FabledGrimoire/.forgejo/workflows/ci.yml

# CI runs first; build only proceeds if all checks pass.
#
# Push to dev:        typecheck + lint + test + build :dev + :<sha>
# Push to main:       typecheck + lint + test + build :latest + :<sha>
# Tag v* (release):   typecheck + lint + test + build :latest + :<version> + :<sha>
#
# main is the production line, so :latest tracks main's tip. Every push also
# gets an immutable :<sha>, the rollback point. A release (a v* tag created on
# main in the Forgejo UI) adds the dated :<version>.
#
# A newer push to the same ref cancels the older run; dev and main never cancel
# each other. Pull requests don't trigger runs: branch pushes already gate.
#
# Forgejo Actions doesn't reliably honour `on.push.branches` as a filter, so
# every job repeats the ref check in its `if:`.
#
# Required secrets (repo → Settings → Secrets → Actions):
#   REGISTRY_USER   — your Forgejo username
#   REGISTRY_TOKEN  — Forgejo PAT with write:packages scope
name: CI & Build

on:
  push:
    branches: [dev, main]
    tags: ["v*"]
    paths:
      - "src/**"
      - "frontend/**"
      - "tests/**"
      - "pyproject.toml"
      # The lock decides what is installed, so a lock-only change must run CI.
      - "uv.lock"
      - "alembic/**"
      - "alembic.ini"
      - "Dockerfile"
      # The plugin ships straight from git (via .claude-plugin/marketplace.json),
      # not in the image, so a push here is its release. See the `plugin` job.
      - "plugin/**"
      - ".claude-plugin/**"
      - "scripts/check_plugin.py"
      - ".forgejo/workflows/ci.yml"
  # Manual run from the Forgejo Actions UI, e.g. to rebuild an image.
  workflow_dispatch: {}

# A newer push cancels the older run on the same branch. Tag runs are never
# cancelled, so a release build can't be cut off mid-flight.
concurrency:
  group: ci-${{ github.ref }}
  cancel-in-progress: ${{ !startsWith(github.ref, 'refs/tags/') }}

# Least-privilege default. Jobs that need more (build pushes to the
# registry) upgrade explicitly.
permissions:
  contents: read

env:
  REGISTRY: git.fabledsword.com
  IMAGE: git.fabledsword.com/bvandeusen/fabledgrimoire

jobs:
  typecheck:
    name: TypeScript typecheck
    # Gate dev, main, and v* tags; reject any other ref (see header note).
    if: github.ref == 'refs/heads/dev' || github.ref == 'refs/heads/main' || startsWith(github.ref, 'refs/tags/v')
    runs-on: python-ci
    container:
      image: git.fabledsword.com/bvandeusen/ci-python:3.14
    steps:
      - uses: actions/checkout@v6

      - name: Cache npm download cache
        uses: actions/cache@v4
        # Non-fatal: a cache-backend error just means installing without it.
        continue-on-error: true
        with:
          path: ~/.npm
          key: npm-cache-${{ hashFiles('frontend/package-lock.json') }}
          restore-keys: npm-cache-

      - name: Install dependencies
        run: npm ci
        working-directory: frontend

      - name: Type check
        run: npx vue-tsc --noEmit
        working-directory: frontend

  # Guards the one part of this repo that ships to users without a build step.
  # scripts/check_plugin.py says what it checks and what it can't.
  plugin:
    name: Plugin hooks
    if: github.ref == 'refs/heads/dev' || github.ref == 'refs/heads/main' || startsWith(github.ref, 'refs/tags/v')
    runs-on: python-ci
    container:
      image: git.fabledsword.com/bvandeusen/ci-python:3.14
    steps:
      # Keep this a bare `uses:`: with a `with:` block the action fails to
      # extract on act_runner ("Cannot find module .../dist/index.js"). The
      # version check diffs two trees, so a depth-1 fetch of main is enough.
      - uses: actions/checkout@v6

      # Per-job rather than in the image: Grimoire is shellcheck's only consumer.
      # Install nothing else here; the hooks use only POSIX tools, and the
      # smoke test runs on a bare image to prove it.
      - name: Install shell tooling
        run: |
          apt-get update -qq
          apt-get install -y -qq --no-install-recommends shellcheck

      # On main the comparison would be against itself, so only the syntax and
      # pattern checks mean anything there.
      - name: Check plugin hooks and manifest
        run: |
          if [ "${{ github.ref }}" = "refs/heads/main" ]; then
            python3 scripts/check_plugin.py --no-version
          else
            git fetch --no-tags --depth=1 origin main:refs/remotes/origin/main
            python3 scripts/check_plugin.py
          fi

  lint:
    name: Python lint
    if: github.ref == 'refs/heads/dev' || github.ref == 'refs/heads/main' || startsWith(github.ref, 'refs/tags/v')
    runs-on: python-ci
    container:
      image: git.fabledsword.com/bvandeusen/ci-python:3.14
    steps:
      - uses: actions/checkout@v6

