Automatic simplification fails with singletons in special cases about portion HOT 4 CLOSED

Of course one of the most important information was only implicit in the examples: the problem concerns the merge function.

from portion.

Hello,

Thank you for having reported this issue! I'll have a look asap and try to release a hotfix asap as well :-)

from portion.

I found the issue. When intervals are simplified, the underlying atomic intervals are sorted by their lower bound. and an attempt to merge them is made. Since (2, 3) and [2] have the same lower bound, the order in which they are provided defines the order in which they will be considered (and that's why (1, 2) | [2] | (2, 3) works): it tries first to merge (1, 2) and (2, 3) (they can't), then (2, 3) and [2] (which results in [2, 3)).

I'll confirm this asap, fix it, and release a hotfix.

from portion.

Fixed in 39e2564

Fix will be released in a few minutes as part of version 2.1.2.

from portion.