pearu commented on PR #50146: URL: https://github.com/apache/arrow/pull/50146#issuecomment-6272143637
Sure. I'll split it into these PRs, each with its own issue: Fixes to the strptime timestamp parser. These affect existing callers (CSV timestamp columns, compute `strptime`, Gandiva `to_date`, Flight cookie `Expires`), and each can land on its own: 1. Windows: the vendored musl `strptime` does not support `%a`, `%b`, `%p` and related directives, because `nl_langinfo` is not available. 2. Impossible dates (e.g. Feb 30) roll over into the next month, and a zero day is taken as the first of the month. These are kept in one PR, since the zero-day fix relies on the date check. 3. Years out of range overflow `int` or wrap around in `date::year` (reachable with `%s`). 4. Leap seconds are wrapped into the next minute. New feature, in two PRs with two new issues under GH-41488: 5. `timestamp_parsers` as a fallback for date32/date64 columns. 6. The same for time32/time64 columns. 5 and 6 build on the fixes and will say which PRs must land first. I'll keep this PR open as a reference until the split PRs have landed. -- 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]
