Improve crd-extractor script

[kishorviswanathan]

• Fetch CRDs in parallel
• Avoid use of associative array
• Handle space in file path

[eyarz]

@kishorviswanathan thank you for improving the script!
did you have a chance to run some tests with your version of the script?

[kishorviswanathan]

@eyarz The networking CRDs I extracted from GKE was using the updated script. It worked fine. I don't have access to any other clusters, so I am not really sure how to test it. Do you have something in mind ?

[eyarz]

@userbradley @ElanHasson can I get you help in reviewing this?
not a bash expert here and because this is a core utility I want to make sure someone will have a second look here before merging :)

[onedr0p]

FWIW I was able to test this and runs it just fine.

[eyarz]

@onedr0p thank you for the feedback!
I will wait for one more reviewer before merging it because it's a critical component in the repo

[ElanHasson]

Looking at this now

[ElanHasson]

@onedr0p thanks for the PR!

One of the things is see often is (example)

Maintainer: "How did you generate this PR?"
PR Creator:  [something other than the extractor]

I wonder if there is a way for the tool to "mark" the files as having been created by it.

I just made #449 to help with this