Creates a new reader that will read from the given hadoop path.
Creates a new reader that will read from the given hadoop path.
The dataSchema represents the schema that was written for the data files. This won't necessarily be the same as the hive metastore schema, because partition values are not written to the data files. We must include this here because some hive formats don't store schema information with the data, eg delimited files.
The projection is the schema required by the user which may be the same, or may be a subset if a projection is being used.