Fix #156, exceptions from FORCED_STOP not rethrown #157
Conversation
|
This doesn't really fix the issue because there is no backtrace saved with the exception so it's a little useless the way this pr is. |
| if e !== nothing && !isa(e, ForcedStop) | ||
| nlopt_exception = nothing | ||
| rethrow(e) | ||
| throw(e) |
There was a problem hiding this comment.
Too bad we lose the original stack trace; I guess in the future we could also potentially save the catch_backtrace() and wrap everything in our own exception type that shows this along with the original exception.
There was a problem hiding this comment.
I think wrapping the exception type might break try-catch blocks around the optimize call that expect specific exception types to be thrown.
I tried this: ikirill@c7aed00
|
I experienced this as well, would be really nice to just pass through the exception directly. However, if a custom exception type is needed to provide additional information, callers would probably be fine catching that and then inspecting the inner exception. What work is needed here? |
|
Sorry I forgot about this PR. I think it works, I've been using my own local NLopt and it throws exceptions with backtraces correctly. |
|
I also added saving backtraces to be shown with the exception. Obviously I can't rethrow the same exception, you have to save it from the function being called by C, and then throw it again on the Julia side once the C function returns. So maybe it's a little awkward. |
|
Just found this after haggling with debugging an error inside my objective without any stacktraces. |
|
@ikirill Any chance you've been working on this? It would be incredible to have this functionality. |
|
Closing in favor of #194 |
6 participants
It's not possible to
rethrowoutside acatchblock (Julia gives an error). It was only checking for errors when ret was INVALID_ARGS which is wrong.