mfinelli / whitebox-challenge

Excel challenge for Whitebox

whitebox code challenge

You need to have docker and docker compose installed and running. You must also provide the data.sql file uncompressed. It must also provide the necessary SQL create table statement e.g.,

CREATE TABLE `rates` (
  `id` int(11) unsigned NOT NULL AUTO_INCREMENT,
  `client_id` int(11) unsigned NOT NULL,
  `start_weight` decimal(10,2) DEFAULT NULL,
  `end_weight` decimal(10,2) DEFAULT NULL,
  `zone` varchar(1) DEFAULT NULL,
  `rate` decimal(10,2) DEFAULT NULL,
  `shipping_speed` varchar(15) DEFAULT NULL,
  `locale` enum('international','domestic') DEFAULT NULL,
  PRIMARY KEY (`id`)
)

There's a run script which will run both the docker build and docker up commands and then you can find the resulting spreadsheet in the output directory.

tl;dr

1. gunzip the provided data.sql.gz into the sql directory.
2. Run run.bash script. (Alternatively, run docker-compose build and docker-compose up --abort-on-container-exit)
3. Resulting excel spreadsheet is in output directory.

notes

• with only several thousand rows this isn't important, but an index on client_id would improve performance
• the provided sample output spreadsheep uses strings everywhere, but all the fields are numbers, so we could use some excel processing afterwards if we entered the data into the spreadsheet as numbers (or we just leave all the processing to SQL...)
• it would be nice to split the huge index.js into some smaller functions and files, write tests, add lint, etc
• i would have liked to use sqlite for this task instead of spinning up mysql, but the provided schema uses some mysql-only stuff (e.g. unsigned) so we stuck with mysql (actually using mariadb in docker compose)
• i only tested this on my macbook, but it's docker so in theory it works the same everywhere that docker runs
• i'm hoping that I set the retry timeout long enough on the initial query. the gist of what happens is when mariadb starts up the first time it needs to load in the schema and data in the sql directory and then it does a restart. if the node process tries to connect and query too quickly it will get a connection refused. right now we're only catching that one time and then waiting ten seconds for it to start back up before trying again. this was enough time on my machine, but if the first run errors out, try just running it again -- it won't need to do the data load and the first connection should work without issue.
• it needs better (or any) error handling. right now it will just blow up if the rates table doesn't exist for some reason (or any of the other myriad of things that might error along the way) -- we're assuming only the happy path for this demo
• if you want to run a different data set you'll need to remove the old mariadb container and then run it again as the data only gets initialized on the first run
• you shouldn't put more than one file into the sql directory since the mariadb image will play them all in alphabetical order so they could conflict with or clobber each other