bake.yml

name: Bake

# Build catalog cells once, centrally, so the website and desktop app can select,
# download, and assemble the same verified inputs.
#
# ── READ THIS BEFORE ADDING A COUNTRY TO THE MATRIX ────────────────────────────────
#
# **A country-scale bake does not fit a GitHub runner, and this workflow is not where
# the real bakes happen.** A `ubuntu-latest` runner has 4 vCPU, 16 GB of RAM and
# ~14 GB of usable disk. The German extract alone is 4.8 GB *before* the packer
# allocates its node table, and the artifacts it produces are hundreds of megabytes
# each. There is no flag that fixes that.
#
# So this workflow is **one caller** of `obc-bake`, deliberately sized for the small
# end of the curated list: city-states and small Bundesländer (Bremen, Hamburg, Berlin,
# Saarland), refreshes of already-cached regions, and — the job that runs on every
# schedule regardless — the published-catalog version guard. The CLI is the product;
# it runs the same way on a workstation, which is where Germany, Bayern, NRW,
# Austria and Switzerland are baked:
#
#     cargo run --release -p obc-bake -- bake --out ~/bake --summary-json ~/bake/run.json
#     OBC_MAPS_BASE_URL=https://maps.openbikecomputer.org/cell-catalog \
#       cargo run --release -p obc-bake -- publish ~/bake --target r2
#
# Nothing in `obc-bake` assumes CI: no runner-only paths, no GitHub-specific output,
# and its self-contained tree can be retained across runs: cells, sidecars, regions,
# schema, skins, satellites, and the root all live below one directory.
#
# ── Publishing ────────────────────────────────────────────────────────────────────
#
# Publishing needs R2 credentials, which live in repository secrets and are only
# passed on an explicit `workflow_dispatch` with `publish: true`. A scheduled run
# bakes and reports; it never uploads. Content first, root last, is the publisher's
# own invariant (`OBCC_Spec.md` §11) and is enforced in `obc-bake`, not
# here — this file only decides *whether* to publish.

on:
  workflow_dispatch:
    inputs:
      regions:
        description: "Regions to bake (comma-separated ids from host/obc-bake/regions.toml; blank = the small-region set)"
        required: false
        default: ""
      force:
        description: "Re-bake even when nothing changed"
        type: boolean
        default: false
      publish:
        description: "Publish the result to R2 (needs the repository secrets)"
        type: boolean
        default: false
  schedule:
    # Weekly, Monday 04:00 UTC — after Geofabrik's nightly extracts have settled.
    # Refreshes the small regions and, whatever happens to those, runs the guard.
    - cron: "0 4 * * 1"
  pull_request:
    paths:
      # The guard's whole job is to notice an OBCM bump before the catalog rots, so
      # it runs on the PRs that could cause one.
      - "firmware/obc-formats/**"
      - "host/obc-pack/src/catalog.rs"
      - "host/obc-bake/**"
      - ".github/workflows/bake.yml"

concurrency:
  # Never two bakes into one bucket. Not cancel-in-progress: killing a bake halfway
  # wastes hours and (unlike a test run) leaves a cache half-warmed.
  group: bake-${{ github.workflow }}
  cancel-in-progress: false

permissions:
  contents: read

env:
  # The regions small enough to bake on a hosted runner. Everything else in
  # regions.toml is a workstation job — see the header.
  SMALL_REGIONS: >-
    europe/germany/bremen,
    europe/germany/hamburg,
    europe/germany/berlin,
    europe/germany/saarland

jobs:
  # ── The mandatory re-bake guard (OBCC_Spec.md §10) ────────────────────────────────
  #
  # Three mechanisms enforce "an OBCM bump invalidates every baked artifact". Two are
  # in the repository and already exist: the generator refuses a mixed tree, and
  # `catalog.rs`'s PINNED_OBCM_VERSION breaks the packer's tests when the format
  # constant moves. Neither can see what is actually being served — a bump can be
  # merged, released, and the pin honestly moved while the CDN still hands v10 files
  # to a v11 firmware until somebody remembers to re-bake.
  #
  # That is this job. It needs the live manifest URL, so it is configured, not
  # hardcoded: set the `OBC_CATALOG_URL` repository *variable* (the same one
  # deploy-site.yml passes to the builder) and it starts checking; leave it unset and
  # it says "skipped" and passes, because a project that has not published a catalog
  # yet must not carry a red check about it.
  #
  # It lives here rather than in ci.yml on purpose: it is a fact about a deployment,
  # not about the tree, and it needs the network. Making it a *required* check once a
  # catalog exists is a branch-protection toggle, not a code change.
  obcm-version-guard:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4
      - uses: actions-rust-lang/setup-rust-toolchain@v1
        with:
          cache: false
      - uses: Swatinem/rust-cache@v2
        with:
          workspaces: .
          key: host
      - uses: ./.github/actions/setup-host-deps
      - name: Is the published catalog still this build's OBCM version?
        env:
          OBC_CATALOG_URL: ${{ vars.OBC_CATALOG_URL }}
        run: cargo run --release --locked -p obc-bake -- check-obcm-version

  bake:
    # Not on pull_request: a PR must not spend twenty minutes packing maps.
    if: github.event_name != 'pull_request'
    runs-on: ubuntu-latest
    timeout-minutes: 300
    steps:
      - uses: actions/checkout@v4
      - uses: actions-rust-lang/setup-rust-toolchain@v1
        with:
          cache: false
      - uses: Swatinem/rust-cache@v2
        with:
          workspaces: .
          key: host
      - uses: ./.github/actions/setup-host-deps

