Add validation for source repos definitions

Adds a script which checks if the source repositories definitions in
ansible/inventory/group_vars/all/source-repositories and
terraform/github/terraform.tfvars.json are out of sync.

Calls this script via a workflow on PR and pushes to main.

Author: MoteHue

.github/workflows/validate-source-repos.yml

@@ -0,0 +1,30 @@
+---
+name: Validate source repositories
+
+on:
+  push:
+    branches:
+      - main
+  pull_request:
+jobs:
+  validate-source-repos:
+    name: Validate source repositories
+
+    runs-on: ubuntu-latest
+
+    steps:
+      - uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # v6.0.2
+
+      # Python version must be pinned because of issue with Ubuntu permissions
+      # See https://github.com/actions/runner-images/issues/11499
+      - name: Set up Python
+        uses: actions/setup-python@a309ff8b426b58ec0e2a45f0f869d46889d02405 # v6.2.0
+        with:
+          python-version: '3.12'
+
+      - name: Install pip dependencies
+        run: |
+          pip install -r requirements.txt
+
+      - name: Validate source repositories
+        run: python3 scripts/validate-source-repos.py

scripts/validate-source-repos.py

@@ -0,0 +1,86 @@
+import json
+import sys
+import yaml
+
+
+def read_repos_data():
+    terraform_repositories = {}
+    ansible_repositories = {}
+
+    with open('terraform/github/terraform.tfvars.json', 'r') as file:
+        data = json.load(file)
+        terraform_repositories = data['repositories']
+
+    with open('ansible/inventory/group_vars/all/source-repositories',
+              'r') as file:
+        data = yaml.safe_load(file)
+        ansible_repositories = data['source_repositories']
+
+    return terraform_repositories, ansible_repositories
+
+
+def get_repos_diff(tf_repos, ans_repos):
+    ans_repos_list = list(ans_repos.keys())
+
+    tf_repos_list = []
+    for team in tf_repos:
+        tf_repos_list += tf_repos[team]
+
+    diff = list(set(tf_repos_list).symmetric_difference(ans_repos_list))
+
+    return diff
+
+
+def _extract_group_from_community_files(files):
+    for file in files:
+        if 'codeowners' in file:
+            content = file['codeowners']['content']
+            return content.strip('{ }').replace('_', '').split('.')[-1]
+
+
+def _sort_ans_repos_by_group(ans_repos):
+    ans_repos_by_group = {}
+    for repo in ans_repos:
+        files = ans_repos[repo]['community_files']
+        group = _extract_group_from_community_files(files)
+        if group in ans_repos_by_group:
+            ans_repos_by_group[group].append(repo)
+        else:
+            ans_repos_by_group[group] = [repo]
+    return ans_repos_by_group
+
+
+def get_mismatched_repos(tf_repos, ans_repos, repos_missing):
+    ans_repos_new = _sort_ans_repos_by_group(ans_repos)
+    tf_repos_new = {k.lower(): v for k, v in tf_repos.items()}
+
+    mismatched_repos = []
+    for group in tf_repos_new:
+        if len(tf_repos_new[group]) == 0 or group not in ans_repos_new:
+            continue
+        else:
+            diff = list(set(tf_repos_new[group]).symmetric_difference(
+                set(ans_repos_new[group])))
+            mismatched_repos.extend(diff)
+    return list(set(mismatched_repos).difference(set(repos_missing)))
+
+
+def main():
+    terraform_repos, ansible_repos = read_repos_data()
+
+    repos_missing = get_repos_diff(terraform_repos, ansible_repos)
+
+    print('The following repos are only present in one of the Ansible '
+          f'source-repositories and the Terraform tfvars: {repos_missing}')
+
+    mismatched_repos = get_mismatched_repos(terraform_repos, ansible_repos,
+                                            repos_missing)
+
+    print('The following repos are assigned to different codeowner groups in '
+         'the Ansible source-repositories and the Terraform tfvars: '
+         f'{mismatched_repos}')
+
+    return len(repos_missing) > 0 or len(mismatched_repos) > 0
+
+if __name__ == "__main__":
+    sys.exit(main())