Support a glob resolver for listing and partitioning files in an object store

[begelundmuller]

This PR makes introduces a glob resolver that lists and optionally partitions files in an object store connector. In connection with that, it also:

• Adds a ListObjects function on drivers.ObjectStore
• Adds a mock_object_store driver for writing tests against object store drivers
• Refactors runtime.Resolve to support efficient iteration over a ResolverResult

This resolver will be used in a follow-up PR to introduce support for incremental ingestion from object stores by calling the glob resolver to obtain values to split ingestion by.

Example usage for listing out individual files:

# Example row:
#   {"path": "data/file.parquet", "uri": "s3://bucket/data/file.parquet", "updated_on": "2024-07-18T12:00:00Z"}
glob:
  connector: s3
  path: s3://bucket/**/*.parquet

Example usage for listing out files partitioned by directory:

# Example row:
#   {
#     "path": "data",
#     "uri": "s3://bucket/data", 
#     "files": ["data/file1.parquet", ...],
#     "updated_on": "2024-07-18T12:00:00Z"
#   } 
glob:
  connector: s3
  path: s3://bucket/**/*.parquet
  partition: directory

Example usage for listing out files partitioned using Hive partitioning:

# Example row: 
#   {
#      "path": "year=2024/month=07",
#      "uri": "s3://bucket/year=2024/month=07", 
#      "files": ["year=2024/month=07/file1.parquet", ...],
#      "updated_on": "2024-07-18T12:00:00Z",
#      "year": "2024",
#      "month": "07"
#   }
glob:
  connector: s3
  path: s3://bucket/**/*.parquet
  partition: hive

Example usage for doing post-processing with DuckDB (note this is an advanced feature needed to address some complex partition processing use cases):

# Example row: 
#   {
#      "path": "2024/07/18.parquet",
#      "previous_path": "2024/07/17.parquet",
#   }
glob:
  connector: s3
  path: s3://bucket/**/*.parquet
  transform_sql: SELECT path, lag(path) OVER (ORDER BY path) AS previous_path FROM {{ .table }}