> START named=node:names("first:*")
   WITH named
> START next=node:names("first: " + current.last), prev=node:names("last: " + current.first)
> RELATE current -[:LEADS_TO]-> next
> RELATE prev -[:LEADS_TO]-> current

> So I have a list of name nodes in an index called "names":
> "foo bar"
> "bar baz"
> "baz qux"
> "bar qux"
> "qux foo"

> Each name node is indexed with a key "first" which is the first word in the name, and "last" which is the last word in the name, and each node has properties on it containing the first and last words.  The following Cypher works:

> START named=node:names("first:bar") RETURN named.name

> returns "bar baz" and "bar qux". So all that is working

> Now what I want to do is use Cypher to link together the name nodes such that the last word of a node is the first word of the outgoing node. In the end, I should have a graph that looks like:

> "foo bar" --> "bar baz"
> "foo bar" --> "bar qux"
> "bar baz" --> "baz qux"
> etc...

> I think I should be able to do this in Cypher with FOREACH and RELATE, but I'm unsure of the exact syntax.  The following is semantically what I want, but is not syntactically correct:

> START named=node:names("first:*")
> FOREACH (current in nodes(named):
>    START next=node:names("first: " + current.last), prev=node:names("last: " + current.first)
>    RELATE current -[:LEADS_TO]-> next
>    RELATE prev -[:LEADS_TO]-> current
> )

> I could do this all in the application, but I was hoping for a more elegant solution using Cypher (preferably one that is idempotent.)

> Any ideas?

> Thanks!
> -- Josh

On Thursday, August 2, 2012 3:54:05 AM UTC-4, Michael Hunger wrote:

> I think in this case you don't need foreach in this case:

> > START named=node:names("first:*")
>    WITH named
> > START next=node:names("first: " + current.last), prev=node:names("last:
> " + current.first)
> > RELATE current -[:LEADS_TO]-> next
> > RELATE prev -[:LEADS_TO]-> current

> Hope that works.

> Can you try to create a console example that we can work on? (The console
> is auto-indexing each property so if you create a node with the first and
> last properties you should find them in the node_auto_index)

> Michael

> Am 02.08.2012 um 05:20 schrieb Josh Adell:

> > So I have a list of name nodes in an index called "names":
> > "foo bar"
> > "bar baz"
> > "baz qux"
> > "bar qux"
> > "qux foo"

> > Each name node is indexed with a key "first" which is the first word in
> the name, and "last" which is the last word in the name, and each node has
> properties on it containing the first and last words.  The following Cypher
> works:

> > START named=node:names("first:bar") RETURN named.name

> > returns "bar baz" and "bar qux". So all that is working

> > Now what I want to do is use Cypher to link together the name nodes such
> that the last word of a node is the first word of the outgoing node. In the
> end, I should have a graph that looks like:

> > "foo bar" --> "bar baz"
> > "foo bar" --> "bar qux"
> > "bar baz" --> "baz qux"
> > etc...

> > I think I should be able to do this in Cypher with FOREACH and RELATE,
> but I'm unsure of the exact syntax.  The following is semantically what I
> want, but is not syntactically correct:

> > START named=node:names("first:*")
> > FOREACH (current in nodes(named):
> >    START next=node:names("first: " + current.last),
> prev=node:names("last: " + current.first)
> >    RELATE current -[:LEADS_TO]-> next
> >    RELATE prev -[:LEADS_TO]-> current
> > )

> > I could do this all in the application, but I was hoping for a more
> elegant solution using Cypher (preferably one that is idempotent.)

> > Any ideas?

> > Thanks!
> > -- Josh

On Fri, Aug 3, 2012 at 1:02 AM, Josh Adell <josh.ad...@gmail.com> wrote:
> This is what I have so far:

> http://console.neo4j.org/?id=yj9cwv

> I think I'm getting a parse error because it doesn't like the string
> concatenation in 'next=node:node_auto_index("first:"+named.last)'

> -- Josh

> On Thursday, August 2, 2012 3:54:05 AM UTC-4, Michael Hunger wrote:

>> I think in this case you don't need foreach in this case:

>> > START named=node:names("first:*")
>>    WITH named
>> > START next=node:names("first: " + current.last), prev=node:names("last:
>> > " + current.first)
>> > RELATE current -[:LEADS_TO]-> next
>> > RELATE prev -[:LEADS_TO]-> current

>> Hope that works.

>> Can you try to create a console example that we can work on? (The console
>> is auto-indexing each property so if you create a node with the first and
>> last properties you should find them in the node_auto_index)

>> Michael

>> Am 02.08.2012 um 05:20 schrieb Josh Adell:

>> > So I have a list of name nodes in an index called "names":
>> > "foo bar"
>> > "bar baz"
>> > "baz qux"
>> > "bar qux"
>> > "qux foo"

>> > Each name node is indexed with a key "first" which is the first word in
>> > the name, and "last" which is the last word in the name, and each node has
>> > properties on it containing the first and last words.  The following Cypher
>> > works:

>> > START named=node:names("first:bar") RETURN named.name

>> > returns "bar baz" and "bar qux". So all that is working

>> > Now what I want to do is use Cypher to link together the name nodes such
>> > that the last word of a node is the first word of the outgoing node. In the
>> > end, I should have a graph that looks like:

>> > "foo bar" --> "bar baz"
>> > "foo bar" --> "bar qux"
>> > "bar baz" --> "baz qux"
>> > etc...

>> > I think I should be able to do this in Cypher with FOREACH and RELATE,
>> > but I'm unsure of the exact syntax.  The following is semantically what I
>> > want, but is not syntactically correct:

>> > START named=node:names("first:*")
>> > FOREACH (current in nodes(named):
>> >    START next=node:names("first: " + current.last),
>> > prev=node:names("last: " + current.first)
>> >    RELATE current -[:LEADS_TO]-> next
>> >    RELATE prev -[:LEADS_TO]-> current
>> > )

>> > I could do this all in the application, but I was hoping for a more
>> > elegant solution using Cypher (preferably one that is idempotent.)

>> > Any ideas?

>> > Thanks!
>> > -- Josh