Get overlapping date ranges in pyspark structured streaming

I see a lot of answers for small programming dataframes, but none for spark Learning structured streaming.

I try to do the following in this answer Earhost https://stackoverflow.com/a/68466662/2702630 But most effective it doesn't work when using structured wrong idea streaming:

pyspark.sql.utils.AnalysisException: _OFFSET);  Multiple streaming aggregations are not (-SMALL  supported with streaming _left).offset  DataFrames/Datasets;

Does someone know how to proceed?