Conversation
…e can release the resource with the destructor
|
I think this goal of this PR makes a lot of sense, but I think this needs a review from someone that's more experienced in C++ than me. @Y-- what do you think about this? Are there easier/better ways to achieve the same? |
This LGTM, a pretty typical use of RAII pattern :-) |
|
Thanks, then I'll add it to more places as required and undraft 👍 |
|
Moved this to 0.2.0, in that release we'll do a more holistic approach to solving these issues. |
|
@Y-- Does this still make sense? Or should we close this one? It seems like you haven't needed this for your cleanup so far? |
Yeah I think something like that still make sense. I'll drive it through or re-implement a similar version when I come to this. |
This PR feels like a continuation of #176 and is also related to #93
When we throw a C++ exception we have to make sure that all postgres related resources get released, this is done through the destructor of the ScopedPostgresResource class
Opening as a draft, hopefully we can talk about the implementation or even alternatives that solve the same issue