ci: migrate deploying doc site job to GHA

Migrate the deployment doc site job from CircleCI to Github Actions

.circleci/config.yml

@@ -1,237 +1,16 @@
-# Configuration file for https://circleci.com/gh/angular/components
+# This config is remaining in place to prevent pull requests failing because of CircleCI config missing.
 
-# Note: YAML anchors allow an object to be re-used, reducing duplication.
-# The ampersand declares an alias for an object, then later the `<<: *name`
-# syntax dereferences it.
-# See http://blog.daemonl.com/2016/02/yaml.html
-# To validate changes, use an online parser, eg.
-# http://yaml-online-parser.appspot.com/
-
-var_1: &docker_image cimg/node:18.17.0
-var_2: &docker-browser-image cimg/node:18.17.0-browsers
-
-# **Note**: When updating the beginning of the cache key, also update the cache key to match
-# the new cache key prefix. This allows us to take advantage of CircleCI's fallback caching.
-# Read more here: https://circleci.com/docs/2.0/caching/#restoring-cache.
-var_3: &cache_key v16-ng-mat-{{ checksum ".bazelversion" }}-{{ checksum "tools/postinstall/apply-patches.js" }}-{{ checksum "WORKSPACE" }}-{{ checksum "yarn.lock" }}
-# We want to invalidate the cache if the postinstall patches change. In order to apply new
-# patches, a clean version of the node modules is needed. Additionally, we invalidate the cache
-# if the Bazel version changes. We do this because otherwise the `bazelisk` cache folder will
-# contain all previously used versions and ultimately cause the cache restoring to be slower.
-var_4: &cache_fallback_key v16-ng-mat-{{ checksum ".bazelversion" }}-{{ checksum "tools/postinstall/apply-patches.js" }}-
-
-# Settings common to each job
-var_5: &job_defaults
-  working_directory: ~/ng
-  docker:
-    - image: *docker_image
-
-# Restores the cache that could be available for the current Yarn lock file. The cache usually
-# includes the node modules and the Bazel repository cache.
-var_7: &restore_cache
-  restore_cache:
-    keys:
-      - *cache_key
-      - *cache_fallback_key
-
-# Saves the cache for the current Yarn lock file. We store the node modules and the Bazel
-# repository cache in order to make subsequent builds faster.
-var_8: &save_cache
-  save_cache:
-    key: *cache_key
-    paths:
-      - 'node_modules'
-      - '~/.cache/bazelisk'
-      - '~/bazel_repository_cache'
-
-# Decryption token that is used to decode the GCP credentials file in ".circleci/gcp_token".
-var_9: &gcp_decrypt_token 'angular'
-
-# Job step that ensures that the node module dependencies are installed and up-to-date. We use
-# Yarn with the frozen lockfile option in order to make sure that lock file and package.json are
-# in sync. Unlike in Travis, we don't need to manually purge the node modules if stale because
-# CircleCI automatically discards the cache if the checksum of the lock file has changed.
-var_10: &yarn_install
-  run:
-    name: 'Installing project dependencies'
-    command: yarn install --frozen-lockfile --non-interactive
-
-# Installs all dependencies but does not enforce a frozen lockfile. Helpful when
-# the "package.json" is updated as part of a CI job. e.g. when setting up snapshots.
-var_11: &yarn_install_loose_lockfile
-  run:
-    name: 'Installing project dependencies'
-    command: yarn install --non-interactive
-
-# Sets up the Bazel config which is specific for CircleCI builds.
-var_12: &setup_bazel_ci_config
-  run:
-    name: 'Setting up Bazel configuration for CI'
-    # Note: We add the remote config flag to the user bazelrc file that is not tracked
-    # by Git. This is necessary to avoid stamping builds with `.with-local-changes`.
-    command: echo "import %workspace%/.circleci/bazel.rc" >> ./.bazelrc.user
-
-# Attaches the release output which has been stored in the workspace to the current job.
-# https://circleci.com/docs/2.0/workflows/#using-workspaces-to-share-data-among-jobs
-var_13: &attach_release_output
-  attach_workspace:
-    at: dist/
-
-# Branch filter that we can specify for jobs that should only run on publish branches. This filter
-# is used to ensure that not all upstream branches will be published as Github builds
-# (e.g. revert branches, feature branches)
-var_14: &publish_branches_filter
-  branches:
-    only:
-      - main
-      # 6.0.x, 7.1.x, etc.
-      - /\d+\.\d+\.x/
-
-# Branch filter that is usually applied to all jobs. Since there is no way within CircleCI to
-# exclude a branch for all defined jobs, we need to manually specify the filters for each job.
-# In order to reduce duplication we use a YAML anchor that just always excludes the "_presubmit"
-# branch. We don't want to run Circle for the temporary "_presubmit" branch which is reserved
-# for the caretaker.
-var_15: &ignore_presubmit_branch_filter
-  branches:
-    ignore:
-      - '_presubmit'
-
-# Runs a script that sets up the Bazel remote execution. This will be used by jobs that run
-# Bazel primarily and should benefit from remote caching and execution.
-var_16: &setup_bazel_remote_execution
-  devinfra/setup-bazel-remote-exec:
-    bazelrc: ./.bazelrc.user
-
-# Sets up the bazel binary globally. We don't want to access bazel through Yarn and NodeJS
-# because it could mean that the Bazel child process only has access to limited memory.
-var_17: &setup_bazel_binary
-  run:
-    name: 'Setting up global Bazel binary'
-    command: ./scripts/circleci/setup_bazel_binary.sh
-
-# Sets up the Angular snapshot builds.
-var_18: &setup_snapshot_builds
-  run:
-    name: 'Setting up Angular snapshot builds'
-    command: node ./scripts/circleci/setup-angular-snapshots.js main
-
-# Filter which ensures that jobs only run for pull requests.
-var_19: &only_on_pull_requests_filter
-  branches:
-    only:
-      - /pull\/\d+/
-
-# Anchor for a step that notifies Slack when preceding steps failed.
-var_20: &slack_notify_on_failure
-  run:
-    name: 'Notifying team about job failure'
-    when: on_fail
-    command: yarn ci-notify-slack-failure
-
-# Branch filter that only matches the main branch.
-var_21: &only_main_branch_filter
-  branches:
-    only:
-      - main
-
-# -----------------------------
-# Container version of CircleCI
-# -----------------------------
 version: 2.1
 
