Preventing terminal hang/memory or CPU spikes when analysing large repositories

[penderi]

Executing the current version of CodeMedic occupies a large amount of CPU and Memory under certain situations.

Current theory is when referencing private nuget repositories.
Tested on ~40k lines of code, with dozens of package references. Nuget feeds include the public Microsoft feed and 2 internal authenticated feeds.

It would help greatly to skip or contain issues around those, or fix potential perf issue - Jeff mentioned caching previously.

Loving the twitch casting Jeff. Outstanding quality. Loving the AI jawn (hope I’ve got that right, I’m from ~London)

[csharpfritz]

For .NET applications, i'm not too worried about memory use because .NET like Java applications has a garbage collector that will clean up memory when it's done use using that memory or when there's memory pressure. I think we're not in bad shape there. We can certainly optimize in the future to use more memory efficient string operations, but I don't think this is an immediate concern while we're trying to get functionality working.

I'd like to zoom in a little bit on the performance numbers that you're seeing 40,000 lines of code and dozens of package references. When we get to performance tuning and tweaking, this will be a good benchmark and use case to start with