Exclude .git directory content when calculating package file count

[qtomlinson]

This was discovered during sanity test for upgrading crawler image to use node:18-bullseye. The file count for pod/cocoapods/-/SoftButton/0.1.0 was different:

• 78 in the node:16 based image
• 80 in the node:18-bullseye based image.
• The difference are due to the files in .git directory: .git/hooks/pre-merge-commit.sample, and .git/hooks/push-to-checkout.sample.

The content of .git directory should be excluded for the package file count.

Task: https://github.com/clearlydefined/crawler/issues/529

[qtomlinson]

@mpcen ready for review

[qtomlinson]

@jeffwilcox @mpcen rebased and ready for review.

[qtomlinson]

@elrayle ready for review

[qtomlinson]

The existing code only exclude .git directory itself, but not its containing files. This fix intends to exclude .git directory and its contents.

[qtomlinson]

@lumaxis @elrayle ready for review