      # ruff is pre-installed in the ci-python image.
      - name: Lint
        run: ruff check src/ scripts/

      # Checks the frontend's CSS against the committed, generated theme.css.
      # Fails only on an unresolvable var() reference; hard-coded literals are
      # reported, not gated.
      - name: Design token check
        run: python3 scripts/check_design_tokens.py --report-literals

      # Reports elements whose classes have modifier rules but no base rule,
      # which vue-tsc can't see. Not gated: a bare wrapper is legitimate, so
      # the signal is the count growing.
      - name: Dangling style check
        run: python3 scripts/check_dangling_styles.py

  test:
    name: Python tests
    if: github.ref == 'refs/heads/dev' || github.ref == 'refs/heads/main' || startsWith(github.ref, 'refs/tags/v')
    runs-on: python-ci
    container:
      image: git.fabledsword.com/bvandeusen/ci-python:3.14
    steps:
      - uses: actions/checkout@v6

      - name: Cache uv packages
        uses: actions/cache@v4
        with:
          path: ~/.cache/uv
          # Keyed on the lock, which decides the installed set.
          key: uv-${{ hashFiles('uv.lock') }}
          restore-keys: uv-

      # Installs exactly what uv.lock pins, so green means these exact versions
      # passed. `--locked` fails if the lock is stale against pyproject, so a
      # dependency change needs a deliberate `uv lock`.
      - name: Install locked dependencies
        env:
          UV_PROJECT_ENVIRONMENT: /opt/venv
        run: uv sync --locked --extra dev

      # The hook execution tests run on a bare image (no jq), like a user's
      # machine might be.
      - name: Run tests
        # Integration tests (real Postgres) run in the `integration` job below.
        run: /opt/venv/bin/python -m pytest tests/ -q -m "not integration"

  # Real Postgres, built by the real migrations: covers the async SQLAlchemy
  # paths unit stubs can't reach.
  #
  # Service hostnames aren't routable on this runner, so the step finds the
  # Postgres container's bridge IP. The job has no `name:` and a separator-free
  # key because act_runner names service containers after it.
  integration:
    if: github.ref == 'refs/heads/dev' || github.ref == 'refs/heads/main' || startsWith(github.ref, 'refs/tags/v')
    runs-on: python-ci
    container:
      image: git.fabledsword.com/bvandeusen/ci-python:3.14
    env:
      # Read at import time. DATABASE_URL is built from the discovered IP below.
      SECRET_KEY: ci_integration_placeholder
    services:
      postgres:
        # Same image as the shipped compose files: Postgres 17 with pgvector.
        image: pgvector/pgvector:pg17
        env:
          POSTGRES_USER: grimoire
          POSTGRES_PASSWORD: ci_integration
          POSTGRES_DB: grimoire_test
        options: >-
          --health-cmd "pg_isready -U grimoire"
          --health-interval 10s
          --health-timeout 5s
          --health-retries 10
    steps:
      - uses: actions/checkout@v6
      # Same locked install as the unit lane, so both test the same versions.
      - name: Install locked dependencies
        env:
          UV_PROJECT_ENVIRONMENT: /opt/venv
        run: uv sync --locked --extra dev
      # Prints two facts about the runner that the steps below depend on.
      # Every command is guarded so the diagnostic can't fail the lane.
      - name: Runner facts (shell, service hostnames)
        run: |
          echo "--- which shell runs a run: step ---"
          readlink -f /bin/sh || echo "/bin/sh: not a symlink"
          ps -p $$ -o comm= || true
          echo "--- is a service reachable by its hostname? ---"
          getent hosts postgres \
            || echo "no — 'postgres' does not resolve; the bridge-IP lookup is still required"
      - name: Integration suite (resolve service IP, migrate, test)
        run: |
          set -eux
          echo "=== container landscape (diagnostic for the name filter) ==="
          docker ps -a --format '{{.ID}} {{.Image}} -> {{.Names}}'
          # Only this job's services: the docker daemon is shared, so a job-name
          # filter can match another repo's Postgres. act names a task's containers
          # GITEA-ACTIONS-TASK-<n>-...; read <n> from our own container (its id is in
          # the /etc/hostname bind mount) and require exactly one match.
          SELF=$(grep -o '/containers/[0-9a-f]\{64\}/' /proc/self/mountinfo | head -n1 | cut -d/ -f3 || true)
          SELF=${SELF:-$(hostname)}
          TASK=$(docker inspect -f '{{.Name}}' "$SELF" | grep -o 'GITEA-ACTIONS-TASK-[0-9]*-')
          test -n "$TASK"
          PG=$(docker ps --filter "name=$TASK" --filter "ancestor=pgvector/pgvector:pg17" -q)
          test "$(printf '%s\n' "$PG" | grep -c .)" -eq 1
          PG_IP=$(docker inspect -f '{{range .NetworkSettings.Networks}}{{.IPAddress}}{{end}}' "$PG")
          test -n "$PG_IP"
          export DATABASE_URL="postgresql+asyncpg://grimoire:ci_integration@${PG_IP}:5432/grimoire_test"
          # Wait for Postgres. The run: shell is dash (no bash /dev/tcp), so use Python.
          /opt/venv/bin/python - "$PG_IP" <<'PY'
          import socket, sys, time
          for _ in range(30):
              try:
                  socket.create_connection((sys.argv[1], 5432), timeout=2).close()
                  break
              except OSError:
                  time.sleep(1)
          else:
              sys.exit("postgres did not become reachable")
          PY
          # Real migrations build the schema; the maintenance tests then run
          # VACUUM (ANALYZE) and read pg_stat_user_tables against it.
          /opt/venv/bin/alembic upgrade head
          /opt/venv/bin/python -m pytest tests/ -v -m integration

