ci: add AGPL license bot (amend added files + PR comment)

2026-10-07 00:44:01 +01:00
parent 72f1e3a917
commit 0af97448c5
3 changed files with 218 additions and 0 deletions
@@ -0,0 +1,174 @@
+name: license-headers
+
+# Enforce AGPL-3.0 headers on every file ADDED by a pull request: amend the PR
+# head commit with the headers and comment the new commit hash on the PR.
+#
+# `runs-on` must match a registered runner label. The org's only runner offers
+# `alpine-latest` (container `node:alpine`), so jobs asking for anything else
+# (e.g. `alpine-3.24`) queue forever — that is why this bot never ran before.
+#
+# No long-lived secret is needed or wanted here: Gitea mints a per-run token
+# (`secrets.GITHUB_TOKEN`) scoped to this repository, which is enough to amend
+# the PR branch and comment. Never add a PAT/SSH key — an action that runs on
+# PRs (including forks) must stay keyless, and Gitea withholds repo/org secrets
+# from fork PRs anyway.
+
+on:
+  pull_request:
+    types: [opened, synchronize, reopened, ready_for_review]
+
+permissions:
+  contents: write
+  issues: write
+  pull-requests: read
+
+concurrency:
+  group: license-headers-${{ gitea.event.pull_request.number }}
+  cancel-in-progress: true
+
+defaults:
+  run:
+    shell: sh
+
+jobs:
+  addlicense:
+    runs-on: alpine-latest
+    steps:
+      - name: Install prerequisites
+        run: apk add --no-cache git curl ca-certificates jq
+
+      - name: Checkout pull request head branch
+        env:
+          TOKEN: ${{ secrets.GITHUB_TOKEN }}
+        run: |
+          set -e
+          REF="${{ gitea.event.pull_request.head.ref }}"
+          HEAD_REPO="${{ gitea.event.pull_request.head.repo.full_name }}"
+          BASE_REPO="${{ gitea.repository }}"
+          if [ "$HEAD_REPO" != "$BASE_REPO" ]; then
+            echo "Fork PR ($HEAD_REPO -> $BASE_REPO): the bot cannot push to a fork; skipping."
+            echo "fork_pr=true" >> "$GITHUB_ENV"
+            exit 0
+          fi
+          # Keep the scheme: the dev Gitea is HTTP-only, so a hardcoded https://
+          # here fails with "TLS connect error: wrong version number".
+          SERVER="${{ gitea.server_url }}"
+          SERVER="${SERVER%/}"
+          SCHEME="${SERVER%%://*}"
+          HOST="${SERVER#*://}"
+          git clone --depth 1 --branch "$REF" \
+            "${SCHEME}://oauth2:${TOKEN}@${HOST}/${BASE_REPO}.git" .
+          # Lease value for the later force push: the exact remote head we based on.
+          git ls-remote origin "refs/heads/${REF}" | cut -f1 > /tmp/head_before.txt
+          echo "checked out ${REF} at $(git rev-parse HEAD)"
+
+      - name: Install addlicense
+        if: env.fork_pr != 'true'
+        run: |
+          set -e
+          curl -sSL -o /tmp/addlicense.tar.gz \
+            https://github.com/google/addlicense/releases/download/v1.2.0/addlicense_v1.2.0_Linux_x86_64.tar.gz
+          tar -xzf /tmp/addlicense.tar.gz -C /tmp
+
+      - name: Add AGPL-3.0 headers to newly added files
+        if: env.fork_pr != 'true'
+        env:
+          TOKEN: ${{ secrets.GITHUB_TOKEN }}
+        run: |
+          set -e
+          SERVER="${{ gitea.server_url }}"
+          SERVER="${SERVER%/}"
+          REPO="${{ gitea.repository }}"
+          PR="${{ gitea.event.pull_request.number }}"
+
+          # Files added by this PR, straight from the API (no base history needed).
