pythonbuild.yml

name: Build

# Schedule runs to run twice a day

on:
  push:
    branches:
      - master
  pull_request:
  schedule:
    - cron: "0 13 * * *" # This schedule runs at 1pm UTC every day

env:
  FLYTE_SDK_LOGGING_LEVEL: 10 # debug

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

jobs:
  detect-python-versions:
    runs-on: ubuntu-latest
    outputs:
      python-versions: ${{ env.python_versions }}
    steps:
      - name: Set Python versions for run
        run: |
          if [[ ${{ github.event_name }} == "schedule" ]]; then
            echo "python_versions=[\"3.8\",\"3.9\",\"3.10\",\"3.11\",\"3.12\"]" >> $GITHUB_ENV
          else
            echo "python_versions=[\"3.9\", \"3.12\"]" >> $GITHUB_ENV
          fi

  build:
    needs:
      - detect-python-versions
    runs-on: ${{ matrix.os }}
    strategy:
      fail-fast: false
      matrix:
        os: [ubuntu-latest, windows-latest, macos-latest]
        python-version: ${{fromJson(needs.detect-python-versions.outputs.python-versions)}}
    steps:
      - uses: actions/checkout@v4
      - name: "Clear action cache"
        uses: ./.github/actions/clear-action-cache
      - name: Set up Python ${{ matrix.python-version }}
        uses: actions/setup-python@v4
        with:
          python-version: ${{ matrix.python-version }}
      - name: Cache pip
        uses: actions/cache@v3
        with:
          # This path is specific to Ubuntu
          path: ~/.cache/pip
          # Look to see if there is a cache hit for the corresponding requirements files
          key: ${{ format('{0}-pip-{1}', runner.os, hashFiles('dev-requirements.in', 'requirements.in')) }}
      - name: Install dependencies
        run: |
          pip install uv
          make setup-global-uv
          uv pip uninstall --system pandas pyarrow
          uv pip freeze
      - name: Test with coverage
        run: |
          make unit_test_codecov
      - name: Codecov
        uses: codecov/codecov-action@v3.1.4
        with:
          fail_ci_if_error: false
          files: coverage.xml

  build-with-extras:
    needs:
      - detect-python-versions
    runs-on: ${{ matrix.os }}
    strategy:
      fail-fast: false
      matrix:
        os: [ubuntu-latest, windows-latest, macos-latest]
        python-version: ${{fromJson(needs.detect-python-versions.outputs.python-versions)}}
    steps:
      - uses: actions/checkout@v4
      - name: "Clear action cache"
        uses: ./.github/actions/clear-action-cache
      - name: Set up Python ${{ matrix.python-version }}
        uses: actions/setup-python@v4
        with:
          python-version: ${{ matrix.python-version }}
      - name: Cache pip
        uses: actions/cache@v3
        with:
          # This path is specific to Ubuntu
          path: ~/.cache/pip
          # Look to see if there is a cache hit for the corresponding requirements files
          key: ${{ format('{0}-pip-{1}', runner.os, hashFiles('dev-requirements.in', 'requirements.in')) }}
      - name: Install dependencies
        run: |
          pip install uv
          make setup-global-uv
          uv pip uninstall --system pandas pyarrow
          uv pip freeze
      - name: Run extras unit tests with coverage
        # Skip this step if running on python 3.12 due to https://github.com/tensorflow/tensorflow/issues/62003
        # and https://github.com/pytorch/pytorch/issues/110436
        if: ${{ matrix.python-version != '3.12' }}
        run: |
          make unit_test_extras_codecov
      - name: Codecov
        uses: codecov/codecov-action@v3.1.4
        with:
          fail_ci_if_error: false
          files: coverage.xml

  build-with-pandas:
    needs:
      - detect-python-versions
    runs-on: ${{ matrix.os }}
    strategy:
      fail-fast: false
      matrix:
        os: [ubuntu-latest]
        python-version: ${{fromJson(needs.detect-python-versions.outputs.python-versions)}}
        pandas: ["pandas<2.0.0", "pandas>=2.0.0"]
        numpy: ["numpy<2.0.0", "numpy>=2.0.0"]
        exclude:
          - numpy: "numpy>=2.0.0"
            pandas: "pandas<2.0.0"
          - numpy: "numpy<2.0.0"
            pandas: "pandas>=2.0.0"
          - numpy: "numpy>=2.0.0"
            python-version: "3.8"

