list of traversed vertices.
3. to use Gremlin (would it be more efficient than point 1?)
Thanks
Andrey.
On Wednesday, 25 June 2014 10:49:41 UTC+1, Lvc@ wrote:
>
> Hi Andrey,
> OrientDB is a GraphDB and a Tree is a kind of graph, so OrientDB well
> suites this use case. You can use indexes for constraints against field
> values. About the rest it's a matter to create vertices (nodes) and edges
> (arcs). You can start from the Tutorial:
>
>
> https://github.com/orientechnologies/orientdb/wiki/Tutorial:-working-with-graphs
>
> Lvc@
>
>
>
> On 24 June 2014 02:23, Andrey Kovalev <[email protected] <javascript:>>
> wrote:
>
>> Hi. I want to store the following structure in a database.
>> System
>>
>> Group-1
>>
>> Group-1
>>
>> Document-1
>>
>> Group-2
>>
>> Group-2
>>
>> Document-1
>>
>> Document-2
>>
>>
>> Initially I was planning to use db4o and to store this hierarchy directly
>> with java classes. But querying is very poor in db4o so I found OrientDB is
>> quite promising in this case.
>> OrientDB is considered as a graph database. On some pages
>> with documentation I saw that OrientDB had been mentioned as a db for tree
>> structures as well.
>> Can anyone share an example (blog post or github) where a simple tree is
>> stored in OrientDB, easily queried and traversed?
>> I want to see what is the best approach to maintain a tree in OrientDB or
>> may be it is better and easier to use another database (mongo?).
>> Can anyone clarify the following items:
>>
>> 1. Is it possible to create an index which will check uniqueness of
>> nodes' fields' values only on a particular level, so I wouldn't be able
>> to
>> add Group-2 to a System node but I would easily add it to a
>> System/Group-2
>> node?
>> 2. If I want to load Groups hierarchy at once and to load Documents
>> lately (lazy) I should store groups as embeddedlist and documents as
>> linklist. Am I right?
>> 1. But in that case in order to add a group I will need to load
>> entire System class manually (in Java) go to a relevant Group, add a
>> child
>> and store entire System. Am I right?
>> 2. If I want to load full hierarchy I can use traverse queries.
>> But I do not get how I can traverse and build a tree structure. As I
>> understand traverse query will return all nodes as a list so again I
>> will
>> need to build a tree structure in java manually?
>>
>> I have much more questions. But depending on how those are answered I may
>> be clear or not.
>> Hoping someone already has thought about it.
>> Thanks in advance.
>>
>>
>>
>>
>>
>>
>>
>>
>>
>> --
>>
>> ---
>> You received this message because you are subscribed to the Google Groups
>> "OrientDB" group.
>> To unsubscribe from this group and stop receiving emails from it, send an
>> email to [email protected] <javascript:>.
>> For more options, visit https://groups.google.com/d/optout.
>>
>
>
--
---
You received this message because you are subscribed to the Google Groups
"OrientDB" 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.
