Missing function to compare `DataFrameSchemas` with order

[Jolanrensen]

df1.schema() == df2.schema() (or df1.schema().compare(df2.schema()).isEqual()) returns true,
even if their columns are ordered like "a, b, c" and "b, c, a", respectively.

This is useful from the Kotlin perspective, as it's safe to cast df1 to the schema of df2; they have the same columns, accessors and could be converted to the same class.

However, in DataFrame, columns have an order. This can be seen when converted to JSON, rendered to HTML, or viewed as a table.
Also, df1 == df2 returns false if there is a different order, which is very inconsistent with the schema equality check above.

We have the move operation to change the order. However, to test move or for a user to check that two DataFrame's schemas match exactly, there's no easy way to do so at present.

[Jolanrensen]

I propose deprecating ComparisonResult.Equals for Matches (or something like that) and .isEqual() to isMatching().

Then we create a new function DataFrameSchema equals (exactly) function that takes order into account, and we use that for == instead.

I think for the library this would not have that much impact, as we are using .compare() almost everywhere we check if schemas match.

There was only one place I found == being used in the codebase to check if schemas match