You definitely can determine this automatically from the existing tags. For anything with patch version > 0, it's trivial since you want the reference for previous version to be `patch_version - 1`. For the first release in a major.minor release line, you would need to figure out the correct previous major.minor.patch release and use that.
Normally I would say this is pretty easy, just list tags, find ones that match the right pattern, split segments on `.` characters, convert each to integers, and sort. However, this does get a bit messier with Kafka because of the switch in release numbering (from 4 digits in pre-1.0 to 3 digits in post-1.0), so you'd have to normalize to 4 digits, sort, then make sure you drop any extra digits from post-1.0 versions. It'd be nice to get this all automated and the ergonomics of the script are nicer if they it is, but I wouldn't block merging this on that. This is still better than what committers do today, which is to just construct this all manually. [ Full content available at: https://github.com/apache/kafka/pull/5572 ] This message was relayed via gitbox.apache.org for [email protected]
