Validate DataRow column length before buffered read

[SeanTAllen]

Closes #216.

_data_row matched the column length against -1 (NULL) and 0, but column lengths in 0x80000000..0xFFFFFFFE — invalid signed Int32 values — fell through to reader.block() without validation. The huge length errored at the buffered read and routed through the protocol-violation path, so no observable behavior changes; this makes the validation explicit at the parse site. Same flavor as #211.

Note on the test: both code paths (with and without the explicit if) error externally — reader.block() will reject any bogus length we can practically construct in a test. The test is therefore a behavioral regression test for "parser rejects malformed column length", not a strict counterfactual for the new if. Flagging in case you'd rather drop the test — I think keeping it is the right call but it's a judgment call.

[ponylang-main]

Hi @SeanTAllen,

The changelog - fixed label was added to this pull request; all PRs with a changelog label need to have release notes included as part of the PR. If you haven't added release notes already, please do.

Release notes are added by creating a uniquely named file in the .release-notes directory. We suggest you call the file 225.md to match the number of this pull request.

The basic format of the release notes (using markdown) should be:

## Title

End user description of changes, why it's important,
problems it solves etc.

If a breaking change, make sure to include 1 or more
examples what code would look like prior to this change
and how to update it to work after this change.

Thanks.