Neo4j密码模式理解和收集加入

Question

I have a following Cypher Pattern Comprehension:

[ (parentD)<-[:DEFINED_BY]-(ch1:Characteristic)<-[v1:HAS_VALUE_ON]-(childD)  
WHERE NOT ((ch1)<-[:DEPENDS_ON]-()) AND ch1.type = 'CHILD'
| {characteristicId: toInt(ch1.id),  value: v1.value, available: v1.available, totalHistoryValues: v1.totalHistoryValues, description: v1.description, valueType: ch1.valueType, visualMode: ch1.visualMode} ] AS valuedCharacteristics

where parentD and childD are the variables from outer query scope.

Right now I need to improve this Pattern Comprehension with additional nodes:

(childD)<-[:DEFINED_BY]-(ch2:Characteristic)<-[v2:HAS_VALUE_ON]-(childD)
WHERE NOT ((ch2)<-[:DEPENDS_ON]-()) AND ch2.type = 'OWN'

in order to be able also return Characteristics ( ch2 ) from this query and their values( v2 ).

Please help to extend the mentioned query in order to return this information under the same valuedCharacteristics

UPDATED

I have updated comprehension to the following:

[ (parentD)<-[:DEFINED_BY]-(ch1:Characteristic)<-[v1:HAS_VALUE_ON]-(childD)<-[:DEFINED_BY]-(ch2:Characteristic)<-[v2:HAS_VALUE_ON]-(childD)  
  WHERE 
    NOT ((ch1)<-[:DEPENDS_ON]-()) AND ch1.type = 'CHILD' AND 
    NOT ((ch2)<-[:DEPENDS_ON]-()) AND ch2.type = 'OWN' | 
  { characteristicId: toInt(ch1.id), value: v1.value, available: v1.available, totalHistoryValues: v1.totalHistoryValues, description: v1.description, valueType: ch1.valueType, visualMode: ch1.visualMode, 
    childCharacteristicId: toInt(ch2.id), childValue: v2.value, childAvailable: v2.available, childTotalHistoryValues: v2.totalHistoryValues, childDescription: v2.description, childValueType: ch2.valueType, childVisualMode: ch2.visualMode 
  } 
] AS valuedCharacteristics

but right now it always returns empty collection.

I may be wrong but I think the reason is that the following query:

[ (parentD)<-[:DEFINED_BY]-(ch1:Characteristic)<-[v1:HAS_VALUE_ON]-(childD)<-[:DEFINED_BY]-(ch2:Characteristic)<-[v2:HAS_VALUE_ON]-(childD)  

can't return the result in some cases.. for example when ch2 is present and ch1 is absent.

Answer 1

This pattern comprehension should work (but it will generate an empty collection unless the entire pattern is matched):

[(parentD)<-[:DEFINED_BY]-(ch1:Characteristic)<-[v1:HAS_VALUE_ON]-(childD)<-[:DEFINED_BY]-(ch2:Characteristic)<-[v2:HAS_VALUE_ON]-(childD)
  WHERE
    NOT ((ch1)<-[:DEPENDS_ON]-()) AND ch1.type = 'CHILD' AND
    NOT ((ch2)<-[:DEPENDS_ON]-()) AND ch2.type = 'OWN' |
  {
    characteristicId: toInt(ch1.id), value: v1.value, available: v1.available, totalHistoryValues: v1.totalHistoryValues, description: v1.description, valueType: ch1.valueType, visualMode: ch1.visualMode,
    childCharacteristicId: toInt(ch2.id), childValue: v2.value, childAvailable: v2.available, childTotalHistoryValues: v2.totalHistoryValues, childDescription: v2.description, childValueType: ch2.valueType, childVisualMode: ch2.visualMode
  }
] AS valuedCharacteristics

[UPDATED]

However, if you want to always get the "parent" data, but optionally get the "child" data (if it exists), then you could switch to using MATCH and OPTIONAL MATCH . For example:

MATCH (parentD)<-[:DEFINED_BY]-(ch1:Characteristic)<-[v1:HAS_VALUE_ON]-(childD)
WHERE NOT ((ch1)<-[:DEPENDS_ON]-()) AND ch1.type = 'CHILD'
OPTIONAL MATCH (childD)<-[:DEFINED_BY]-(ch2:Characteristic)<-[v2:HAS_VALUE_ON]-(childD)
WHERE NOT ((ch2)<-[:DEPENDS_ON]-()) AND ch2.type = 'OWN'
RETURN COLLECT({
  parent: {characteristicId: toInt(ch1.id), value: v1.value, available: v1.available, totalHistoryValues: v1.totalHistoryValues, description: v1.description, valueType: ch1.valueType, visualMode: ch1.visualMode},
  child: {characteristicId: toInt(ch2.id), value: v2.value, available: v2.available, totalHistoryValues: v2.totalHistoryValues, description: v2.description, valueType: ch2.valueType, visualMode: ch2.visualMode}
}) AS valuedCharacteristics