3 ms·
If you get an .xls which doesn't have very esoteric functions, I expect it to open about the same way in any Excel program and any other office suite. With CSV
by jcattle 6mo ago
If you get an .xls which doesn't have very esoteric functions, I expect it to open about the same way in any Excel program and any other office suite.
With CSV I do not have that expectation. I know that for some random user-submitted CSVs, I will have to fiddle. Even if that means finding the one row in thousand rows which has some null value placeholder, messing up the whole automatic inference.
- mcdonje 6mo agoYou're just saying when there's no filetype transfer, you don't have to deal with issues related to filetype transfer.
- jcattle 6mo agoNo. That's not at all what I'm saying. I am saying that a fixed CSV file will open differently depending on the program you open it with. Don't even need to transfer it. Opening a csv in pandas can be different than opening with polars, can be different to DuckDB, can be different to Excel. You've got not guarantees. There's no spec, and how edge cases (if you want to call how to serialize and deserialize a float an edge case) are handled is open to the implementation.