Ritchie Vink

Yeap, we should be able to do that. I can work on that once https://github.com/jorgecarleitao/arrow2/pull/1320 is merged.

I was working towards this in arrow2. That's why I need https://github.com/jorgecarleitao/arrow2/pull/1320. I shall pick it up.

This has been added

I think this discussion should be upstream in pyo3. https://pyo3.rs/main/doc/pyo3/panic/struct.panicexception A panic should not be handled as something that can be caught. We should treat it as a bug in...

I will close this as I think we have explained the situation.

You can use python glob to read the lazyframes separately and then concatenate them sorted by the columns. This is similar to what we do internally save for the sorting.

I will close this as this is expected behavior. If the schemas of different files differ, they have to be handled accordingly by the user as I think it is...

Could you do this one @universalmind303? You can simply copy what we do in the project, hstack and filter method of the LogicalPlanBuilders.

This is not needed in `scan_csv` because the query optimizer will automatically select the columns you use. No need to clutter the api with that. ```python pl.scan_csv(file).select(["foo" , "bar"]) ```...

This is not how `with_context` is meant to work. You need to filter both `LazyFrame`s.