exercises-in-programming-style/31-double-inverse-multiplexer/README.md

Style #30

Very similar to style #29, but with an additional twist

Constraints:

• Input data is divided in chunks, similar to what an inverse multiplexer does to input signals

• A map function applies a given worker function to each chunk of data, potentially in parallel

• The results of the many worker functions are reshuffled in a way that allows for the reduce step to be also parallelized

• The reshuffled chunks of data are given as input to a second map function that takes a reducible function as input

Possible names:

• Double inverse multiplexer
• Map-reduce
• Hadoop style