LZO Parquet compression is not supported

[nuno-faria]

Describe the bug

According to the docs (both https://datafusion.apache.org/user-guide/configs.html and https://datafusion.apache.org/user-guide/sql/format_options.html#parquet-format-options), lzo should be supported for Parquet files. However, it is not implemented in the parquet crate, and it seems that there are no plans to implement it (apache/arrow-rs#6970).

It might be better to update the docs to mention that lzo is not supported, or simply remove it altogether.

To Reproduce

> copy (select 1) to 't.parquet' options (compression 'lzo');

thread 'tokio-runtime-worker' (36172) panicked at ...
called `Result::unwrap()` on an `Err` value: NYI("The codec type LZO is not supported yet")
note: run with `RUST_BACKTRACE=1` environment variable to display a backtrace

Expected behavior

No response

Additional context

It's also GPLv2, so maybe it will never be supported anyway.

[SAJALVIJAY19]

Hi, I’m interested in contributing to DataFusion and starting with this issue.
I noticed that LZO is documented as supported but not implemented.
I’d like to start by improving the documentation and/or error messaging to make this clearer.
Please let me know if this approach makes sense.

[nuno-faria]

Hi, I’m interested in contributing to DataFusion and starting with this issue. I noticed that LZO is documented as supported but not implemented. I’d like to start by improving the documentation and/or error messaging to make this clearer. Please let me know if this approach makes sense.

Thanks @SAJALVIJAY19, but there is already a PR to fix it: #19726.