You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
MATCH (a:Person)
WHERE a.firstName = "Ramon"
RETURN a
And we are given a graph that has 1000 different possible unstructured properties across Person nodes (which is an information kept in the catalog), we will try to construct a plan with 1000 property readers, what will try to read each possible property one after another, and put a null or a value in one ValueVector. So for a node that has only 1 unstructured property, we will try to read 1000 possible properties. In addition, we will output 999 nulls.
Instead, we should have a special ValueVector (possibly of dataType Map or currently we can call it UnstructuredPropertyMap) that and a special AlllUnstructuredPropertiesReader that reads all unstructured properties of a node and puts it into a single ValueVector.
The text was updated successfully, but these errors were encountered:
Currently, when a user asks the query:
And we are given a graph that has 1000 different possible unstructured properties across Person nodes (which is an information kept in the catalog), we will try to construct a plan with 1000 property readers, what will try to read each possible property one after another, and put a null or a value in one ValueVector. So for a node that has only 1 unstructured property, we will try to read 1000 possible properties. In addition, we will output 999 nulls.
Instead, we should have a special ValueVector (possibly of dataType Map or currently we can call it UnstructuredPropertyMap) that and a special AlllUnstructuredPropertiesReader that reads all unstructured properties of a node and puts it into a single ValueVector.
The text was updated successfully, but these errors were encountered: