File CDK: S3 config adapter (top-level config options)

[clnoll]

The existing S3 connector offers some config options that we do not want to provide for new connections. However, we still want to support them rather than making changes that will break users' syncs.

At the top-level, the options provided by the S3 connector include:

• dataset: should be mapped to name.
• path_pattern: should be mapped to globs. I don't think additional changes are required but we should double-check.
• path_prefix: we should attempt to create the appropriate globs by combining this with path_pattern.
• user_schema: should be mapped to input_schema.
• start_date: this will be wired through the file-based CDK in File CDK: add start_date config option #28136

This will also include handling config options used by the existing S3 connector including:

• bucket
• aws_access_key_id
• aws_secret_access_key
• endpoint

These options are currently nested under a provider key (along with path_prefix and start_date), but for the new S3 connector we'll have them be at the top level.

This ticket will involve 1) adding a version to the new file-based CDK config object so that we can differentiate it from the existing S3 config and 2) creating a new S3-specific FileBasedConfig object, to be used by the new S3 connector. The new config object will include logic to adapt the old config to the new config format. In addition to top-level config options, this pattern will be used by the config options specific to each file type (described in separate tickets).