Make initial Mock cleanup more performant

[nohwnd]

PR Summary

There is initial cleanup of orphaned mocks that happens in Invoke-Pester and makes sure that there are no Mock aliases in place coming from previously cancelled runs. In Pester v4 we only cleaned up the main session state, but in v5 we pick up every loaded module and clean up there. This adds up to the point where it can take several seconds (in docker container) to do this cleanup.

This PR optimizes the cleanup to just query functions, rather than all cmdlets and filtering them.

PR Checklist

• PR has meaningful title
• Summary describes changes
• PR is ready to be merged
  • If not, use the arrow next to Create Pull Request to mark it as a draft. PR can be marked Ready for review when it's ready.
• Tests are added/update (if required) (there are no tests for this functionality in the first place imho)
• Documentation is updated/added (if required)

[nohwnd]

It is tested: [n] ERROR: - Invoke-Pester cleans up orphaned mock hooks and aliases in modules ->

will need to fix.

[fflaten]

Agree with the idea, but are we able to mute these? Could cause errors in a follow-up run.

> & (Get-Module PSReadLine) { dir function:\ }                                                                                                   
Exception: /workspaces/Pester/src/functions/Pester.SessionState.Mock.ps1:592:9
Line |
 592 |          throw "Mock data are not setup for this scope, what happened? …
     |          ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
     | Mock data are not setup for this scope, what happened?

Not sure which property being enumerated is causing this.

> & (Get-Module PSReadLine) { Get-Command Pester* } | % parameters
InvalidOperation: 
Line |
  34 |  … icparam { & $MyInvocation.MyCommand.Mock.Get_MockDynamicParameter -Cm …
     |                ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
     | The expression after '&' in a pipeline element produced an object that was not valid. It must result in a command name, a script block, or a CommandInfo object.

Maybe silenced by checking for Mock-property before invoking it? Even better, are we able to self-delete based on similar condition there and/or begin?

[nohwnd]

I've reverted the change, and remove the functions as before, but now just querying the item provider for functions rather than asking Get-Command to give me all publc commands (including cmdlets) that are available in all modules. This seems to take 7ms (and 1 ms after warmup), vs 55ms (and 30ms after warmup), and should not depend on the number of modules that are imported.

[nohwnd]

@SQLDBAWithABeard could you try this before I merge please?

[SQLDBAWithABeard]

@SQLDBAWithABeard could you try this before I merge please?

Totally removes those calls from the top of my performance list and gives me other things to work on :-)

[nohwnd]

Rob reported this comparison of calls to Get-Command -Name PesterMock_* vs Get-ChildItem function:/PesterMock_*

With 8 modules loaded
20 Milliseconds
2 Milliseconds
With 19 Modules loaded
314 Milliseconds
5 Milliseconds
With 21 Modules loaded inc dbatools/dbacheks
220 Milliseconds
8 Milliseconds

Insignificant difference between 1st and following runs