Creating nodes from nested JSON using neo4J query - json

Question

I'm new in neo4j and i have this json file:
{
"locations_connections": {
"locations": [
{
"id": "aws.us-east-1",
"longitude": 72.8777,
"latitude": 19.0760
},
{
"id": "aws.us-east-2",
"longitude": 126.9780,
"latitude": 37.5665
},
{
"id": "aws.us-west-1",
"longitude": 103.8517837,
"latitude": 1.287950
}
],
"connections": [
{
"aws.us-west-1": [
{
"id": "aws.us-west-1",
"latency": 3.16,
"cost": 0.02
},
{
"id": "aws.us-east-1",
"latency": 53.47,
"cost": 0.02
},
{
"id": "aws.us-east-2",
"latency": 53.47,
"cost": 0.02
}
]
},
{
"aws.us-east-1": [
{
"id": "aws.us-east-1",
"latency": 3.16,
"cost": 0.02
},
{
"id": "aws.us-east-2",
"latency": 53.47,
"cost": 0.02
}
]
},
{
"aws.us-east-2": [
{
"id": "aws.us-east-2",
"latency": 53.47,
"cost": 0.02
}
]
}
]
}
}
After reading the json using the apoc.load.json(URL) procedure , what query do I write to represent this as a graph?
where the Node will contain the information name like for example aws.us-east-1, value of longitude and value of latitude and the edges will have the latency and the cost
I have this code:
call apoc.load.json("/file.json") yield value
UNWIND value.locations_connections.locations as loc
UNWIND value.locations_connections.connections as con
MERGE (e:Element {id:loc.id}) ON CREATE
SET e.longitude = loc.longitude, e.latitude = loc.latitude
WITH con
FOREACH (region_source IN KEYS(con)|
FOREACH (data in con[region_source]|
MERGE (e1:Element {id: region_source})
MERGE (e1)<-[:CONNECT]-(e2:Element {id:data.id, latency:data.latency, cost:data.cost})
))
and the execution result is incorrect:
Added 9 labels, created 9 nodes, set 27 properties, created 6 relationships, completed after 60 ms.and I have seen this one,But this is not what I expected

Answer 1

You cannot use match inside a FOREACH so when you put MERGE and :CONNECT inside the for loop, it is creating multiple nodes. This is what I did and tell us if it works for you or not.
call apoc.load.json("/file.json") yield value
// read the json file
WITH value, value.locations_connections.locations as locs
// for loop to create the locations (or regions)
FOREACH (loc in locs | MERGE (e:Element {id:loc.id}) ON CREATE
SET e.longitude = loc.longitude, e.latitude = loc.latitude
)
// get the data for the connections
WITH value.locations_connections.connections as cons
UNWIND cons as con
// the keys and value are assigned to variables region and data
WITH KEYS(con)[0] as region_source, con[KEYS(con)[0]] as dat
// unwind is similar to a for loop
UNWIND dat as data
// look for the nodes that we want
MATCH (e1:Element {id: region_source}), (e2:Element {id: data.id})
// create the connection between regions
MERGE (e1)<-[:CONNECT {latency:data.latency, cost:data.cost}]-(e2)
RETURN e1, e2
See result below: