Shaka T-Shirt Raffle Code

find-all-contributors.sh

#!/bin/bash
# Find all contributors to date.  These users will get increased odds of winning
# in the raffle.

# These repos will be checked to list all contributors.
REPOS="shaka-player shaka-packager shaka-streamer shaka-player-embedded"

API_URL="https://api.github.com/repos/"

# 4 pages of data covers all PRs to date in all repos.  Some pages will be empty
# for some repos, and that's fine.
for i in 1 2 3 4; do
  for repo in $REPOS; do
    # Make API requests as the shaka-bot account, using a pre-generated token.
    curl -u shaka-bot:$(cat token) \
        "$API_URL/google/$repo/pulls?state=all&per_page=100&page=$i" \
        > prs.$repo.$i.json
  done
done

# Pull the GitHub usernames from all PRs, normalize to lowercase, and output a
# list of unique contributor usernames to a text file.
jq -r -s '.[][].user.login' prs.*.json | \
    tr A-Z a-z | \
    sort -u > contributors-nov-1.txt

select-winners.py

#!/usr/bin/env python3
"""Select winners from the Shaka T-Shirt Raffle."""

import csv
import random

# Get a set of all past contributors, which was pre-computed using the GitHub
# API.
with open('contributors-nov-1.txt') as f:
  contributors = set(f.read().split('\n'))

entrants = []

# Open a CSV download of the data from the Google Form for the raffle.
with open('entrants.csv') as f:
  reader = csv.reader(f)
  # Throw away the header row.
  next(reader)
  # Track usernames we have seen.  No stuffing the ballot box!
  seen = set()

  for row in reader:
    # Columns are timestamp, email, GitHub username.
    _, email, username = row

    # Usernames in the contributors list have already been normalized to
    # lowercase.  Do the same now for the CSV inputs.
    username = username.lower()
    # Some people managed to add spaces to the form, so strip the usernames of
    # leading and trailing whitespace, too.
    username = username.strip()

    # Track usernames we have seen.  No stuffing the ballot box!
    if username in seen:
      continue
    seen.add(username)

    # If the entrant is a contributor, they get 5x the chances to win.
    chances = 5 if username in contributors else 1
    for i in range(chances):
      entrants.append((email, username, chances))

# Shuffle the list in place.
random.shuffle(entrants)

# Now the top entries are the winners.  Past contributors got multiple entries,
# and they are therefore more likely to have an entry close to the top of the
# list.  But now we should remove second and subsequent entries so that each
# person is in the list only once.

winners = []
for entry in entrants:
  if entry not in winners:
    winners.append(entry)

# Finally, a human process will screen the top entrants for eligibilty.  Some
# may have to be skipped if, for example, US law would prohibit us from shipping
# to them for some reason.  We don't know in advance how many times this could
# happen, so we will not limit the output to 20 entries.

for winner in winners:
  print(winner)