data-refresh.yml

name: Refresh dataset

# Regenerates lib/sepomex_db.csv from the latest official SEPOMEX export and
# opens a PR when the data changes.
#
# The scheduled run is self-contained: `rake data:refresh` downloads the export
# straight from Correos de México (the CodigoPostal_Exportar.aspx WebForms
# postback, handled by FetchSepomexExport) and regenerates the CSV — no mirror
# URL to provision. Pass a `xml_url` on manual dispatch to import a specific,
# already-hosted `CPdescarga.xml` instead (e.g. to reproduce a past export).
on:
  workflow_dispatch:
    inputs:
      xml_url:
        description: Optional URL to a specific SEPOMEX CPdescarga.xml (skips the live download)
        required: false
  schedule:
    - cron: "0 6 1 * *" # 06:00 UTC on the 1st of each month

permissions:
  contents: write
  pull-requests: write

concurrency:
  group: data-refresh
  cancel-in-progress: false

jobs:
  refresh:
    name: Refresh dataset
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v7

      - name: Set up Ruby
        uses: ruby/setup-ruby@v1
        with:
          bundler-cache: true

      - name: Regenerate lib/sepomex_db.csv from the latest export
        if: github.event.inputs.xml_url == ''
        run: bundle exec rake data:refresh

      - name: Regenerate lib/sepomex_db.csv from a provided URL
        if: github.event.inputs.xml_url != ''
        run: |
          curl -fSL "${{ github.event.inputs.xml_url }}" -o "${RUNNER_TEMP}/CPdescarga.xml"
          bundle exec rake "data:import_xml[${RUNNER_TEMP}/CPdescarga.xml]"

      - name: Open a PR if the dataset changed
        uses: peter-evans/create-pull-request@v8
        with:
          branch: chore/refresh-dataset
          add-paths: lib/sepomex_db.csv
          commit-message: Refresh lib/sepomex_db.csv from the latest SEPOMEX export
          title: Refresh postal-code dataset
          labels: data
          body: |
            Automated refresh of `lib/sepomex_db.csv` from the latest official
            SEPOMEX export (`rake data:refresh`).

            After merging, the next release image bakes in the new data; to update
            a running/local database, run `bin/rails db:reset && rake data:loadev`.