Given a csv file with "buckets" and another csv file with purchase records, categorize the records into buckets based on some specificity rules.
Assumptions Assumed English words in UTF-8 in the files, thus didn't use any character decoding (https://stackoverflow.com/questions/6797984/how-to-convert-string-to-lowercase-in-python). Seems like Python3 handles it though.
If the "*,*,*" bucket does not exist in the purchase buckets, algorithm creates one at the beginning as per the example above.
Design
I initially wanted to use a regex to capture the data fields needed from the csv file, but it became too cryptic to read and switched to regular string methods. I also learned that reading from a csv can be done with pandas, but I suspect it's an overkill.
My solution started without using classes and object. Looking for the abstractions as I solved.
Testing
Ideally I would run different sets of csv files on every run, but I kept it at one set for now and focused on writing a comprehensive test suite.
Other Notes
Must use python 3.6 or newer. I had a working version with OrderedDict (for lower versions of python, but decided to switch to the regular dictionary structure since ordering is supported starting python3.6).