Feat: Add metadata validation workflow and scripts

Signed-off-by: alyssacgoins <agoins@redhat.com>

Author: alyssacgoins

.github/resources/scripts/get_components_for_validation.sh

@@ -0,0 +1,91 @@
+#!/usr/bin/env bash
+
+# Determine which components should be validated by a given validation script.
+#
+# Usage:
+#   get_components_for_validation.sh <BASE_COMMIT> <HEAD_COMMIT> <VALIDATION_SCRIPT_DIR>
+#
+# Outputs a GitHub Actions output variable named `new_components_list` containing
+# a comma-separated list of component directories to validate.
+
+set -euo pipefail
+
+BASE_COMMIT=${1:-}
+HEAD_COMMIT=${2:-}
+VALIDATION_SCRIPT_DIR=${3:-}
+
+# Validate required parameters
+if [[ -z "${BASE_COMMIT}" || -z "${HEAD_COMMIT}" || -z "${VALIDATION_SCRIPT_DIR}" ]]; then
+  echo "Usage: $0 <BASE_COMMIT> <HEAD_COMMIT> <VALIDATION_SCRIPT_DIR>" >&2
+  exit 1
+fi
+
+COMMIT_RANGE="${BASE_COMMIT}..${HEAD_COMMIT}"
+SCRIPT_DIR="scripts/${VALIDATION_SCRIPT_DIR}"
+
+echo "Retrieving components to run against the following script: ${VALIDATION_SCRIPT_DIR}"
+echo "Using BASE_COMMIT: ${BASE_COMMIT}"
+echo "Using HEAD_COMMIT: ${HEAD_COMMIT}"
+
+# If diff is detected within the input script directory, validate all components.
+SCRIPT_DIFF=$(git diff --name-only --diff-filter=ACM "${COMMIT_RANGE}" -- "${SCRIPT_DIR}" || true)
+if [[ -n "${SCRIPT_DIFF}" ]]; then
+  ALL_COMPONENT_FILES=$(find components -mindepth 2 -maxdepth 2 -type d | \
+    sort -u | \
+    tr '\n' ',' | \
+    sed 's/,$//')
+
+if [[ -z "${ALL_COMPONENT_FILES//[[:space:]]/}" ]]; then
+    echo "Changes detected in ${VALIDATION_SCRIPT_DIR}, but there are no existing components to validate."
+else
+    echo "Changes detected in ${VALIDATION_SCRIPT_DIR}. All existing components will be validated: ${ALL_COMPONENT_FILES}"
+fi
+
+  if [[ -n "${GITHUB_OUTPUT:-}" ]]; then
+    echo "new_components_list=${ALL_COMPONENT_FILES}" >> "${GITHUB_OUTPUT}"
+  else
+    echo "new_components_list=${ALL_COMPONENT_FILES}"
+  fi
+
+else
+  # If no script diff is detected, check for new components.
+  GIT_DIFF=$(git diff --name-only --diff-filter=A "${COMMIT_RANGE}" -- "components" || true)
+  NEW_COMPONENTS=()
+  FORMATTED_COMPONENTS=""
+
+  # Retrieve list of component categories.
+  COMPONENT_TYPES=$(find components -mindepth 2 -maxdepth 2 -type d | \
+    sed 's/^components\///' | \
+    sort -u)
+
+  # Identify all new components in each category by checking all existing components against GIT_DIFF.
+  for dir in ${COMPONENT_TYPES}; do
+    echo "DIR: components/${dir}"
+    if [[ " ${GIT_DIFF} " =~ components/"${dir}" ]]; then
+      NEW_COMPONENTS+=("components/${dir}")
+    fi
+  done
+
+  # Format the list of new components for output.
+  if ((${#NEW_COMPONENTS[@]} > 0)); then
+    FORMATTED_COMPONENTS=$(printf '%s\n' "${NEW_COMPONENTS[@]}" | \
+      sort -u | \
+      tr '\n' ',' | \
+      sed 's/,$//')
+  else
+    FORMATTED_COMPONENTS=""
+  fi
+
+  if [[ -n "${FORMATTED_COMPONENTS}" ]]; then
+    echo "The following new components were found and will be validated: ${FORMATTED_COMPONENTS}"
+  else
+    echo "No new components detected."
+  fi
+
+  if [[ -n "${GITHUB_OUTPUT:-}" ]]; then
+    echo "new_components_list=${FORMATTED_COMPONENTS}" >> "${GITHUB_OUTPUT}"
+  else
+    echo "new_components_list=${FORMATTED_COMPONENTS}"
+  fi
+
+fi

