Morgana-engine

Main changes
- Fixes the parsing of queries with `IN` and `NOT IN` keywords with some specific column names (15)
- Implements internally the lexing and parsing of the SQL query
- Removes package `sqlparse` as a dependency

Main changes
- Fixes the parsing of queries with `IN` and `NOT IN` keywords by previously grouping tokens (7)

Main changes
- Exports the selected dataframe in a spark-compatible formated. coercing the precision of datetime columns (2)
- Treats date columns with the appropriate backend (3)
- Columns that define partitions now can receive aliases in the query (4)
- The file format for partitioned tables in both S3 and FS storages don't need a final dash `-` before the extension (5)
- Schema data types for columns were changes: `integer` -> `int` and `number`-> `float`

- Schema formats changed for more tests
- Support for both relative and absolute paths in schema URIs
- Support for both paths and URIs in the schemas

- Initial version for enabling querying partitioned values in local FS or S3 storage
- No validations are made with respect to the query
- Only SELECT statement is supported