Extraction Process

Game data comes from Heroes.DataParser (https://github.com/koliva8245/HeroesDataParser) with as little intervention as possible. Direct output from HDP is available in /raw. Automated process is as follows:

• Whenever there's an update to https://github.com/Blizzard/heroprotocol IFTTT notifies a member of the team (via Pushover)
• Team member first launches Battle.net and patches HotS, then runs the extract script
• The extract scripts checks all dependencies and repos for updates then calls HeroesDataParser with the following: "HeroesData" --description 3 --storage-path "$dirLive/dirPTR" --extract all --output-directory "$tmpDir" --hero-warnings --localization all --build 99999 --json
• The extract script compresses image files using ImageMagick: magick mogrify -resize 64x64 -strip -depth 8 *.png
• The extract script commits all extracted data to https://github.com/tattersoftware/heroes-talents and notifies the server
• The server pulls extracted game data and runs GameData->import (to update the internal database) then GameData->export (to create heroes-talent JSON hero files)
• The server updates and commits https://github.com/tattersoftware/heroes-talents the new JSON files and notifies a collaborator for review
• A collaborator submits a pull request to heroes-talents with the updated branch data