Ronan Lamy

The SQL queries we generate are often unnecessarily complicated, particularly because they use way too many nested SELECTs. For instance, this: ```python from datachain import DataChain, C chain = DataChain.from_values(a=range(10)).mutate(b=C("a")).order_by("b").select("b")...

Pre-UDF table logic creates unnecessary copies

2

The pre-UDF logic in https://github.com/iterative/datachain/blob/ee43fd16b751db751a3b70e7833483aea3591232/src/datachain/query/dataset.py#L589-L598 unconditionally copies the input query into a new table, which is expensive and useless in most cases. For context, this was introduced in https://github.com/iterative/dvcx/pull/1068