fix(GeneratorQueryHandler): replay fetchMore requests made while active#1618
fix(GeneratorQueryHandler): replay fetchMore requests made while active#1618brenns10 wants to merge 1 commit into
Conversation
Do not drop fetchMore calls that arrive while a batch is still active. Queue one pending fetch request and replay it after activeChanged(false). This handles reentrant fetchMore calls triggered during results insertion and prevents multi-chunk generators from stalling after the first chunk. Fixes: albertlauncher/albert-plugin-python#12
ty, once again steph.
thats okay unless i am the first to review their output ;D Regarding this PR: In my opinion, this is really hard to argue without a formal protocol for the lazy fetch mechanism using It makes sense to assume that the view calls |
|
Given that you sent a fix already I have to note that the entire query handling related code is really easy to mess up and I have been bitten by nasty bugs stemming from Qt itself. Although this PR may fix the build for your Qt version this may not hold for all others out there. I just want to really understand what is going on there before solving a problem. |
2 participants
Hello, I ended up digging in a bit and finding a fix (and test case) for the issue I reported earlier with the Python plugin. With this change my example plugin works as expected.
To be totally transparent - I did use AI/LLM tools in this PR, but I have read and understood everything I'm submitting here.