Fixes #634

Note: there is a memory footprint reduction, there are some ideas how to optimise in follow-up PR

• added benchmark tests, and created a fixture to instantiate a large chart (copy files)
• added Rules, to easy maintain add/remote file patterns
• created an issue Function loader.Load() potential leaking resources helm/helm#31007
• added LoaderWithRules, helm itself not sure if required such capability, but required optimisation for helm-unittest, as we executing dozens of helm.Load operations, while helm <command> only runs it once

I have tried deep_copy https://github.com/helm-unittest/helm-unittest/blob/main/pkg/unittest/deep_copy.go, it does not work very well,

Following approach has it's own pros/cons

From the pros side

• we could maintain our set of rules for Load operation to improve resource consumption, like include/exclude values files or snapshots for each test
• in follow-up we could easily add caching, so instead of evaluating rules on each for idx, testJob := range s.Tests , we could do them ouside of a loop and pass only files required for current test.

cons:

• loose coupling with helm

CPU reduction 4 times
Memory reduced in 2.5x times

Before

With the fix

At some point, depends on where or not issue is accepted, I could try to submit few PRs to helm, but the process there is quite lengthy.