ci-lint-validate-convert.yml

name: Lint, validate and convert CSV to JSON-LD

on:
  pull_request:
    branches: 'main'
    paths: 'HTAN.model.csv'
  workflow_dispatch:

concurrency:
# cancel the current running workflow from the same branch, PR when a new workflow is triggered 
# when the trigger is not a PR but a push, it will use the commit sha to generate the concurrency group
# {{ github.workflow }}: the workflow name is used to generate the concurrency group. This allows you to have more than one workflows
# {{ github.ref_type }}: the type of Git ref object created in the repository. Can be either branch or tag 
# {{ github.sha }}: full commit sha 
# credit: https://github.com/Sage-Bionetworks-Workflows/sagetasks/blob/main/.github/workflows/ci.yml
  group: >-
    ${{ github.workflow }}-${{ github.ref_type }}-
    ${{ github.sha }}
  cancel-in-progress: true

jobs:
  lint:
    name: Lint CSV
    runs-on: ubuntu-latest
    steps:
    - uses: actions/checkout@v3
    - name: csvlinter
      uses: kcheriyath/csvlinter@V0.6.0
      with:
        find_pattern: "*.csv"

  validate:
    name: Validate CSV
    runs-on: ubuntu-latest
    env:
      GITHUB_PAT: ${{ secrets.GITHUB_TOKEN }}

    steps:
      - uses: actions/checkout@v3
      - uses: actions/setup-python@v4 
        with:
          python-version: '3.10'

      - name: Install Schematic
        shell: bash
        run: |
          pip3 install poetry
          git clone --single-branch --branch main https://github.com/Sage-Bionetworks/schematic.git
          cd schematic
          poetry build
          pip3 install dist/schematicpy-*-py3-none-any.whl
          pip install --force-reinstall typing-extensions==4.5.0
      
      - name: Convert CSV schema
        shell: bash
        run: |
          schematic schema convert .github/CSV.model.csv   

      - name: Validate data model CSV
        shell: bash
        run: |
          schematic model -c .github/CSV_schematic_config.yml validate -mp HTAN.model.csv -dt "DataModel" | 
            grep "Your manifest has been validated successfully. There are no errors in your manifest, and it can be submitted without any modifications."

  component_check:
    name: Check Component DependsOn entries
    runs-on: ubuntu-latest
    env:
      GITHUB_PAT: ${{ secrets.GITHUB_TOKEN }}

    steps:
      - uses: actions/checkout@v3
      - uses: actions/setup-python@v4 
        with:
          python-version: '3.10'

      - name: Install Python dependencies
        run: python -m pip install --upgrade pip pandas
        
      - name: Run script
        shell: bash
        run: |
          python .github/workflows/check_components.py

  convert:
    name: Convert CSV to JSON-LD
    needs:
      - lint
      - validate
      - component_check
    runs-on: ubuntu-latest
    env:
      GITHUB_PAT: ${{ secrets.GITHUB_TOKEN }}

    steps:
      - uses: actions/checkout@v3
      - uses: actions/setup-python@v4 
        with:
          python-version: '3.9'

      - name: Install Schematic
        shell: bash
        run: |
          pip3 install poetry
          git clone --single-branch --branch main https://github.com/Sage-Bionetworks/schematic.git
          cd schematic
          poetry build
          pip3 install dist/schematicpy-*-py3-none-any.whl

      - name: Convert .csv to .jsonld
        shell: bash
        run: |
          schematic schema convert HTAN.model.csv

      - uses: r-lib/actions/pr-fetch@v2
        with:
          repo-token: ${{ secrets.GITHUB_TOKEN }}

      - name: Commit the changes
        run: |
          git config --local user.name "$GITHUB_ACTOR"
          git config --local user.email "$GITHUB_ACTOR@users.noreply.github.com"
          git add HTAN.model.jsonld
          git commit -m "GitHub Action: convert *.model.csv to *.model.jsonld" || echo "No changes to commit"

      - uses: r-lib/actions/pr-push@v2
        with:
          repo-token: ${{ secrets.GITHUB_TOKEN }}