+          : > /tmp/added.txt
+          page=1
+          while : ; do
+            body="$(curl -sS -H "Authorization: token ${TOKEN}" \
+              "${SERVER}/api/v1/repos/${REPO}/pulls/${PR}/files?limit=100&page=${page}")"
+            n="$(printf '%s' "$body" | jq 'length')"
+            [ "$n" = "0" ] && break
+            printf '%s' "$body" | jq -r '.[] | select(.status=="added") | .filename' >> /tmp/added.txt
+            [ "$n" -lt 100 ] && break
+            page=$((page + 1))
+          done
+
+          # addlicense validates -ignore with doublestar and silently skips file
+          # types it does not understand, so pass every added file and let
+          # .licenseignore be the only escape hatch.
+          set -- -v -f .license-header.tpl -c "Dcntrlized" -y "$(date +%Y)"
+          if [ -f .licenseignore ]; then
+            while IFS= read -r line; do
+              case "$line" in ""|\#*) continue ;; esac
+              set -- "$@" -ignore "$line"
+            done < .licenseignore
+          fi
+          found=0
+          while IFS= read -r f; do
+            [ -n "$f" ] || continue
+            [ -f "$f" ] || continue
+            found=1
+            /tmp/addlicense "$@" "$f"
+          done < /tmp/added.txt
+          [ "$found" = "1" ] || echo "No newly added files."
+
+          if git diff --quiet; then
+            echo "no_changes=true" >> "$GITHUB_ENV"
+          else
+            echo "no_changes=false" >> "$GITHUB_ENV"
+          fi
+
+      - name: Amend the commit and force-push
+        if: env.fork_pr != 'true' && env.no_changes == 'false'
+        env:
+          TOKEN: ${{ secrets.GITHUB_TOKEN }}
+        run: |
+          set -e
+          REF="${{ gitea.event.pull_request.head.ref }}"
+          SERVER="${{ gitea.server_url }}"
+          SERVER="${SERVER%/}"
+          SCHEME="${SERVER%%://*}"
+          HOST="${SERVER#*://}"
+          git config user.name "Dcntrlized Auto CopyLeft Bot"
+          git config user.email "[email protected]"
+          git add -A
+          git commit --amend --no-edit
+          SHA="$(git rev-parse HEAD)"
+          printf '%s\n' "$SHA" > /tmp/commit_sha.txt
+          echo "amended to $SHA"
+          BEFORE="$(cat /tmp/head_before.txt)"
+          git push --force-with-lease="refs/heads/${REF}:${BEFORE}" \
+            "${SCHEME}://oauth2:${TOKEN}@${HOST}/${{ gitea.repository }}.git" \
+            "HEAD:refs/heads/${REF}"
+
+      - name: Comment on the pull request
+        if: env.fork_pr != 'true' && env.no_changes == 'false'
+        env:
+          TOKEN: ${{ secrets.GITHUB_TOKEN }}
+        run: |
+          set -e
+          SERVER="${{ gitea.server_url }}"
+          SERVER="${SERVER%/}"
+          REPO="${{ gitea.repository }}"
+          PR="${{ gitea.event.pull_request.number }}"
+          SHA="$(cat /tmp/commit_sha.txt)"
+          cat > /tmp/pr-comment.md <<EOF
+          ## Uh oh! Unlicensed code detected!
+
+          You just tried to push **unlicensed code**! This repository is covered by the
+          [AGPL-3.0](https://www.gnu.org/licenses/agpl-3.0.html) license, and all code
+          requires a license header.
+
+          Thankfully, the bot caught it and amended your commit
+          ([\`${SHA}\`](${SERVER}/${REPO}/commit/${SHA})) with the correct licensing.
+
+          If this looks wrong, tag **@Admin** so we can take a look!
+          EOF
+          BODY="$(jq -n --rawfile b /tmp/pr-comment.md '{body:$b}')"
+          curl -sS -X POST \
+            "${SERVER}/api/v1/repos/${REPO}/issues/${PR}/comments" \
+            -H "Authorization: token ${TOKEN}" \
+            -H "Content-Type: application/json" \
+            -d "$BODY"
+          echo "comment posted"