-# Configures CircleCI orbs for this pipeline. Orbs allow consumption of publicly shared
-# CircleCI commands, jobs, configuration elements or executors. Read more about orbs here:
-# https://circleci.com/docs/2.0/orb-intro/
-orbs:
-  # Set up the `queue` orb that allows us to queue up builds across jobs. We use it
-  # to make sure that snapshot builds are not deployed out of order, resulting in Git
-  # push conflicts.
-  queue: eddiewebb/queue@1.5.0
-  devinfra: angular/dev-infra@1.0.8
-
-commands:
-  checkout_and_rebase:
-    description: Checkout and rebase the repository
-    steps:
-      - checkout
-      # After checkout, rebase on top of target branch.
-      - devinfra/rebase-pr-on-target-branch:
-          base_revision: << pipeline.git.base_revision >>
-          head_revision: << pipeline.git.revision >>
-
-# -----------------------------------------------------------------------------------------
-# Job definitions. Jobs which are defined just here, will not run automatically. Each job
-# must be part of a workflow definition in order to run for PRs and push builds.
-# -----------------------------------------------------------------------------------------
 jobs:
-  # -------------------------------------------------------------------------------------------
-  # Job that builds all release packages. The built packages can be then used in the same
-  # workflow to publish snapshot builds.
-  # -------------------------------------------------------------------------------------------
-  build_release_packages:
-    <<: *job_defaults
-    resource_class: xlarge
-    environment:
-      GCP_DECRYPT_TOKEN: *gcp_decrypt_token
-    steps:
-      - checkout_and_rebase
-      - *restore_cache
-      - *setup_bazel_ci_config
-      - *setup_bazel_remote_execution
-      - *yarn_install
-      - *setup_bazel_binary
-
-      - run: yarn build-and-check-release-output
-      - run: yarn check-tooling-setup
-
-      # TODO(devversion): replace this with bazel tests that run Madge. This is
-      # cumbersome and doesn't guarantee no circular deps for other entry-points.
-      - run: yarn madge --circular dist/releases/cdk/schematics/index.js
-
-      # Store the release output in the workspace storage. This means that other jobs
-      # in the same workflow can attach the release output to their job.
-      - persist_to_workspace:
-          root: dist
-          paths:
-            - 'releases/**/*'
-      - *slack_notify_on_failure
-
-  # ----------------------------------------
-  # Job that publishes the docs site
-  # ----------------------------------------
-  deploy_docs_site:
+  pass:
     docker:
-      - image: *docker-browser-image
-    resource_class: xlarge
-    environment:
-      GCP_DECRYPT_TOKEN: *gcp_decrypt_token
+      - image: cimg/base:2022.05
     steps:
-      - checkout_and_rebase
-      - *restore_cache
-      - *setup_bazel_ci_config
-      - *setup_bazel_remote_execution
-      - *yarn_install
-      - *setup_bazel_binary
-
-      - run: yarn ci-push-deploy-docs-app
-      - *slack_notify_on_failure
+      - run: echo "This too shall pass (always)"
 
