analysis_workflow.yml

name: Build with analysis tools
on:
  workflow_dispatch:
    inputs:
      run_all_benchmarks:
        type: boolean
      force_publish_result:
        description: Force publish result (requries admin access)
        type: boolean

  schedule: # Schdeule the job to run at 12 a.m. daily
    - cron: '0 0 * * *'

  pull_request:
    paths-ignore:
    - "**/*.md"

jobs:
  cibw_docker_image:
    uses: ./.github/workflows/cibw_docker_image.yml
    permissions: {packages: write}
    with:
        cibuildwheel_ver: "2.12.1"
        force_update: false

  code_coverage:
    needs: [cibw_docker_image]  
    runs-on: "ubuntu-latest"
    container:
        image: ${{needs.cibw_docker_image.outputs.tag}}
    services:
      mongodb:
        image: mongo:4.4
        ports:
          - 27017:27017
    env:
      VCPKG_NUGET_USER: ${{secrets.VCPKG_NUGET_USER || github.repository_owner}}
      VCPKG_NUGET_TOKEN: ${{secrets.VCPKG_NUGET_TOKEN || secrets.GITHUB_TOKEN}}
      VCPKG_MAN_NUGET_USER: ${{secrets.VCPKG_MAN_NUGET_USER}} # For forks to download pre-compiled dependencies from the Man repo
      VCPKG_MAN_NUGET_TOKEN: ${{secrets.VCPKG_MAN_NUGET_TOKEN}}
      ARCTIC_CMAKE_PRESET: linux-debug
    steps:
      - uses: actions/checkout@v3.3.0
        with:
          submodules: recursive

      - name: Get number of CPU cores
        uses: SimenB/github-actions-cpu-cores@v1.1.0
        id: cpu-cores

      - name: Extra envs
        run: |
          . build_tooling/vcpkg_caching.sh # Linux follower needs another call in CIBW
          echo -e "VCPKG_BINARY_SOURCES=$VCPKG_BINARY_SOURCES
          VCPKG_ROOT=$PLATFORM_VCPKG_ROOT" | tee -a $GITHUB_ENV
          cmake -P cpp/CMake/CpuCount.cmake | sed 's/^-- //' | tee -a $GITHUB_ENV
          echo "ARCTICDB_CODE_COVERAGE_BUILD=1" | tee -a $GITHUB_ENV
        env:
          CMAKE_BUILD_PARALLEL_LEVEL: ${{vars.CMAKE_BUILD_PARALLEL_LEVEL}}

      - name: Prepare C++ compilation env
        run: . build_tooling/prep_cpp_build.sh 
      
      - name: CMake compile
        # We are pinning the version to 10.6 because >= 10.7, use node20 which is not supported in the container
        uses: lukka/run-cmake@v10.6
        with:
          cmakeListsTxtPath: ${{github.workspace}}/cpp/CMakeLists.txt
          configurePreset: ${{env.ARCTIC_CMAKE_PRESET}}
          buildPreset: ${{env.ARCTIC_CMAKE_PRESET}}
        env:
          ARCTICDB_DEBUG_FIND_PYTHON: ${{vars.ARCTICDB_DEBUG_FIND_PYTHON}}
          python_impl_name: 'cp311'
        
      - name: Run C++ Tests
        shell: bash -l {0}
        run: |
          cd cpp/out/linux-debug-build/
          ls arcticdb
          make -j ${{ steps.cpu-cores.outputs.count }} arcticdb_rapidcheck_tests
          make -j ${{ steps.cpu-cores.outputs.count }} test_unit_arcticdb
          ctest

      # We are chainging the python here because we want to use the default python to build (it is devel version)
      # and this python for the rest of the testing
      - name: Select Python (Linux)
        run: echo /opt/python/cp36-cp36m/bin >> $GITHUB_PATH

      - name: Install local dependencies with pip
        shell: bash
        run: |
          python -m pip install --upgrade pip
          ARCTIC_CMAKE_PRESET=skip pip install -ve .[Testing]

      # - name: Test with pytest
      #   uses: ./.github/actions/run_local_pytest
      #   with:
      #     build_type: debug
      #     threads: 1
      #     fast_tests_only: 0
      #     other_params: '-m coverage run '

