Summary

Add flatten logic into the ViaQ plugin to make it easier to selectively choose which labels to flatten for this feature

Acceptance Criteria

• Plugin optionally configurable to flatten labels
• Plugin optionally removes "kubernetes.labels" except for configurable exclusions
• Verify the "flatten" transformation results in `record_key = [key=value,key=value]` to satisfy regression functionality
• Verify transformations leaves exclusions in 'kubernetes.labels'

Notes

• The original intent for flatten labels was to preserve the number of fields consumed when pushing to Elasticsearch. Too many field could could potential result in exceeding the max limit due to the possibility of any service being capable of having any number of labels.