Reduce unnecessary PR full-suite CI runs

Author: MaxGhenis

.github/workflows/pr_code_changes.yaml

@@ -6,11 +6,18 @@ on:
     branches:
       - main
     paths:
+      - pyproject.toml
+      - uv.lock
+      - modal_app/**
       - policyengine_us_data/**
       - tests/**
       - .github/workflows/**
       - Makefile
 
+concurrency:
+  group: pr-code-changes-${{ github.event.pull_request.number }}
+  cancel-in-progress: true
+
 jobs:
   check-fork:
     runs-on: ubuntu-latest
@@ -30,6 +37,88 @@ jobs:
           fi
           echo "✅ PR is from the correct repository"
 
+  decide-test-scope:
+    name: Decide PR test scope
+    runs-on: ubuntu-latest
+    needs: check-fork
+    outputs:
+      full_suite: ${{ steps.decide.outputs.full_suite }}
+      reason: ${{ steps.decide.outputs.reason }}
+    steps:
+      - uses: actions/checkout@v4
+        with:
+          fetch-depth: 0
+
+      - id: decide
+        env:
+          BASE_SHA: ${{ github.event.pull_request.base.sha }}
+          HEAD_SHA: ${{ github.event.pull_request.head.sha }}
+          PR_LABELS_JSON: ${{ toJson(github.event.pull_request.labels.*.name) }}
+        run: |
+          python - <<'PY'
+          import fnmatch
+          import json
+          import os
+          import subprocess
+
+          labels = set(json.loads(os.environ["PR_LABELS_JSON"]))
+          changed_files = subprocess.check_output(
+              [
+                  "git",
+                  "diff",
+                  "--name-only",
+                  os.environ["BASE_SHA"],
+                  os.environ["HEAD_SHA"],
+              ],
+              text=True,
+          ).splitlines()
+
+          full_suite_label = "full-data-ci"
+          critical_patterns = [
+              ".github/workflows/pr_code_changes.yaml",
+              ".github/workflows/reusable_test.yaml",
+              "modal_app/**",
+              "policyengine_us_data/calibration/**",
+              "policyengine_us_data/datasets/**",
+              "policyengine_us_data/db/**",
+              "policyengine_us_data/storage/download_private_prerequisites.py",
+              "policyengine_us_data/utils/loss.py",
+              "policyengine_us_data/utils/mortgage_interest.py",
+              "policyengine_us_data/utils/soi.py",
+              "policyengine_us_data/utils/uprating.py",
+          ]
+
+          matched_files = [
+              path
+              for path in changed_files
+              if any(fnmatch.fnmatch(path, pattern) for pattern in critical_patterns)
+          ]
+
+          if full_suite_label in labels:
+              full_suite = True
+              reason = f"label:{full_suite_label}"
+          elif matched_files:
+              full_suite = True
+              reason = f"critical-path:{matched_files[0]}"
+          else:
+              full_suite = False
+              reason = "basic-pytest-only"
+
+          with open(os.environ["GITHUB_OUTPUT"], "a", encoding="utf-8") as output:
+              output.write(f"full_suite={'true' if full_suite else 'false'}\n")
+              output.write(f"reason={reason}\n")
+
+          summary = [
+              "### PR test scope",
+              f"- full suite: `{'true' if full_suite else 'false'}`",
+              f"- reason: `{reason}`",
+          ]
+          if matched_files:
+              summary.append(f"- first matching file: `{matched_files[0]}`")
+          with open(os.environ["GITHUB_STEP_SUMMARY"], "a", encoding="utf-8") as out:
+              out.write("\n".join(summary) + "\n")
+          PY
+
   check-lock-freshness:
     name: Check uv.lock freshness
     runs-on: ubuntu-latest
@@ -81,10 +170,10 @@ jobs:
         run: python -c "from policyengine_core.data import Dataset; print('Core import OK')"
 
   Test:
-    needs: [check-fork, Lint]
+    needs: [check-fork, Lint, decide-test-scope]
     uses: ./.github/workflows/reusable_test.yaml
     with:
-      full_suite: true
+      full_suite: ${{ needs.decide-test-scope.outputs.full_suite == 'true' }}
       upload_data: false
       deploy_docs: false
-    secrets: inherit
+    secrets: inherit

changelog.d/cheaper-pr-ci.fixed.md

@@ -0,0 +1,2 @@
+Reduce unnecessary PR CI spend by canceling superseded runs and limiting
+the Modal-backed full data build to labeled or high-risk data-pipeline changes.