    steps:
      - uses: actions/checkout@v4
      - name: "Clear action cache"
        uses: ./.github/actions/clear-action-cache
      - name: Set up Python ${{ matrix.python-version }}
        uses: actions/setup-python@v4
        with:
          python-version: ${{ matrix.python-version }}
      - name: Cache pip
        uses: actions/cache@v3
        with:
          # This path is specific to Ubuntu
          path: ~/.cache/pip
          # Look to see if there is a cache hit for the corresponding requirements files
          key: ${{ format('{0}-pip-{1}', runner.os, hashFiles('dev-requirements.in', 'requirements.in')) }}
      - name: Install dependencies
        run: |
          pip install uv
          make setup-global-uv
          uv pip install --system --force-reinstall "${{ matrix.pandas }}" "${{ matrix.numpy }}"
          uv pip freeze
      - name: Test with coverage
        run: |
          make unit_test_codecov
      - name: Codecov
        uses: codecov/codecov-action@v3.1.4
        with:
          fail_ci_if_error: false
          files: coverage.xml

  test-hypothesis:
    needs:
      - detect-python-versions
    runs-on: ${{ matrix.os }}
    strategy:
      fail-fast: false
      matrix:
        os: [ubuntu-latest]
        python-version: ${{fromJson(needs.detect-python-versions.outputs.python-versions)}}
    steps:
      - uses: actions/checkout@v4
      - name: Set up Python ${{ matrix.python-version }}
        uses: actions/setup-python@v4
        with:
          python-version: ${{ matrix.python-version }}
      - name: Cache pip
        uses: actions/cache@v3
        with:
          # This path is specific to Ubuntu
          path: ~/.cache/pip
          # Look to see if there is a cache hit for the corresponding requirements files
          key: ${{ format('{0}-pip-{1}', runner.os, hashFiles('dev-requirements.in', 'requirements.in')) }}
      - name: Install dependencies
        run: |
          pip install uv
          make setup-global-uv
          uv pip freeze
      - name: Test with coverage
        env:
          FLYTEKIT_HYPOTHESIS_PROFILE: ci
        run: |
          make unit_test_hypothesis
      - name: Codecov
        uses: codecov/codecov-action@v3.1.4
        with:
          fail_ci_if_error: false
          files: coverage.xml

  test-serialization:
    needs:
      - detect-python-versions
    runs-on: ${{ matrix.os }}
    strategy:
      fail-fast: false
      matrix:
        os: [ubuntu-latest]
        python-version: ${{fromJson(needs.detect-python-versions.outputs.python-versions)}}
    steps:
      - uses: actions/checkout@v4
      - name: Set up Python ${{ matrix.python-version }}
        uses: actions/setup-python@v4
        with:
          python-version: ${{ matrix.python-version }}
      - name: Cache pip
        uses: actions/cache@v3
        with:
          # This path is specific to Ubuntu
          path: ~/.cache/pip
          # Look to see if there is a cache hit for the corresponding requirements files
          key: ${{ format('{0}-pip-{1}', runner.os, hashFiles('dev-requirements.in', 'requirements.in')) }}
      - name: Install dependencies
        run: |
          pip install uv
          make setup-global-uv
          uv pip freeze
      - name: Test with coverage
        run: |
          make test_serialization_codecov
      - name: Codecov
        uses: codecov/codecov-action@v3.1.4
        with:
          fail_ci_if_error: false
          files: coverage.xml

