bec_widgets/.github/workflows/knowledge-graph.yml

# Builds the graphify knowledge graph and publishes it with the release it describes.
#
# Triggered after "Continuous Delivery" (semantic-release) completes: releases happen
# exactly when releasable code (feat/fix/perf) lands on the default branch, so the graph
# tracks actual code changes rather than every push. A release-tag gate skips CD runs
# that released nothing — and, because it only proceeds when HEAD carries a v* tag, it
# also guarantees the graph describes exactly the commit that was released. Note:
# chaining via workflow_run is deliberate — tags pushed by a workflow's own GITHUB_TOKEN
# do not fire `on: push: tags` workflows.
#
# Two publication paths:
#   * release runs attach knowledge-graph.tar.gz to the GitHub Release for that tag, so
#     every developer can fetch the map for a given version, unauthenticated and
#     permanently, at a URL derived from the version alone;
#   * all runs also upload a job artifact, which is what makes PR and workflow_dispatch
#     dry-runs inspectable.
#
# The graph is never committed (graphify-out/ is gitignored). actions/cache carries the
# AST cache between runs, so rebuilds are incremental. CI builds are AST-only
# (deterministic, no LLM key); the semantic docs layer and community labels ride along
# via the cache when a locally refreshed one was uploaded.

name: knowledge-graph

on:
  workflow_run:
    workflows: ["Continuous Delivery"]
    types: [completed]
  workflow_dispatch: {}

permissions:
  contents: read

jobs:
  build-graph:
    if: github.event_name != 'workflow_run' || github.event.workflow_run.conclusion == 'success'
    runs-on: ubuntu-latest
    timeout-minutes: 30

    # write is needed only to attach the tarball to an existing release; the workflow
    # never pushes commits, tags or branches.
    permissions:
      contents: write

    steps:
      - uses: actions/checkout@v4
        with:
          fetch-tags: true

      - name: Release gate
        id: gate
        run: |
          TAG="$(git tag --points-at HEAD | grep '^v' | head -n 1 || true)"
          echo "sha=$(git rev-parse HEAD)" >> "$GITHUB_OUTPUT"
          echo "tag=$TAG" >> "$GITHUB_OUTPUT"
          if [ "${{ github.event_name }}" != "workflow_run" ]; then
            # dry-run contexts: build, but there is no release to attach to
            echo "run=true" >> "$GITHUB_OUTPUT"
          elif [ -n "$TAG" ]; then
            echo "Released $TAG - building and attaching the graph."
            echo "run=true" >> "$GITHUB_OUTPUT"
          else
            echo "CD run made no release - skipping graph build."
            echo "run=false" >> "$GITHUB_OUTPUT"
          fi

      - uses: actions/setup-python@v5
        if: steps.gate.outputs.run == 'true'
        with:
          python-version: "3.12"

      - name: Restore graph cache
        if: steps.gate.outputs.run == 'true'
        uses: actions/cache@v4
        with:
          path: graphify-out
          key: graphify-${{ runner.os }}-${{ github.sha }}
          restore-keys: |
            graphify-${{ runner.os }}-

      - name: Install graphify
        if: steps.gate.outputs.run == 'true'
        # pinned so CI node ids stay identical to locally built graphs;
        # bump deliberately together with local installs
        run: pip install --quiet 'graphifyy==0.9.5'

      - name: Build knowledge graph
        if: steps.gate.outputs.run == 'true'
        env:
          PYTHONHASHSEED: "0" # deterministic clustering
          # identity written into build_meta.json; GITHUB_SHA is overridden because on
          # workflow_run it names the triggering run's head, not what we checked out
          GITHUB_SHA: ${{ steps.gate.outputs.sha }}
          GRAPH_RELEASE_TAG: ${{ steps.gate.outputs.tag }}
        run: python .github/scripts/graphify/graph_rebuild.py . .github/scripts/graphify/graph_spec.json

      - name: Package graph
        id: package
        if: steps.gate.outputs.run == 'true'
        run: |
          # name the asset after the version it describes, so a downloaded file is
          # identifiable on disk without unpacking it first
          VERSION=$(python -c "import json; print(json.load(open('graphify-out/build_meta.json'))['version'])")
          NAME="knowledge-graph-v${VERSION}.tar.gz"
          tar -czf "$NAME" -C graphify-out graph.json GRAPH_REPORT.md build_meta.json
          echo "name=$NAME" >> "$GITHUB_OUTPUT"
          echo "Packaged $NAME ($(du -h "$NAME" | cut -f1)):"
          cat graphify-out/build_meta.json

      - name: Attach graph to release
        # only release runs have a tag to attach to; dry-runs stop at the artifact below
        if: steps.gate.outputs.run == 'true' && steps.gate.outputs.tag != ''
        env:
          GH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
        run: gh release upload "${{ steps.gate.outputs.tag }}" "${{ steps.package.outputs.name }}" --clobber

      - name: Publish graph artifact
        if: steps.gate.outputs.run == 'true'
        uses: actions/upload-artifact@v4
        with:
          name: knowledge-graph
          path: ${{ steps.package.outputs.name }}
          retention-days: 90