Testing automated tests for "big data"-style programming

What is it about?

Stream APIs exist in mainstream programming languages to enable processing large data sets (e.g., databases, data structures, and files) in relatively succinct commands, safely hiding the complexity of parallel programming. Changing from old data manipulation structures to the new APIs' usage frequently comes with the unintentional introduction of bugs, so good automated tests are a must. But how good are your tests? This work is all about proving the value of existing test suites by introducing common stream-related bugs in the original program and verifying if tests can detect them.

Featured Image

Photo by Gabriel Heinzer on Unsplash

Photo by Gabriel Heinzer on Unsplash

Why is it important?

This strategy, called mutation testing, helps improve the quality of existing test suites by exposing their fragilities. The better the tests, the less is spent removing bugs.