      - name: Get python Coverage report
        shell: bash -l {0}
        run: |
          cd python
          python -m coverage report -m | tee output.txt
          python -m coverage html
          zip -r python_cov.zip htmlcov/

          echo "PYTHON_COV_PERCENT=$(cat output.txt | grep 'TOTAL' | awk '{print $NF}' | tr -d '%')" >> $GITHUB_ENV

      - name: Run Gcovr manually post-pytest
        shell: bash -l {0}
        run: |
          cd cpp/out/linux-debug-build/
          python -m pip install gcovr
          mkdir coverage
          python -m gcovr --txt --html-details coverage/index.html -e vcpkg_installed/ -e proto/ -e ../../third_party -e ../../arcticdb/util/test/ -r ../.. --exclude-throw-branches --exclude-unreachable-branches -u --exclude-function-lines | tee output.txt
          zip -r coverage.zip coverage/

          echo "CPP_COV_PERCENT=$(cat output.txt | grep 'TOTAL' | awk '{print $NF}' | tr -d '%')" >> $GITHUB_ENV

      - name: Upload Coverage
        uses: actions/upload-artifact@v3.1.3
        with:
          name: cpp-coverage-artifact
          path: cpp/out/linux-debug-build/coverage.zip

      - name: Upload Python Coverage
        uses: actions/upload-artifact@v3.1.3
        with:
          name: python-coverage-artifact
          path: python/python_cov.zip

      - name: Restore cached CPP Coverage Percentage from the previous run
        id: cache-cov-restore
        uses: actions/cache/restore@v3.3.2
        with:
          path: prev_coverage.txt
          key: coverage
      
      - name: Get and compare coverage if cache was restored
        run: |
          # if cache was restored, compare coverage
          if [ -f coverage.txt ]; then
            PREV_COVERAGE=$(cat prev_coverage.txt | cut -d' ' -f2)
            echo "Previous coverage: $PREV_COVERAGE"
            CURR_COVERAGE=${{env.CPP_COV_PERCENT}}
            echo "CPP_COV_PREV_PERCENT=$PREV_COVERAGE" >> $GITHUB_ENV
            echo "Current coverage: $CURR_COVERAGE"
            if [ $CURR_COVERAGE -gt $PREV_COVERAGE ]; then
              echo "Coverage increased"
            elif [ $CURR_COVERAGE -lt $PREV_COVERAGE ]; then
              echo "Coverage decreased"
            else
              echo "Coverage unchanged"
            fi
          fi

      - name: Save CPP Coverage Percentage to file
        run: |
          echo "Coverage: ${{ env.CPP_COV_PERCENT }}" > current_coverage.txt

      - name: Save the current CPP Coverage Percentage to the cache
        id: cache-cov-save
        uses: actions/cache/save@v3.3.2
        with:
          path: current_coverage.txt
          key: coverage

      - name: Check percentage and send Slack notification
        if: ${{ env.CPP_COV_PREV_PERCENT && env.CPP_COV_PERCENT && env.CPP_COV_PERCENT < env.CPP_COV_PREV_PERCENT }}
        uses: slackapi/slack-github-action@v1.24.0
        with:
          # For posting a rich message using Block Kit
          payload: |
            {
              "text": "The CPP Code Coverage has been reduced",
              "blocks": [
                {
                  "type": "section",
                  "text": {
                    "type": "mrkdwn",
                    "text": "The CPP Code Coverage from the current run(${{ env.CPP_COV_PERCENT }}%) is lower the previous one(${{ env.CPP_COV_PREV_PERCENT }}%)."
                  }
                }
              ]
            }
        env:
          SLACK_WEBHOOK_URL: ${{ secrets.ARCTICDB_DEV_WEBHOOK_URL }}
          SLACK_WEBHOOK_TYPE: INCOMING_WEBHOOK

  start_ec2_runner:
    if: github.event_name == 'schedule'
    uses: ./.github/workflows/ec2_runner_jobs.yml
    secrets: inherit
    with:
      job_type: start

