The QueryManager takes center stage when it comes to handling cross-entity filtering scenarios in GraphWeaver.
Its primary objective is to resolve queries that span multiple data sources, enabling smooth data retrieval across disparate entities.
It is very rare that you would need to use the QueryManager directly but this page explains its use.
When the QueryManager receives a query filter, it treats it as a tree structure, where each branch of the tree can exist in different data sources.
The QueryManager's task, along with the help of the visit function, is to traverse the branches of the tree until reaching the leaf nodes.
Then, it traverses back up the tree, checking at each level whether the data source matches the level below.
If the data source differs at any level, it means that the query needs to be executed at that point.
The QueryManager will then collapse the filter into a set of IDs, making it understandable to the parent data source.
To better understand this complex process, let's consider an example of a cross-data source query: