Hey there,
When dealing with filters in PQL, the core limitation is that the language is designed to use variables for values, not for structural elements like table or column names. The filter must know exactly which column it is checking. Thus, your direct substitution, like FILTER "${case_table}"."number" = '${number}', fails because PQL cannot dynamically resolve the variable case_table into a live table name at runtime. To overcome this rigidity when dealing with multiple data sources (like overlay Views across different case tables), the recommended solution is to use the CASE_TABLE() operator. This function allows you to reference the current, correct Case Table based on its relationship to the Activity Table, providing a standardized way to pull the required column (.number) regardless of the underlying data source being used. For truly complex scenarios, especially with overlay Views, the best practice is to define the filter logic for each specific case table and use variable mapping only to pass the filter's value (e.g., ${number}), ensuring precise filtering while maintaining the necessary structural definition in the PQL.
Let me know if this helps!