Nsryjdtyk

I have a map say

car: {make: toyota, color: blue, model: camry} 

and I have a node Car with the property

make: toyota

Now I want to add the properties in the map car to my node Car.

However,

MATCH (n:Car {make:'toyota'}) 

SET n +=car 

RETURN n;

will overwrite the make property on my node.

Is there a way I can avoid it doing this?

The use case is when make is the key property for the node I wouldn't want it to be accidentally changed and of course I wouldn't want to trow an error either.

If you try to do this in one query, you will have to make one WHERE clause that catches all of the properties that are NULL. From that you would need to use a WITH clause to pass on those records and use another WHERE clause to select just the records missing one property. The problem is getting back to the records from the first WHERE clause. That may be possible but if so I don't know how to do that with Cypher.

Using multiple queries for each property this is easily done by adding an IS NULL to the WHERE clause.

To demonstrate this a set of records with different combinations of properties set and missing are created.

CREATE (:Car {make: "toyota", color: "red", model: "prius", car_id: 01} )

CREATE (:Car {make: "toyota", color: "red", car_id: 02} )

CREATE (:Car {make: "toyota", model: "prius", car_id: 03} )

CREATE (:Car {make: "toyota", car_id: 04} );

To see what was created

MATCH (n) RETURN n.make,n.model,n.color,n.car_id,LABELS(n);



╒════════╤═════════╤═════════╤══════════╤═══════════╕

│"n.make"│"n.model"│"n.color"│"n.car_id"│"LABELS(n)"│

╞════════╪═════════╪═════════╪══════════╪═══════════╡

│"toyota"│"prius"  │"red"    │1         │["Car"]    │

├────────┼─────────┼─────────┼──────────┼───────────┤

│"toyota"│null     │"red"    │2         │["Car"]    │

├────────┼─────────┼─────────┼──────────┼───────────┤

│"toyota"│"prius"  │null     │3         │["Car"]    │

├────────┼─────────┼─────────┼──────────┼───────────┤

│"toyota"│null     │null     │4         │["Car"]    │

└────────┴─────────┴─────────┴──────────┴───────────┘

To update just the color property

MATCH (c:Car {make:"toyota"}) 

WHERE c.color IS NULL 

SET c += {color: "blue"}

RETURN c;

and to see what changed

MATCH (n) RETURN n.make,n.model,n.color,n.car_id,LABELS(n);



╒════════╤═════════╤═════════╤══════════╤═══════════╕

│"n.make"│"n.model"│"n.color"│"n.car_id"│"LABELS(n)"│

╞════════╪═════════╪═════════╪══════════╪═══════════╡

│"toyota"│"prius"  │"red"    │1         │["Car"]    │

├────────┼─────────┼─────────┼──────────┼───────────┤

│"toyota"│null     │"red"    │2         │["Car"]    │

├────────┼─────────┼─────────┼──────────┼───────────┤

│"toyota"│"prius"  │"blue"   │3         │["Car"]    │

├────────┼─────────┼─────────┼──────────┼───────────┤

│"toyota"│null     │"blue"   │4         │["Car"]    │

└────────┴─────────┴─────────┴──────────┴───────────┘

Notice that just the cars with no color property are updated.

Doing the same for model

MATCH (m:Car {make:"toyota"}) 

WHERE m.model IS NULL

SET m += {model: "camry"} 

RETURN m;



MATCH (n) RETURN n.make,n.model,n.color,n.car_id,LABELS(n);



╒════════╤═════════╤═════════╤══════════╤═══════════╕

│"n.make"│"n.model"│"n.color"│"n.car_id"│"LABELS(n)"│

╞════════╪═════════╪═════════╪══════════╪═══════════╡

│"toyota"│"prius"  │"red"    │1         │["Car"]    │

├────────┼─────────┼─────────┼──────────┼───────────┤

│"toyota"│"camry"  │"red"    │2         │["Car"]    │

├────────┼─────────┼─────────┼──────────┼───────────┤

│"toyota"│"prius"  │"blue"   │3         │["Car"]    │

├────────┼─────────┼─────────┼──────────┼───────────┤

│"toyota"│"camry"  │"blue"   │4         │["Car"]    │

└────────┴─────────┴─────────┴──────────┴───────────┘

You can use apoc.map.clean to remove certain keys and values from a map:

SET c += apoc.map.clean(car, ['make'],)