fix: add missing nop param for full-sync

[PendaGTP]

Related to #378 (#378 (comment))

Now the full-sync script will use the new env var FULL_SYNC_NOOP and so run is noop mode if true.
This helps run safe-settings GHA since you can add this var conditionaly based on event.

Example (based from https://github.com/UCL-MIRSG/.github/blob/4695e545829b91dcddc6e36358454bc4a879f751/.github/workflows/safe-settings.yaml thanks @paddyroddy) :

name: Safe Settings Sync
on:
  push:
    branches:
      - main
  pull_request:
    paths:
      - safe-settings/**
      - .github/workflows/safe-settings.yaml
  schedule:
    - cron: 0 */4 * * *
  # -->
  workflow_dispatch:
    inputs:
      NOOP:
        description: 'Run in no-op mode'
        required: false
        type: boolean
        default: false

concurrency:
  cancel-in-progress: true
  group: >-
    ${{ github.workflow }}-${{ github.event.pull_request.number || github.ref }}

jobs:
  safe-settings-sync:
    runs-on: ubuntu-latest
    env:
      SAFE_SETTINGS_VERSION: 2.1.14
      SAFE_SETTINGS_CODE_DIR: .safe-settings-code
    steps:
      - name: Checkout source
        uses: actions/checkout@11bd71901bbe5b1630ceea73d27597364c9af683 # v4

      - name: Checkout GitHub Safe-Settings repository
        uses: actions/checkout@11bd71901bbe5b1630ceea73d27597364c9af683 # v4
        with:
          path: ${{ env.SAFE_SETTINGS_CODE_DIR }}
          ref: ${{ env.SAFE_SETTINGS_VERSION }}
          repository: github/safe-settings

      - name: Setup Node.js
        uses: actions/setup-node@39370e3970a6d050c480ffad4ff0ed4d3fdee5af # v4
        with:
          cache-dependency-path:
            ${{ env.SAFE_SETTINGS_CODE_DIR }}/package-lock.json
          cache: npm
          node-version-file: ${{ env.SAFE_SETTINGS_CODE_DIR }}/.nvmrc

      - name: Install dependencies
        run: npm install
        working-directory: ${{ env.SAFE_SETTINGS_CODE_DIR }}

      # -->
      - name: Set no-op mode flag
        run: |
          echo "FULL_SYNC_NOOP=false" >> $GITHUB_ENV

          if [[ "${{ github.event_name }}" == "pull_request" ]]; then
            echo "FULL_SYNC_NOOP=true" >> $GITHUB_ENV
          fi

          if [[ "${{ github.event_name }}" == "workflow_dispatch" ]]; then
            if [[ "${{ inputs.NOOP }}" != "" ]]; then
              echo "FULL_SYNC_NOOP=${{ inputs.NOOP }}" >> $GITHUB_ENV
            fi
          fi

      - name: Run application
        run: npm run full-sync
        working-directory: ${{ env.SAFE_SETTINGS_CODE_DIR }}
        env:
          ADMIN_REPO: .github
          APP_ID: ${{ vars.SAFE_SETTINGS_APP_ID }}
          BLOCK_REPO_RENAME_BY_HUMAN: false
          CONFIG_PATH: safe-settings
          DEPLOYMENT_CONFIG_FILE:
            ${{ github.workspace }}/safe-settings/deployment.yaml
          ENABLE_PR_COMMENT: true
          GH_ORG: ${{ vars.SAFE_SETTINGS_GH_ORG }}
          GITHUB_CLIENT_ID: ${{ vars.SAFE_SETTINGS_GITHUB_CLIENT_ID }}
          GITHUB_CLIENT_SECRET:
            ${{ secrets.SAFE_SETTINGS_GITHUB_CLIENT_SECRET }}
          LOG_LEVEL: trace
          PRIVATE_KEY: ${{ secrets.SAFE_SETTINGS_PRIVATE_KEY }}
          SETTINGS_FILE_PATH: organisation.yaml
          # -->
          FULL_SYNC_NOOP: ${{ env.FULL_SYNC_NOOP }}

[HagegeR]

Nice! maybe you can update the readme too?
https://github.com/github/safe-settings/blob/main-enterprise/docs/github-action.md

[PendaGTP]

Nice! maybe you can update the readme too? https://github.com/github/safe-settings/blob/main-enterprise/docs/github-action.md

Yes, I agree that updating the documentation would be beneficial.

I think it would be more appropriate to create a separate PR for GHA documentation updates, as I'm planning to make other documentation improvements unrelated to this PR.

But if updating GHA documentation in this current PR is required to merge it, I can update it.

[PendaGTP]

@decyjphr

We have different implementations for boolean environment variables.
For example, CREATE_PR_COMMENT is handled as a string, while with this changes FULL_SYNC_NOP is implemented as a boolean value. This is reflected in the different approaches in our unit tests.

I think it would be beneficial to standardize how we handle boolean environment variables across the project. Once we decide on the preferred implementation (string or boolean), I can create a new PR to refactor the existing code for consistency.
I can also update my changes to fit current implementation.

[sofuture]

This would be very useful!

[klutchell]

I would love to see this implemented!

[decyjphr]

@PendaGTP I like the approach of treating env variables as boolean. I will create a PR for making the rest also as boolean.