bug: use credentials when scanning

[sa7mon]

Bug

As reported in #328, credentials stored in ~/.aws/credentials or a path located by the AWS_SHARED_CREDENTIALS_FILE environment variable were not being used when scanning buckets. Because of this, any results from scanning AWS buckets would be potentially missing AUTH_USERS permissions.
This was an unfortunate oversight when re-writing S3Scanner in Go mostly due to the account ownership of the test buckets not being fully understood at that time. Consequently, the unit tests to verify this functionality did not report accurately.

Changes

• A new IAM user has been provisioned to use for unit tests. This user has custom policies applied to it so that permission checks produce the results that would be produced if using an account that does not own the bucket (blocking access to READ and READ_ACL operations on certain test buckets).
• Fix grant check in ParseAclOutputv2
• Minor cleanup refactors
• Load profile and credentials when scanning
  • S3Scanner will follow the procedure documented here when choosing which profile to use and where to load credentials from. For example, an alternative profile can be used by setting the AWS_PROFILE environment variable.
  • Note: the only exception is that the EC2 IMDS feature has been disabled. If you want to use S3Scanner inside an EC2 instance, you will need to store the credentials to use in the credentials file.
• Update ClientMap to use a compound key when storing clients. Now we can store both anonymous and authenticated clients for each region.
• Add additional logging when -verbose is supplied

Closes #328

[github-actions]

Package	Line Rate	Complexity	Health
github.com/sa7mon/s3scanner/bucket	91%	0	✔
github.com/sa7mon/s3scanner/cmd/regioncheck	55%	0	➖
github.com/sa7mon/s3scanner/cmd/s3scanner	19%	0	❌
github.com/sa7mon/s3scanner/db	73%	0	➖
github.com/sa7mon/s3scanner/permission	28%	0	❌
github.com/sa7mon/s3scanner/provider	79%	0	✔
github.com/sa7mon/s3scanner/worker	49%	0	❌
Summary	62% (595 / 959)	0	➖