      - name: Disk and memory before the bake
        # Printed because the first thing to check when a bake dies is whether it was
        # the runner's 14 GB of disk rather than the packer.
        run: |
          df -h /
          free -h

      # Extracts are cached across runs: a refresh that re-downloads 4 GB every week
      # is not a refresh. The key rolls weekly so the cache follows Geofabrik's own
      # cadence; `restore-keys` lets a new week start from last week's files, which the
      # runner then revalidates by HEAD (Last-Modified + Content-Length) per region.
      - name: Cache Geofabrik extracts
        uses: actions/cache@v4
        with:
          path: ~/.cache/obcm/geofabrik
          key: geofabrik-${{ github.run_id }}
          restore-keys: geofabrik-

      # The bake tree survives across runs so an unchanged region can be *skipped*
      # rather than re-packed; the skip is keyed on content hashes, so a restored tree
      # that disagrees with the extracts re-bakes rather than lying.
      - name: Cache the bake tree
        uses: actions/cache@v4
        with:
          path: ~/bake
          key: bake-tree-${{ github.run_id }}
          restore-keys: bake-tree-

      - name: Bake
        # Dispatch inputs arrive as environment variables, never interpolated into the
        # script text: `${{ }}` inside `run:` is textual substitution, and a region id
        # is a string a human typed.
        env:
          INPUT_REGIONS: ${{ inputs.regions }}
          INPUT_FORCE: ${{ inputs.force }}
          OBC_MAPS_BASE_URL: ${{ vars.OBC_MAPS_BASE_URL }}
        run: |
          regions=()
          if [ -n "$INPUT_REGIONS" ]; then
            IFS=',' read -ra ids <<< "$INPUT_REGIONS"
            for id in "${ids[@]}"; do regions+=("${id// /}"); done
          else
            IFS=',' read -ra ids <<< "$SMALL_REGIONS"
            for id in "${ids[@]}"; do regions+=("${id// /}"); done
          fi
          force=()
          if [ "$INPUT_FORCE" = "true" ]; then force+=("--force"); fi
          # No `set -e` games: obc-bake exits non-zero if any job failed or any region
          # ended with no artifact, and that is exactly the signal this step wants.
          cargo run --release --locked -p obc-bake -- bake \
            --out ~/bake \
            --summary-json ~/bake-summary.json \
            "${regions[@]}" "${force[@]}"

      # `if: always()` — the summary is most valuable exactly when the step above
      # failed, because it names which regions did not bake and why.
      - name: Run summary
        if: always()
        run: |
          if [ -f ~/bake-summary.json ]; then
            python3 - <<'PY' >> "$GITHUB_STEP_SUMMARY"
          import json, pathlib
          data = json.loads(pathlib.Path.home().joinpath("bake-summary.json").read_text())
          print(f"### Bake — OBCM v{data['obcm_version']}, recipe v{data['recipe_version']}\n")
          print("| band | cells | partial | bytes |")
          print("| --- | --- | --- | --- |")
          for band in data["bands"]:
              print(f"| {band['band']} | {band['cells']} | {band['partial_cells']} | {band['bytes']} |")
          failures = [plan for plan in data["plans"] if plan.get("error")]
          if failures:
              print("\n**Failed source sets:**")
              for plan in failures:
                  print(f"- {', '.join(plan['sources'])}: {plan['error']}")
          if data["uncovered_regions"]:
              print("\n**Regions with no artifact:** " + ", ".join(data["uncovered_regions"]))
          PY
          fi

      - name: Upload the run summary
        if: always()
        uses: actions/upload-artifact@v4
        with:
          name: bake-summary
          path: ~/bake-summary.json
          if-no-files-found: warn

      # Only on an explicit dispatch. The credentials are never present on a
      # scheduled run, so a schedule cannot publish even by accident.
      #
      # Note what this job's tree is: the cached tree plus any regions refreshed in
      # this run. The generated root offers exactly what is complete in that tree.
      - name: Publish to R2
        if: ${{ github.event_name == 'workflow_dispatch' && inputs.publish }}
        env:
          OBC_R2_ACCOUNT_ID: ${{ secrets.OBC_R2_ACCOUNT_ID }}
          OBC_R2_BUCKET: ${{ secrets.OBC_R2_BUCKET }}
          OBC_R2_PREFIX: ${{ vars.OBC_R2_PREFIX }}
          OBC_R2_ACCESS_KEY_ID: ${{ secrets.OBC_R2_ACCESS_KEY_ID }}
          OBC_R2_SECRET_ACCESS_KEY: ${{ secrets.OBC_R2_SECRET_ACCESS_KEY }}
          OBC_MAPS_BASE_URL: ${{ vars.OBC_MAPS_BASE_URL }}
        run: |
          # `update` first: a runner image whose package index has aged out 404s on
          # the install otherwise, and it would do so at the last step of a long job.
          sudo apt-get update
          sudo apt-get install -y rclone
          test -n "$OBC_MAPS_BASE_URL" || { echo "set the OBC_MAPS_BASE_URL variable first"; exit 1; }
          cargo run --release --locked -p obc-bake -- publish ~/bake \
            --base-url "$OBC_MAPS_BASE_URL" --target r2