pr-check.yml

name: Test Changed Jupyter Notebooks

on:
  pull_request:
    branches:
      - main

jobs:
  test-notebooks:
    runs-on: ubuntu-latest
    
    steps:
    - name: Checkout repository
      uses: actions/checkout@v2

    - name: Find changed notebooks
      id: find_notebooks
      run: |
        git fetch origin ${{ github.event.pull_request.base.ref }} ${{ github.event.pull_request.head.ref }}
        git diff --name-only origin/${{ github.event.pull_request.base.ref }}...origin/${{ github.event.pull_request.head.ref }} > changed_files.txt
        grep '\.ipynb$' changed_files.txt > changed_notebooks.txt || echo "No notebooks changed" > changed_notebooks.txt

    - name: Check if any notebooks changed
      id: check_notebooks
      run: |
        if grep -q '\.ipynb$' changed_notebooks.txt; then
          echo "notebooks_changed=true" >> GITHUB_ENV
        else
          echo "notebooks_changed=false" >> GITHUB_ENV
        fi
  
    - name: No notebooks changed
      if: env.notebooks_changed == 'false'
      run: |
        echo "No Jupyter notebooks changed in this PR."
        exit 0

    - name: Set up Python
      uses: actions/setup-python@v2
      with:
        python-version: '3.10'  # Use Python 3.10
  
    - name: Install flake8
      run: |
        python -m pip install --upgrade pip
        pip install flake8 flake8-nb
      shell: bash
  
    - name: Run Flake8 linting on changed notebooks
      if: env.notebooks_changed == 'true'
      id: lint
      run: |
        mkdir -p linting_logs
        flake8_errors=0
        while IFS= read -r notebook; do
          [ "$(basename "$notebook")" = "python-dosearch.ipynb" ] && continue
          echo "Linting $notebook"
          flake8-nb --config=.flake8 "$notebook" > "linting_logs/$(basename "$notebook" .ipynb)_linting.txt" 2>&1 || flake8_errors=$((flake8_errors+1))
          cat "linting_logs/$(basename "$notebook" .ipynb)_linting.txt"
        done < changed_notebooks.txt
        echo "flake8_errors=$flake8_errors" >> $GITHUB_ENV
      shell: bash
    
    - name: Upload linting logs
      uses: actions/upload-artifact@v2
      with:
        name: linting-logs
        path: linting_logs

    - name: Check for linting errors
      if: "(env.flake8_errors != '0')"
      run: exit 1
  
    - name: Install git
      run: |
        sudo apt-get update
        sudo apt-get install -y git
      shell: bash

    - name: Install libgraphviz-dev if folder is CM3
      run: sudo apt-get install -y libgraphviz-dev
      shell: bash

    - name: Install R if folder is CM3
      run: |
        sudo apt-get update
        sudo apt-get install -y r-base
      shell: bash

    - name: Install dependencies
      run: |
        python -m pip install --upgrade pip
        if [ -f requirements.txt ]; then pip install -r requirements.txt; fi
        pip install jupyter nbconvert ipython
      shell: bash

    - name: Convert changed Jupyter notebooks to Python scripts
      id: convert
      run: |
        mkdir -p converted_scripts
        while IFS= read -r notebook; do
          echo "Processing $notebook"
          [ -e "$notebook" ] || continue
          [ "$(basename "$notebook")" = "DoubleML_and_Feature_Engineering_with_BERT.ipynb" ] && continue
          jupyter nbconvert --to script "$notebook" --output-dir converted_scripts
        done < changed_notebooks.txt
      shell: bash

    - name: Run converted Python scripts with IPython
      id: execute
      run: |
        mkdir -p logs
        script_errors=0
        for script in converted_scripts/*.py; do
          [ -e "$script" ] || continue
          echo "Running $script"
          ipython "$script" > "logs/$(basename "$script" .py).txt" 2>&1 || script_errors=$((script_errors+1))
        done
        echo "script_errors=$script_errors" >> $GITHUB_ENV
      shell: bash

    - name: Aggregate and report errors and warnings
      run: |
        echo "Aggregating errors and warnings..."
        grep -E "Traceback|Error:|Exception:|ModuleNotFoundError|FutureWarning|TypeError" logs/*.txt linting_logs/*.txt > logs/errors_and_warnings.txt || true
        echo "Errors and Warnings:"
        cat logs/errors_and_warnings.txt
      shell: bash

    - name: Upload execution logs
      uses: actions/upload-artifact@v2
      with:
        name: execution-logs
        path: logs

    - name: Check for errors
      if: "(env.flake8_errors != '0' || env.script_errors != '0')"
      run: exit 1