.github/workflows/validate-component-metadata-schema.yml

@@ -0,0 +1,45 @@
+name: validate-metadata-schema
+env:
+  PYTHON_VERSION: 3.11
+  COMPONENT_SCRIPT_PATH: $GITHUB_WORKSPACE/.github/resources/scripts/get_components_for_validation.sh
+  VALIDATION_SCRIPT_PATH: $GITHUB_WORKSPACE/scripts/validate_metadata/validate_metadata.py
+on:
+  pull_request:
+    paths:
+      - 'components/**'
+      - 'scripts/**'
+
+jobs:
+  validate-component-metadata-schema:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v5
+        with:
+          fetch-depth: 0
+
+      - name: Set up Python
+        uses: actions/setup-python@v6
+        with:
+          python-version: ${{ env.PYTHON_VERSION }}
+
+      - name: Install Test dependencies
+        run: |
+          pip install -r scripts/validate_metadata/requirements.txt
+
+      - name: Retrieve new components
+        id: get-new-components
+        run: ${{  env.COMPONENT_SCRIPT_PATH  }} "${{ github.event.pull_request.base.sha }}" "${{ github.event.pull_request.head.sha }}" validate_metadata
+
+      - name: Validate new components
+        if: ${{ steps.get-new-components.outputs.new_components_list != '' }}
+        run: |
+          NEW_COMPONENTS_ARRAY="${{ steps.get-new-components.outputs.new_components_list }}"
+
+          # 2. Set IFS to a comma, so that the shell will split the string by commas.
+          IFS=','
+          
+          for component in $NEW_COMPONENTS_ARRAY; do
+            COMPONENT_PATH="$GITHUB_WORKSPACE/$component"
+            echo "Processing component: $component"
+            python "${{  env.VALIDATION_SCRIPT_PATH  }}" --component $COMPONENT_PATH
+          done

scripts/validate_metadata/__main__.py

@@ -0,0 +1,8 @@
+"""Entry point for running validate_metadata as a module.
+"""
+
+from .validate_metadata import main
+
+if __name__ == "__main__":
+    main()
+

scripts/validate_metadata/requirements.txt

@@ -0,0 +1,2 @@
+PyYAML==6.0.3
+semver==3.0.4

scripts/validate_metadata/test_data/component_directories_metadata/component_missing_metadata_file/OWNERS

@@ -0,0 +1,2 @@
+approvers:
+- sample-approver

scripts/validate_metadata/test_data/component_directories_metadata/component_missing_owners_file/metadata.yaml

@@ -0,0 +1,22 @@
+tier: third_party
+name: happy-path-component
+stability: alpha | beta | stable
+dependencies:
+  kubeflow:
+    - name: Pipelines
+      version: '>=2.5'
+    - name: Trainer
+      version: '>=2.0'
+  external_services:
+    - name: Argo Workflows
+      version: "3.6"
+tags:
+  - training
+  - evaluation
+lastVerified: 2025-03-15T00:00:00Z
+ci:
+  skip_dependency_probe: false
+  pytest: optional
+links:
+  documentation: https://kubeflow.org/components/happy-path-component
+  issue_tracker: https://github.com/kubeflow/kfp-components/issues

scripts/validate_metadata/test_data/component_directories_metadata/component_valid/OWNERS

@@ -0,0 +1,2 @@
+approvers:
+- sample-approver

scripts/validate_metadata/test_data/component_directories_metadata/component_valid/metadata.yaml

@@ -0,0 +1,21 @@
+name: component_valid
+tier: third-party
+stability: alpha
+dependencies:
+  kubeflow:
+    - name: Pipelines
+      version: '2.5.0'
+    - name: Trainer
+      version: '2.0.0'
+  external_services:
+    - name: Argo Workflows
+      version: "3.6.0"
+tags:
+  - training
+  - evaluation
+lastVerified: 2025-03-15T00:00:00Z
+ci:
+  skip_dependency_probe: false
+links:
+  documentation: https://kubeflow.org/components/happy-path-component
+  issue_tracker: https://github.com/kubeflow/kfp-components/issues

scripts/validate_metadata/test_data/component_directories_metadata/not_a_dir.txt

@@ -0,0 +1,20 @@
+name: invalid-ci
+tier: core
+stability: alpha
+dependencies:
+  kubeflow:
+    - name: Pipelines
+      version: '>=2.5.0'
+    - name: Trainer
+      version: '>=2.0.0'
+  external_services:
+    - name: Argo Workflows
+      version: "3.6.0"
+tags:
+  - training
+  - evaluation
+lastVerified: 2025-03-15T00:00:00Z
+ci: invalid-ci-value
+links:
+  documentation: https://kubeflow.org/components/happy-path-component
+  issue_tracker: https://github.com/kubeflow/kfp-components/issues