Ritchie Vink

Yeap, would accept a PR on this.

I think we should raise an error here, informing the use to use a `Duration/timedelta` to offset datetime types.

What if we just make `date + time` arithmetic work? That's most intuitive to me.

Alright. Shall I setup the skeleton for the rust expression so we can move it down?

Ok, I will take a look at this one tomorrow (if the bugs allow me).

Thanks a lot! This saves us a lot of work and we are happy we can continue with xlsx2csv. :raised_hands:

Does this occur in the dataframe API as well?

I always have to look up the spec again when dealing with decimal. Thanks for the fix. Can you leave some comments explaining your rationale?

Thank you for the additional comments! Our decimial support is not near where we want it to be. So this is super helpful. Thanks a lot.

Can we see if we can support this without allocating a boxed iterator on every row?