When a script terminates before Undo_EndBlock() ...

Just a question for JCS about whether or not it's possible to have improved handling of what happens to Reaper's undo system if a script that sets its own Undo_BeginBlock() ends at an error before reaching the corresponding end block. If it's not already being done, would it be possible for Reaper to force the end of an undo point if a script throws an error, to prevent the undo queue from getting messed up? I've noticed this when making my own scripts - occasionally (:p) one makes mistakes and gets errors, and currently this means we must re-open Reaper to sort out undo/redo behaviour. It would also improve the use of Reascripts for all users, as if a buggy script is run it won't corrupt their project without them realising :)

You can use something like this to ensure execution of undo endblock: Code:

^ Ok (thanks I didn't know about contextmanager), but more to the point - wouldn't it be good for non-scrip-writing users, if this was done automatically, so that if they had a script that resulted in error, and hadn't been made like the above, reaper would protect them from it? I cannot imagine why, in a situation where a script is meant to set the end block, and doesn't because of an error, it would not be desirable for it to be sorted by Reaper - rather than just being added to the library, could it be built-in and automatic?