phiresky
Especially simple refactorings like renaming a variable cause huge amount of work when merging for minimal benefit.
fyi: I've copied the unit tests part of this to #4559 and fixed it a bit, I'll work more on it the next days.
this was done in #3420
Thank you for the issue. I'm a bit confused as to how this happens. I see in your PR you add the path to the db to the cache key....
Thanks for the info. That could indeed be, if you close one connection the new one may well get the same pointer location.. I didn't consider that case. _If_ that...
Another (a bit hacky) workaround might be to call [sqlite3_auto_extension()](https://sqlite.org/c3ref/auto_extension.html) with a callback that clears the cache. Then the cache should be cleared whenever a new connection is created, which...
Actually I found an even easier method. Since we already have a hook when a new connection is opened, I can simply invalidate the cache when that happens! I tested...
in my opinion, the least pino should do is compare the number of printf args to the number of arguments passed and either log warnings or simply log the objects...
> That is an expensive operation I concede that the code I sent is fairly expensive. > it iterates through the format string once in order to replace the tokens...
sounds fine El dom, 7 jul 2024, 13:15, Mimoja ***@***.***> escribió: > Hi :) > I forked your repository to add a few small commits which rename the pyhon >...