LoLBurst

Benchmarking

• $ chmod +x testme.sh
• $ ./testme.sh

Tests against data that we already validated using grep -E "^PREFIX.*" | sort and stored in output files to determine that each backend works properly.

Uses measure program to determine memory usage.

Uses Python's time module to determine search / insertion time for each backend.

Installing

Dependencies: Python 3, pip, Bash, gcc

• $ git clone https://github.com/noyoshi/LoLBurst
• $ cd LoLBurst
• $ chmod +x download.sh
• $ ./download.sh
• This will install all requirements for python, and dowload the necessary asset
  files from the API

Running the server

• $ FLASK_APP=project.py flask run -h [IPADDRESS]

Using back ends for prefix search

• $ chmod +x benchmark.py
• $ cat INPUT_FILE | ./benchmark.py -[ust] "PREFIX"

Overview

• project.py is the flask application used to manage the server and handle
  recieving and sending data to and from the front end
• base.py contains the backend implementatins for the auto complete / prefix
  searching
• download.py and download.sh are used to download assets from the Riot Games API and League of
  Legends asset API
• helpers.py and item_helper.py handle API calls and create, manage, and
  parse JSON files (where we store and cache data)
• measure.c is the provided measure program to determine memory usage
• benchmark.py is a script to run the backends independently and check for
  time and accuracy
• test_project.py uses the unit testing libraries for Python and Flask to
  create and check accuracy of function calls within the project.py file
• testme.sh is the script we run to test the program
• tests/ is the folder that contains the outputs for various tests done in
  test_project.py
• templates/ is the folder for flask containing the HTML templates
• static/ contain static files for the front end, such as javascript and css
  files, and any downloaded image files
• *.json are all json files containing Riot API static data (cached locally as
  we are limited to 10 calls per hour for this data)

Contribution Breakdown

Timothy Blazek

• Initial front end development
• Created many of the api calling functions
• Created the prefix tree

Edward Atkinson

• Implemented the sorted and unsorted backends
• Created the javascript used to perform autocompletion on the front end
• Used javascript to send data to flask backend

Noah Yoshida

• Created tests and integrated them with Travis CI
• Uploaded code to remote server
• Created benchmarking functions and gathered data for them
• Created final demo video and presentation