Add a deduplicate preprocessor

[dotsdl]

See this comment for the context.

We are taking the approach of having composable preprocessing functions to make it easy for users to manipulate our DataFrame-based data structures without having to reinvent the wheel each time or quickly dive into complex and error-prone pandas-fu. To that end, we would like a preprocessor that simply and safely deduplicates records from our standard form DataFrames.

[xiki-tempula]

I think this is already done through the drop_duplicates keyword.