`polars-cli` complain that you're using `scan_csv` to read a compressed file even when using `read_csv`

[apcamargo]

Checks

• I have checked that this issue has not already been reported.

• I have confirmed this bug exists on the latest version of the Polars CLI.

Reproducible example

〉select * FROM read_csv('data.tsv.gz');
Error: cannot scan compressed csv; use `read_csv` for compressed data

### Issue description

`polars-cli` complains that it can't scan compressed files, and suggests using `read_csv`. My understanding here is that`polars-cli` uses `scan_csv` behind the scenes even though the command we use is called `read_csv`. In this case, maybe the error message should be adjusted to make sense in the context of `polars-cli`?

### Expected behavior

The function should either (1) be able to read compressed files (which would require using `read_csv` as a backend), or (2) have an error message saying that `polars-cli` can't read compressed files (or suggesting an alternative way of reading those)

### Installed version

0.7.0

[apcamargo]

I've been looking at table_functions.rs and noticed that read_csv there actually uses a LazyCsvReader. This got me thinking, because if you're familiar with the Python API, read_csv there doesn't give you a lazy frame. So, it kind of feels like the name read_csv might be a bit misleading. Wouldn't scan_csv be a more fitting name? Although, I'm sure there's a reason for these naming choices.

Also, I'm curious if there's a way to use the eager API with these table functions. What's the use case for the table functions, aside from their use in polars-cli? Is there any documentation available that goes into more detail about this?