This is the inverse of parsePathFragment().
Normalize the column names in partition specification, w.r.t.
Normalize the column names in partition specification, w.r.t. the real partition column names
and case sensitivity. e.g., if the partition spec has a column named monTh
, and there is a
partition column named month
, and it's case insensitive, we will normalize monTh
to
month
.
Given a partition path fragment, e.g.
Given a partition path fragment, e.g. fieldOne=1/fieldTwo=2
, returns a parsed spec
for that fragment as a TablePartitionSpec
, e.g. Map(("fieldOne", "1"), ("fieldTwo", "2"))
.
Given a partition path fragment, e.g.
Given a partition path fragment, e.g. fieldOne=1/fieldTwo=2
, returns a parsed spec
for that fragment as a Seq[(String, String)]
, e.g.
Seq(("fieldOne", "1"), ("fieldTwo", "2"))
.
Resolves possible type conflicts between partitions by up-casting "lower" types.
Resolves possible type conflicts between partitions by up-casting "lower" types. The up- casting order is:
NullType -> IntegerType -> LongType -> DoubleType -> StringType