  integration:
    needs:
      - detect-python-versions
    runs-on: ${{ matrix.os }}
    strategy:
      fail-fast: false
      matrix:
        os: [ubuntu-latest]
        python-version: ${{fromJson(needs.detect-python-versions.outputs.python-versions)}}
        makefile-cmd:
          [integration_test_codecov, integration_test_lftransfers_codecov]
    steps:
      # As described in https://github.com/pypa/setuptools_scm/issues/414, SCM needs git history
      # and tags to work.
      - uses: actions/checkout@v4
        with:
          fetch-depth: 0
      - name: "Clear action cache"
        uses: ./.github/actions/clear-action-cache # sandbox has disk pressure, so we need to clear the cache to get more disk space.
      - name: Set up Python ${{ matrix.python-version }}
        uses: actions/setup-python@v4
        with:
          python-version: ${{ matrix.python-version }}
      - name: Cache pip
        uses: actions/cache@v3
        with:
          # This path is specific to Ubuntu
          path: ~/.cache/pip
          # Look to see if there is a cache hit for the corresponding requirements files
          key: ${{ format('{0}-pip-{1}', runner.os, hashFiles('dev-requirements.in', 'requirements.in')) }}
      - name: Install dependencies
        run: |
          pip install uv
          make setup-global-uv
          uv pip freeze
      - name: Install FlyteCTL
        uses: unionai-oss/flytectl-setup-action@master
      - name: Setup Flyte Sandbox
        run: |
          flytectl demo start
      - name: Set up Docker Buildx
        uses: docker/setup-buildx-action@v3
        with:
          driver-opts: |
            image=moby/buildkit:master
            network=host
      - name: Build and push to local registry
        uses: docker/build-push-action@v5
        with:
          context: .
          file: Dockerfile.dev
          build-args: |
            PYTHON_VERSION=${{ matrix.python-version }}
            PSEUDO_VERSION=1.999.0dev0
          push: true
          tags: localhost:30000/flytekit:dev
          cache-from: type=gha
          cache-to: type=gha,mode=max
      - name: Integration Test with coverage
        env:
          FLYTEKIT_IMAGE: localhost:30000/flytekit:dev
          FLYTEKIT_CI: 1
          PYTEST_OPTS: -n2
        run: |
          make ${{ matrix.makefile-cmd }}
      - name: Codecov
        uses: codecov/codecov-action@v3.1.0
        with:
          fail_ci_if_error: false

