Neo4j Load CSV only when unique

问题

If i have csv as follows:

"route_id","service_id","trip_id","trip_headsign"
4973_3, 2, 14519396, "Bf. Tostedt"
4973_3, 2, 14519395, "Bf. Tostedt"
4973_3, 1, 14519379, "Sittensen, Schule"
4973_3, 1, 14519391, "Sittensen, Bahnhofstraße"
4973_3, 2, 14519394, "Bf. Tostedt"
4973_3, 1, 14519390, "Bf. Tostedt"
4973_3, 3, 14519381, "Bf. Tostedt"
4973_3, 4, 14519392, "Bf. Tostedt"

and I want to load only when trip_headsign is unique, so in above case I would only create Trip 3 times. How do i do it?

I am so far as follows:

load csv with headers from  
 'file:///hamburg/trips.txt' as csv   
 create (t:Trip {id: csv.trip_id, service_id: csv.service_id, headsign: csv.trip_headsign}); 

But I get this error:

Node(76020) already exists with label `Trip` and property `headsign` = 'Bf. Tostedt'

I have intentionally keps headsign as unique constraint because I only want everyone of it to exist only once.

回答1:

I figured out now. Thanks

load csv with headers from  
'file:///hamburg/trips.txt' as csv   
merge (t:Trip {headsign: csv.trip_headsign})
on match set t.id =  csv.trip_id
on match set t.service_id = csv.service_id

来源：https://stackoverflow.com/questions/48543121/neo4j-load-csv-only-when-unique