In 3.31.1, this self-contained script, which joins to an unnecessary table and adds what should be a redundant but harmless WHERE constraint on it:
https://gist.github.com/TallJimbo/d819876a77cfd79312ad48508cfdd8a2 returns incorrect results (which clearly violate the redundant constraint). Removing either of the two redundant constraints fixes the problem. The results in 3.27.2 and 3.30.1 were correct. Tests were done with SQLite packages from the Anaconda Python distribution, against regular file-based databases. I didn't see any active tickets that appeared to be related, but I did find https://www.sqlite.org/src/tktview?name=cf5ed20fc8, which was reported against 3.25 and seems to have fixed before any of the versions I tested - I'm not sure if this is just a reappearance of that bug, but it's at least very similar. Hope this is useful! Jim Bosch _______________________________________________ sqlite-users mailing list sqlite-users@mailinglists.sqlite.org http://mailinglists.sqlite.org/cgi-bin/mailman/listinfo/sqlite-users
