Implement custom format for object keys uploaded by S3Writer

[jfzunigac]

This PR introduces 2 new configs to S3Writer to allow for more customizable format for object uploads:

• filenamePattern : A regex to extract fields/tokens out of the filename
• filenameTokens : A list of tokens extracted from filenamePattern, these are extracted automatically but can also be set for testing

keyPrefix and filenameFormat are also replaced with keyFormat so that tokens extracted from the flename can be used in keyFormat by following the syntax: %{TOKEN}, e.g:

# Configuration for writer
writer.type=s3
writer.s3.maxFileSizeMB=5
writer.s3.minUploadTimeInSeconds=30
writer.s3.bucket=my-bucket
writer.s3.keyFormat=%{service}/%{index}/my_log.%TIMESTAMP
writer.s3.maxRetries=3
writer.s3.filenamePattern=^(?<service>[a-zA-Z0-9]+)_.*_(?<index>\\d+)\\.log$

Additionally, we add some default formatters:

• %UUID
• %TIMESTAMP
• %HOST
• %LOGNAME