There are some issues with _extract_embedded_uri in Redirectors.pm. Sometimes it returns things that aren't valid, sometimes it doesn't find the embedded URI, and often it includes query parameters that aren't part of the embedded URI. The current regex approach may be too simple to handle complex cases (e.g. nested, %-encoded parameters). I'm trying to figure out if there's a simple fix, or if we need a full-blown URI parser. I can work on this over the next few days.
On Tue, Sep 1, 2026 at 02:29 PM, Giovanni Bechis <[email protected]> wrote: ------------------------------------------------------------------------------------------------------------------ CAUTION: External email from: gbechis@apache.org Use caution before clicking on links or opening attachments. ------------------------------------------------------------------------------------------------------------------ Hi, Apache SpamAssassin "trunk" branch seems to work well in production since some months and I would like to proceed to check regression tests on as much machines as possible to later release a rc1 version. If nobody has anything that they are working on that they intend to commit before release, or any other issues to bring up, I would like to go to RTC mode[¹] and I'll then prepare a release candidate that people can test. If all works fine I can later work on a final 4.0.3 release. I don't think we need a formal vote, a lazy consensus over the next ~96 hours should be enough, if anyone has a reason not to switch to RTC just speak up. Lazy consensus will expire on 09/05/2026 00:00:00 UTC. Cheers Giovanni [¹] https://www.apache.org/foundation/glossary.html#ReviewThenCommit (https://www.apache.org/foundation/glossary.html#ReviewThenCommit) -- Giovanni Bechis V.P. Apache SpamAssassin [email protected] (mailto:[email protected])
