To run the script Babashka is preferred, you can find installation instructions here

To run the tests I opted for clojure/kaocha as Babshka was fiddly to get working with the tests, installation instructions are here

The data is required to be in the data directory otherwise the script will just fail

To run the script, from the project root run:

bb src/calculate_price_differential.clj

To run the tests, from the project roo run:

./bin/kaocha

For a REPL, I used vim-iced and to start that up (with kaocha) I used:

iced repl --dependency=kaocha-nrepl:0.1.1 --middleware=kaocha-nrepl.core/wrap-kaocha

• I spent far too long trying to get the data into a format that was easy to output, resulting in many iterations through the entire data set
• I avoided using regexes for no real reason, which led to parsing being pretty easy to break
• With a larger data set this script would not scale well, however in a real setting you are most likely to only improve the script when required, so I felt this more adequately reflected real world situations
• Lack of confiurability of location of the data, this would be useful, but again would likely be added later
• Wasted far too much time trying to get tests working in Babashka
• It would have been good to have kept track of the receipt in which the line item was, as it became hard to manually test the script
• The tests should really test edge cases and invalid data, but as I was using them as a design tool I found it hard to add that on later