  build-plugins:
    needs:
      - build
      - detect-python-versions
    runs-on: ubuntu-latest
    strategy:
      fail-fast: false
      matrix:
        python-version: ${{fromJson(needs.detect-python-versions.outputs.python-versions)}}
        plugin-names:
          # Please maintain an alphabetical order in the following list
          - flytekit-airflow
          - flytekit-async-fsspec
          - flytekit-aws-athena
          - flytekit-aws-batch
          - flytekit-aws-sagemaker
          - flytekit-bigquery
          - flytekit-comet-ml
          - flytekit-dask
          - flytekit-data-fsspec
          - flytekit-dbt
          - flytekit-deck-standard
          # TODO: remove dolt plugin - https://github.com/flyteorg/flyte/issues/5350
          # flytekit-dolt
          - flytekit-duckdb
          - flytekit-envd
          - flytekit-flyteinteractive
          - flytekit-greatexpectations
          - flytekit-hive
          - flytekit-huggingface
          - flytekit-identity-aware-proxy
          - flytekit-inference
          - flytekit-k8s-pod
          - flytekit-kf-mpi
          - flytekit-kf-pytorch
          - flytekit-kf-tensorflow
          - flytekit-mlflow
          - flytekit-mmcloud
          - flytekit-modin
          - flytekit-neptune
          - flytekit-onnx-pytorch
          - flytekit-onnx-scikitlearn
          # onnx-tensorflow needs a version of tensorflow that does not work with protobuf>4.
          # The issue is being tracked on the tensorflow side in https://github.com/tensorflow/tensorflow/issues/53234#issuecomment-1330111693
          # flytekit-onnx-tensorflow
          - flytekit-omegaconf
          - flytekit-openai
          - flytekit-pandera
          - flytekit-papermill
          - flytekit-polars
          - flytekit-pydantic
          - flytekit-ray
          - flytekit-snowflake
          - flytekit-spark
          - flytekit-sqlalchemy
          - flytekit-vaex
          - flytekit-whylogs
        exclude:
          - python-version: 3.8
            plugin-names: "flytekit-aws-sagemaker"
          - python-version: 3.9
            plugin-names: "flytekit-aws-sagemaker"
          # flytekit-modin depends on ray which does not have a 3.11 wheel yet.
          # Issue tracked in https://github.com/ray-project/ray/issues/27881
          - python-version: 3.11
            plugin-names: "flytekit-modin"
          - python-version: 3.11
            plugin-names: "flytekit-ray"
          # Great-expectations does not support python 3.11 due to sqlachemy>=2.0.0
          # not being supported yet:
          # https://github.com/great-expectations/great_expectations/issues/7020
          - python-version: 3.11
            plugin-names: "flytekit-greatexpectations"
          # onnxruntime does not support python 3.10 yet
          # https://github.com/microsoft/onnxruntime/issues/9782
          - python-version: 3.11
            plugin-names: "flytekit-onnx-pytorch"
          - python-version: 3.11
            plugin-names: "flytekit-onnx-scikitlearn"
          - python-version: 3.11
            plugin-names: "flytekit-onnx-tensorflow"
          # vaex currently doesn't support python 3.11
          - python-version: 3.11
            plugin-names: "flytekit-vaex"
          # whylogs does support python 3.11 dataclass restrictions
          # See: https://github.com/flyteorg/flytekit/actions/runs/4493746408/jobs/7905368664
          - python-version: 3.11
            plugin-names: "flytekit-whylogs"
          # apache-beam, one of flytekit-airflow dependencies, does not support python 3.12: https://github.com/apache/beam/issues/29149
          - python-version: 3.12
            plugin-names: "flytekit-airflow"
          # Tensorflow is a dependency of flytekit-mlflow tests and that is not supported yet: https://github.com/tensorflow/tensorflow/issues/62003
          - python-version: 3.12
            plugin-names: "flytekit-mlflow"
          # modin[ray] is a dependency needed to run the tests, unfortunately this is not supported in python 3.12 yet
          - python-version: 3.12
            plugin-names: "flytekit-modin"
          # vaex currently doesn't support python 3.12
          - python-version: 3.12
            plugin-names: "flytekit-vaex"
          # Ray does not support python 3.12 yet: https://github.com/ray-project/ray/issues/40211
          - python-version: 3.12
            plugin-names: "flytekit-ray"
          # Segmentation fault on python 3.12: https://github.com/flyteorg/flyte/issues/5020
          - python-version: 3.12
            plugin-names: "flytekit-kf-pytorch"
    steps:
      - uses: actions/checkout@v4
      - name: "Clear action cache"
        uses: ./.github/actions/clear-action-cache
      - name: Set up Python ${{ matrix.python-version }}
        uses: actions/setup-python@v4
        with:
          python-version: ${{ matrix.python-version }}
      - name: Cache pip
        uses: actions/cache@v3
        with:
          # This path is specific to Ubuntu
          path: ~/.cache/pip
          # Look to see if there is a cache hit for the corresponding requirements files
          key: ${{ format('{0}-pip-{1}', runner.os, hashFiles('dev-requirements.txt', format('plugins/{0}/requirements.txt', matrix.plugin-names ))) }}
      - name: Install dependencies
        run: |
          pip install uv
          # TODO: double-check if checking out all tags solves the issue
          export SETUPTOOLS_SCM_PRETEND_VERSION="3.0.0"
          make setup-global-uv
          cd plugins/${{ matrix.plugin-names }}
          uv pip install --system .
          if [ -f dev-requirements.in ]; then uv pip install --system -r dev-requirements.in; fi
          # TODO: move to protobuf>=5. Github issue: https://github.com/flyteorg/flyte/issues/5448
          uv pip install --system -U $GITHUB_WORKSPACE "protobuf<5" "git+https://github.com/flyteorg/flyte.git@master#subdirectory=flyteidl"
          # TODO: remove this when numpy v2 in onnx has been resolved
          if [[ ${{ matrix.plugin-names }} == *"onnx"* || ${{ matrix.plugin-names }} == "flytekit-sqlalchemy" || ${{ matrix.plugin-names }} == "flytekit-pandera" ]]; then
            uv pip install --system "numpy<2.0.0"
          fi
          uv pip freeze
      - name: Test with coverage
        run: |
          cd plugins/${{ matrix.plugin-names }}
          # onnx plugins does not support protobuf>4 yet (in fact it is tensorflow that
          # does not support that yet). More details in https://github.com/onnx/onnx/issues/4239.
          if [[ ${{ matrix.plugin-names }} == *"onnx"* || ${{ matrix.plugin-names }} == "flytekit-whylogs" || ${{ matrix.plugin-names }} == "flytekit-mlflow" ]]; then
             PROTOCOL_BUFFERS_PYTHON_IMPLEMENTATION=python coverage run -m pytest tests --cov=./ --cov-report=xml --cov-append
          else
             coverage run -m pytest tests --cov=./ --cov-report=xml --cov-append
          fi
      - name: Codecov
        uses: codecov/codecov-action@v3.1.0
        with:
          fail_ci_if_error: false
  lint:
    runs-on: ubuntu-latest
    steps:
      - name: Fetch the code
        uses: actions/checkout@v4
      - name: Set up Python 3.12
        uses: actions/setup-python@v4
        with:
          python-version: 3.12
      - uses: actions/cache@v3
        with:
          path: ~/.cache/pip
          key: ${{ runner.os }}-pip-${{ hashFiles('**/dev-requirements.in') }}
          restore-keys: |
            ${{ runner.os }}-pip-
      - name: Install dependencies
        run: |
          python -m pip install --upgrade pip
          pip install uv
          make setup-global-uv
          uv pip freeze
      - name: Lint
        run: |
          make lint