  benchmark_linux_medium:
    timeout-minutes: 1200
    if: github.event_name == 'schedule'
    needs: [start_ec2_runner, cibw_docker_image]
    runs-on: ${{ needs.start_ec2_runner.outputs.label }}
    container:
      image: ${{needs.cibw_docker_image.outputs.tag}}
    permissions:
      contents: write # Technically not necessary since the ARCTICDB_TEST_PAT token override exists
    env:
      # this is potentially overflowing the cache, so should be looked into after we address issue #1057
      SCCACHE_GHA_VERSION: ${{vars.SCCACHE_GHA_VERSION || 1}} # Setting this env var enables the caching
      VCPKG_NUGET_USER: ${{secrets.VCPKG_NUGET_USER || github.repository_owner}}
      VCPKG_NUGET_TOKEN: ${{secrets.VCPKG_NUGET_TOKEN || secrets.GITHUB_TOKEN}}
      CMAKE_C_COMPILER_LAUNCHER: sccache
      CMAKE_CXX_COMPILER_LAUNCHER: sccache

    steps:
      - uses: actions/checkout@v3.3.0
        with:
          fetch-depth: 0
          token: ${{ secrets.ARCTICDB_TEST_PAT }}

      - name: Set persistent storage variables
        uses: ./.github/actions/set_persistent_storage_env_vars
        with:
          aws_access_key: "${{ secrets.AWS_S3_ACCESS_KEY }}"
          aws_secret_key: "${{ secrets.AWS_S3_SECRET_KEY }}"

      - name: Configure sccache
        uses: mozilla-actions/sccache-action@v0.0.3
        with:
          version: "v0.4.0"
        
      # We are changing the python here because we want to use the default python to build (it is devel version)
      # and this python for the rest of the testing
      - name: Select Python (Linux)
        run: |
          ls /opt/python  
          echo /opt/python/cp36-cp36m/bin >> $GITHUB_PATH

      - name: Install ASV
        shell: bash -el {0}
        run: |
          git config --global --add safe.directory .
          python -m pip install --upgrade pip
          pip install asv virtualenv
          python -m asv machine -v --yes --machine ArcticDB-Runner-Medium

      - name: Benchmark all commits
        if: inputs.run_all_benchmarks == true
        shell: bash -el {0}
        run: |
          python -m asv run -v --show-stderr HASHFILE:python/hashes_to_benchmark.txt

      - name: Benchmark only latest commit
        if: github.event_name != 'pull_request' && inputs.run_all_benchmarks == false
        shell: bash -el {0}
        run: |
          python -m asv run -v --show-stderr HEAD^!
          git log HEAD^..HEAD --oneline -n1 --decorate=no | awk '{print $1;}' >> python/hashes_to_benchmark.txt

      - name: Benchmark against master
        if: github.event_name == 'pull_request' && inputs.run_all_benchmarks == false
        shell: bash -el {0}
        run: |
          python -m asv continuous -v --show-stderr origin/master HEAD -f 1.15

      - name: Publish results
        if: inputs.force_publish_result == true || github.event_name == 'schedule'
        shell: bash -el {0}
        run: |
          git config --global user.name "${GITHUB_ACTOR}"
          git config --global user.email "${GITHUB_ACTOR_ID}+${GITHUB_ACTOR}@users.noreply.github.com"

          git fetch --tags --all

          python -m asv publish -v

          if ${{ github.event_name != 'pull_request' }} ; then
            git add python/.asv/results/*
            git add python/hashes_to_benchmark.txt
            git commit -m "Update ASV Results and hashes to test"
            git push origin HEAD
          fi

          if ${{ github.ref == 'refs/heads/master' }} ; then
            python -m asv gh-pages -v --rewrite
          fi

  stop-ec2-runner:
    needs: [start_ec2_runner, benchmark_linux_medium]
    if: github.event_name == 'schedule'
    uses: ./.github/workflows/ec2_runner_jobs.yml
    secrets: inherit
    with:
      job_type: stop
      label: ${{ needs.start_ec2_runner.outputs.label }}
      ec2-instance-id: ${{ needs.start_ec2_runner.outputs.ec2-instance-id }}