This application reads a CSV file of browser IDs and attempts to match them to common users. Benchmarking was done using both a 10 thousand and 100 thousand record CSV file. Both were processed in under 3 seconds on average on my machine so I believe this meets the requirement of working 'at scale'. If I were to spend more time on this I would incorporate Active Record into the models and implement a database to allow for querying and support of real life use scenarios.
Clone the repo and use Bundler to install the dependencies.
git clone https://github.com/eking1082/browser_id_challenge.git
cd browser_id_challenge
gem install bundler
bundle install
Run app.rb with the path to the CSV file you would like to process. The CSV headers are expected to be formatted as local id {number}.
For example, you can process the test CSV file by running:
ruby app.rb tests/test.csv
Simply use rake test to run all of the tests.