provider-map-jobs.yml

name: Provider Map Jobs
on:
  # TODO: Remove this once behavior has been finalized
  push:
    branches:
      - vlad/provider-map-gha
  workflow_dispatch:
  schedule:
    - cron: "0 0 * * 0"

jobs:
  fetch_warehouse:
    name: Fetch Warehouse Updates
    runs-on: ubuntu-latest
    steps:
      - name: Checkout repository
        uses: actions/checkout@v4

      - name: Authenticate with Google
        uses: google-github-actions/auth@v1
        with:
            credentials_json: '${{ secrets.GCP_SA_KEY }}'

      - name: Set up Cloud SDK
        uses: google-github-actions/setup-gcloud@v1

      - name: Download providers.csv from warehouse
        run: |
          bq query \
              --quiet \
              --headless \
              --format=csv \
              --use_legacy_sql=false \
              --max_rows=500 \
              "SELECT * FROM \`mart_transit_database.dim_mobility_mart_providers\`" \
          > src/metadata/providers/providers.csv

      - name: Fix our CSV file
        run: |
          # Workaround because of...
          #   https://issuetracker.google.com/issues/315160970
          sed -i -n -e '/agency_name/,$p' src/metadata/providers/providers.csv

      - uses: actions/setup-python@v4
        with:
          python-version: '3.10'

      - name: Install dependencies, process providers, and update GeoJSON
        run: |
          pip install -r requirements.txt
          python .github/resources/process_providers.py

      - name: Create Pull Request
        uses: peter-evans/create-pull-request@v5
        with:
          title: Provider Map Data Auto Update
          body: |
            It's that time again! The warehouse has delivered new data for us to use. This is an automatic pull request created by the `provider-map-jobs.yml` workflow; it is triggered via a cron that runs every Sunday at midnight UTC.
          commit-message: "chore: auto-update provider data from warehouse"
          add-paths: |
            src/metadata/providers/providers.csv
            src/metadata/providers/counties.geojson
          base: main