  build:
    name: Build & push image
    # `plugin` is not in needs: the plugin ships from git, not in the image, so
    # holding the image back wouldn't stop a bad hook. A red `plugin` job still
    # turns the run red. `integration` is: the image carries the migrations and
    # queries that lane tests.
    needs: [typecheck, lint, test, integration]
    # dev → :dev, main → :latest, tag → :latest + :<version>; all get :<sha>.
    if: github.ref == 'refs/heads/dev' || github.ref == 'refs/heads/main' || startsWith(github.ref, 'refs/tags/v')
    runs-on: python-ci
    container:
      image: git.fabledsword.com/bvandeusen/ci-python:3.14
    permissions:
      contents: read
      packages: write
    steps:
      - uses: actions/checkout@v6
        with:
          # Full history for the version derivation. Today it needs only HEAD's
          # commit time, but a shallow clone fails silently (a too-low value) if
          # the derivation ever reads further back.
          fetch-depth: 0

      - name: Generate image tags and version
        id: tags
        # POSIX `case`, not bash `[[ ]]`: the run: shell is dash, where `[[`
        # fails silently and only the :<sha> tag would be pushed.
        run: |
          TAGS="${{ env.IMAGE }}:${{ github.sha }}"

          # Three values kept separate; the Dockerfile describes each.

          # 1. Ordering key: build time in minutes since 2020-01-01, so it only
          #    goes up. Commit time would go down when an older commit is
          #    rebuilt; a commit count isn't monotonic across branches.
          BUILD_KEY=$(( ( $(date -u +%s) - 1577836800 ) / 60 ))

          # 2. Name: commit time, so the same source reads the same on every lane.
          COMMIT_TS=$(git log --format=%ct -1 HEAD)
          BUILD_NAME=$(date -u -d "@$COMMIT_TS" +%Y.%m.%d.%H%M)

          # 3. Channel: its own value, never part of the name.
          CHANNEL="dev"
          case "${{ github.ref }}" in
            refs/heads/dev)
              TAGS="$TAGS,${{ env.IMAGE }}:dev"
              ;;
            refs/heads/main)
              # main is the production line, so it publishes :latest.
              TAGS="$TAGS,${{ env.IMAGE }}:latest"
              CHANNEL="stable"
              ;;
            refs/tags/*)
              TAGS="$TAGS,${{ env.IMAGE }}:latest,${{ env.IMAGE }}:${{ github.ref_name }}"
              CHANNEL="stable"
              ;;
          esac
          echo "value=$TAGS" >> $GITHUB_OUTPUT
          echo "build_name=$BUILD_NAME" >> $GITHUB_OUTPUT
          echo "build_key=$BUILD_KEY" >> $GITHUB_OUTPUT
          echo "channel=$CHANNEL" >> $GITHUB_OUTPUT

      # No disk-space prune step: the runner's Docker daemon is shared, and
      # `docker system prune -af` here deleted images other repos' jobs had
      # just pulled and were about to start (CI-runner #16). The runner host
      # prunes nightly instead.

      - name: Set up Docker Buildx
        uses: docker/setup-buildx-action@v4

      - name: Log in to Forgejo registry
        uses: docker/login-action@v4
        with:
          registry: ${{ env.REGISTRY }}
          username: ${{ secrets.REGISTRY_USER }}
          password: ${{ secrets.REGISTRY_TOKEN }}

      - name: Build and push
        uses: docker/build-push-action@v7
        with:
          context: .
          push: true
          provenance: false
          tags: ${{ steps.tags.outputs.value }}
          # The commit lets the image's self-report be checked against :<sha>.
          build-args: |
            BUILD_VERSION=${{ steps.tags.outputs.build_name }}
            BUILD_KEY=${{ steps.tags.outputs.build_key }}
            BUILD_CHANNEL=${{ steps.tags.outputs.channel }}
            BUILD_COMMIT=${{ github.sha }}
          # Layer cache kept in the registry, so a build starts warm even after
          # the runner prunes. `mode=max` caches intermediate layers too.
          cache-from: type=registry,ref=${{ env.IMAGE }}:cache
          cache-to: type=registry,ref=${{ env.IMAGE }}:cache,mode=max,ignore-error=true