Experiment with finding feature flag references in the code

Author: emyller

.github/workflows/poc-github-code-references.yml

@@ -0,0 +1,114 @@
+name: 'PoC: GitHub Code References'
+permissions:
+  contents: read
+
+on:
+  schedule:
+    - cron: '0 0 * * *'  # Runs daily at midnight UTC
+  workflow_dispatch:
+  pull_request:  # DROPME
+
+env:
+  FLAGSMITH_PUBLIC_KEY: ENktaJnfLVbLifybz34JmX
+  FLAGSMITH_EDGE_API_URL: https://edge.api.flagsmith.com
+  PYTHON_VERSION: '3.13'
+  PYTHON_REQUESTS_VERSION: '2.32.4'
+
+jobs:
+  collect-code-references:
+    runs-on: depot-ubuntu-latest
+    steps:
+      - name: Checkout code
+        uses: actions/checkout@v4
+
+      - uses: actions/setup-python@v5
+        with:
+          python-version: ${{ env.PYTHON_VERSION }}
+
+      - name: Install dependencies
+        run: |
+          pip install requests==${{ env.PYTHON_REQUESTS_VERSION }}
+
+      - name: Collect code references
+        id: collect
+        shell: python
+        run: |
+          import json
+          import os
+          import re
+          from collections import deque
+          from pathlib import Path
+          from typing import Generator
+
+          import requests
+
+          EXCLUDE_PATTERNS = ["node_modules", "venv", ".git", "cache", "build", "htmlcov", "docs", ".json"]
+
+          def is_file_binary(file_path: Path) -> bool:
+              """Check if a file is binary."""
+              with file_path.open("rb") as file:
+                  chunk = file.read(1024)
+              return b'\0' in chunk
+
+          def find_references(feature_names: list[str]) -> Generator[tuple[str, str, int], None, None]:
+              """Search for references to a feature name in the codebase."""
+              all_files = Path('.').glob("**/*")
+              for path in all_files:
+                  if any(pattern in str(path).lower() for pattern in EXCLUDE_PATTERNS):
+                      continue
+                  if not path.is_file():
+                      continue
+                  if is_file_binary(path):
+                      continue
+                  context: deque[str] = deque(maxlen=2)
+                  with path.open("r", encoding="utf-8", errors="ignore") as file:
+                      for line_number, line in enumerate(file, start=1):
+                          context.append(line)
+                          for feature_name in feature_names:
+                              if feature_name not in line:  # Match feature name
+                                  continue
+                              if (
+                                  re.search(fr"""[A-Z][A-Z0-9_]{3,}\s*=\s*(['"]){feature_name}\1""", line) or  # Constant assignment
+                                  re.search(fr"""(?ism:(?:feature|flag)[^\(]*\(.*(["']){feature_name})\1""", "".join(context))  # Function call
+                              ):
+                                  yield feature_name, str(path), line_number
+
+          # Fetch visible features
+          all_flags = requests.get(f"${{ env.FLAGSMITH_EDGE_API_URL }}/api/v1/flags", headers={"X-Environment-Key": "${{ env.FLAGSMITH_PUBLIC_KEY }}"}).json()
+          feature_names = sorted([flag["feature"]["name"] for flag in all_flags])
+          print("Feature names:", feature_names)
+
+          # Find code references
+          code_references = [
+              (feature_name, file_path, line_number)
+              for feature_name, file_path, line_number in find_references(feature_names)
+          ]
+
+          # Output to GHA
+          json_references = json.dumps(code_references)
+          with open(os.environ["GITHUB_OUTPUT"], "a") as gh_output:
+              print(f"code_references='{json_references}'", file=gh_output)
+
+      - name: Display code references
+        shell: python
+        run: |
+          import json
+          from collections import defaultdict
+
+          code_references = json.loads('''${{ steps.collect.outputs.code_references }}''')
+          if not code_references:
+              print("No code references found.")
+              exit(0)
+
+          references_by_feature = defaultdict(list)
+          for feature_name, file_path, line_number in sorted(code_references):
+              references_by_feature[feature_name].append((file_path, line_number))
+
+          print("Code References:")
+          for feature_name, references in references_by_feature.items():
+              print(f"\nFeature: {feature_name}")
+              for file_path, line_number in references:
+                  print(f"  - {file_path}:{line_number}")
+
+      # TODO
+      # - name: Upload code references