JeonDaehong commented on PR #17774: URL: https://github.com/apache/iceberg/pull/17774#issuecomment-5979630199
Since you mentioned having some doubts about the math, I compared this PR against the old bit-at-a-time implementation from `main` in a throwaway test. * **Every single-bit input:** 1–10 columns, 1–16 bytes per input, and every possible output size. Since both implementations move each input bit to exactly one output bit, this covers the complete mapping space. **4,607,680 cases, 0 differences.** * **Random and patterned bytes** (`0x00`, `0xFF`, `0x80`, etc.): 1–10 columns, 1–17 bytes, every truncated output size, with a reused buffer pre-filled with garbage. **504,900 cases, 0 differences.** Hope this helps. -- 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] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