-# ----------------------------------------------------------------------------------------
-# Workflow definitions. A workflow usually groups multiple jobs together. This is useful if
-# one job depends on another.
-#
-# NOTE: When updating this configuration section, make sure to update GitHub robot
-#       config to match the new workflow jobs.
-# ----------------------------------------------------------------------------------------
 workflows:
   version: 2
-
   default_workflow:
     jobs:
-      - build_release_packages:
-          filters: *ignore_presubmit_branch_filter
-      - deploy_docs_site:
-          filters: *publish_branches_filter
-          requires:
-            - build_release_packages
+      - pass

.github/actions/slack/action.yml

@@ -0,0 +1,21 @@
+name: 'Notify slack on failure'
+description: 'Notifying Slack of failure'
+
+inputs:
+  JOB_NAME:
+    description: 'Name of the job'
+    required: true
+  ANGULAR_ROBOT_SLACK_TOKEN:
+    description: 'Name of the job'
+    required: true
+
+runs:
+  using: 'composite'
+  steps:
+    - name: Notify about failed test
+      uses: slackapi/slack-github-action@e28cf165c92ffef168d23c5c9000cffc8a25e117 # v=v1.24.0
+      with:
+        channel-id: 'C015EBF2XB6'
+        slack-message: '${{ inputs.JOB_NAME }} job failed for ${{ github.base_ref }} branch failed on build ${{ github.event.after }}: ${{ github.server_url }}/${{ github.repository }}/actions/runs/${{ github.run_id }}'
+      env:
+        SLACK_BOT_TOKEN: ${{ inputs.ANGULAR_ROBOT_SLACK_TOKEN }}

.github/angular-robot.yml

@@ -52,10 +52,6 @@ merge:
       - 'pr: needs*'
       - 'presubmit failures'
 
-    # list of PR statuses that need to be successful
-    requiredStatuses:
-      - 'ci/circleci: build_release_packages'
-
   # the comment that will be added when the merge label is added despite failing checks, leave empty or set to false to disable
   # {{MERGE_LABEL}} will be replaced by the value of the mergeLabel option
   # {{PLACEHOLDER}} will be replaced by the list of failing checks