DKER2 opened a new pull request, #11508: URL: https://github.com/apache/arrow-rs/pull/11508
# Which issue does this PR close? - Part of #9493. # Rationale for this change #11232 changes how `VariantPath` strings are parsed, and there is currently no benchmark covering path parsing. As requested in https://github.com/apache/arrow-rs/pull/11232#issuecomment-5879196321, this adds the benchmark on its own so it is available on `main` as a baseline before that PR proceeds. # What changes are included in this PR? A new Criterion target, `variant_path`, in `parquet-variant`. It parses six path strings with `VariantPath::try_from`: | Case | Path | |---|---| | `dot_fields` | `foo.bar.baz` | | `bracket_quoted_fields` | `['foo']['bar']['baz']` | | `bracket_indexes` | `[0][1][2]` | | `bracket_wildcards` | `[*][*][*]` | | `bracket_escaped_field` | `['escaped\]field']` | | `mixed` | `a['b.c'].d[2]['3']` | Run with `cargo bench -p parquet-variant --bench variant_path`. For reference, results on `main` (aarch64 macOS): | Case | Time | |---|---:| | `dot_fields` | 30.6 ns | | `bracket_quoted_fields` | 259 ns | | `bracket_indexes` | 140 ns | | `bracket_wildcards` | 137 ns | | `bracket_escaped_field` | 144 ns | | `mixed` | 293 ns | # Are these changes tested? This PR only adds a benchmark; no library code changes. # Are there any significant user-facing changes? No. # AI usage Claude Code was used to write the benchmark and run it. I reviewed the result and can own and debug the change. 🤖 Generated with [Claude Code](https://claude.com/claude-code) -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
