fix: add pytest to build.sh, fix hanging tests, add request timeouts

pipeline-runner/build.sh

@@ -49,10 +49,16 @@ fi
 echo "==> Build"
 cmake --build "${BUILD_DIR}" -j$(nproc)
 
-echo "==> Run tests"
+echo "==> Run C++ tests"
 rm -rf /tmp/pipeline-runner-test/
 "${BUILD_DIR}/test-pipeline-runner" \
   --gtest_brief=1 \
   --gtest_output=xml:"${BUILD_DIR}/test-results.xml"
 
+if [ "${FAKE}" = true ]; then
+  echo "==> Run Python tests"
+  timeout 60 python3 -m pytest "${SCRIPT_DIR}/tests" --tb=line -q \
+    --junitxml="${BUILD_DIR}/python-test-results.xml"
+fi
+
 echo "==> Done"

pipeline-runner/tests/conftest.py

@@ -82,16 +82,16 @@ class PipelineServer:
             except subprocess.TimeoutExpired:
                 self.process.kill()
                 self.process.wait(timeout=5)
+            # Close pipes to prevent resource leaks and zombie processes
+            if self.process.stdout:
+                self.process.stdout.close()
+            if self.process.stderr:
+                self.process.stderr.close()
 
 
 @pytest.fixture(scope="session")
 def pipeline_server():
-    """Session-scoped fixture that builds, starts, and stops the pipeline-runner server."""
+    """Session-scoped fixture that starts and stops the pipeline-runner server."""
-    project_dir = os.path.dirname(os.path.dirname(os.path.abspath(__file__)))
-    build_script = os.path.join(project_dir, "build.sh")
-
-    subprocess.check_call([build_script, "--fake"], cwd=project_dir)
-
     work_dir = tempfile.mkdtemp(prefix="pipeline-test-")
     port = get_free_port()
 

pipeline-runner/tests/test_integration.py

@@ -3,8 +3,11 @@
 import time
 import uuid
 
+import pytest
 import requests
 
+REQUEST_TIMEOUT = 5
+
 
 def _unique_id():
     return str(uuid.uuid4())
@@ -26,7 +29,9 @@ def test_full_pipeline_lifecycle(pipeline_server):
             {"id": _unique_id(), "text": "echo step3", "title": "Step 3", "order": 2},
         ],
     }
-    resp = requests.post(f"{base}/api/pipelines", json=pipeline)
+    resp = requests.post(
+        f"{base}/api/pipelines", json=pipeline, timeout=REQUEST_TIMEOUT
+    )
     assert resp.status_code == 200
     created = resp.json()
     assert created["id"] == pipeline_id
@@ -34,13 +39,15 @@ def test_full_pipeline_lifecycle(pipeline_server):
     assert len(created["prompts"]) == 3
 
     # --- List pipelines ---
-    resp = requests.get(f"{base}/api/pipelines")
+    resp = requests.get(f"{base}/api/pipelines", timeout=REQUEST_TIMEOUT)
     assert resp.status_code == 200
     pipelines = resp.json()
     assert any(p["id"] == pipeline_id for p in pipelines)
 
     # --- Get single pipeline ---
-    resp = requests.get(f"{base}/api/pipelines/{pipeline_id}")
+    resp = requests.get(
+        f"{base}/api/pipelines/{pipeline_id}", timeout=REQUEST_TIMEOUT
+    )
     assert resp.status_code == 200
     fetched = resp.json()
     assert fetched["id"] == pipeline_id
@@ -49,7 +56,9 @@ def test_full_pipeline_lifecycle(pipeline_server):
     assert fetched["prompts"][0]["order"] == 0
 
     # --- Start a run ---
-    resp = requests.post(f"{base}/api/runs", json={"pipeline_id": pipeline_id})
+    resp = requests.post(
+        f"{base}/api/runs", json={"pipeline_id": pipeline_id}, timeout=REQUEST_TIMEOUT
+    )
     assert resp.status_code == 200
     run_data = resp.json()
     run_id = run_data["run_id"]
@@ -58,7 +67,9 @@ def test_full_pipeline_lifecycle(pipeline_server):
 
     # --- Poll run status until completed ---
     for _ in range(60):
-        resp = requests.get(f"{base}/api/runs/{run_id}/status")
+        resp = requests.get(
+            f"{base}/api/runs/{run_id}/status", timeout=REQUEST_TIMEOUT
+        )
         assert resp.status_code == 200
         status_data = resp.json()
         if status_data["status"] in ("completed", "error"):
@@ -71,15 +82,19 @@ def test_full_pipeline_lifecycle(pipeline_server):
     assert len(status_data["steps"]) == 3
 
     # --- Get log ---
-    resp = requests.get(f"{base}/api/runs/{run_id}/log")
+    resp = requests.get(f"{base}/api/runs/{run_id}/log", timeout=REQUEST_TIMEOUT)
     assert resp.status_code == 200
     assert len(resp.text) > 0
 
     # --- Delete pipeline ---
-    resp = requests.delete(f"{base}/api/pipelines/{pipeline_id}")
+    resp = requests.delete(
+        f"{base}/api/pipelines/{pipeline_id}", timeout=REQUEST_TIMEOUT
+    )
     assert resp.status_code == 200
     assert resp.json()["deleted"] == pipeline_id
 
     # --- Verify deletion ---
-    resp = requests.get(f"{base}/api/pipelines/{pipeline_id}")
+    resp = requests.get(
+        f"{base}/api/pipelines/{pipeline_id}", timeout=REQUEST_TIMEOUT
+    )
     assert resp.status_code == 404

pipeline-runner/tests/test_smoke.py

@@ -1,16 +1,15 @@
 """Smoke tests for pipeline-runner REST API endpoints."""
 
-import json
 import os
 import shutil
 import socket
-import subprocess
 import tempfile
-import time
 
 import pytest
 import requests
 
+REQUEST_TIMEOUT = 5
+
 
 test_get_endpoints = [
     ("api/pipelines", 200, list),
@@ -21,7 +20,9 @@ test_get_endpoints = [
 @pytest.mark.parametrize("endpoint,expected_status,expected_type", test_get_endpoints)
 def test_get_endpoints(pipeline_server, endpoint, expected_status, expected_type):
     """Check HTTP status code and response content-type for GET endpoints."""
-    response = requests.get(f"{pipeline_server.base_url}/{endpoint}")
+    response = requests.get(
+        f"{pipeline_server.base_url}/{endpoint}", timeout=REQUEST_TIMEOUT
+    )
     assert response.status_code == expected_status
     assert "application/json" in response.headers.get("Content-Type", "")
     data = response.json()
@@ -30,7 +31,9 @@ def test_get_endpoints(pipeline_server, endpoint, expected_status, expected_type
 
 def test_cors_headers(pipeline_server):
     """Check that CORS headers are present on responses."""
-    response = requests.get(f"{pipeline_server.base_url}/api/pipelines")
+    response = requests.get(
+        f"{pipeline_server.base_url}/api/pipelines", timeout=REQUEST_TIMEOUT
+    )
     assert response.headers.get("Access-Control-Allow-Origin") == "*"
 
 
@@ -45,10 +48,6 @@ def _get_free_port():
 @pytest.fixture(scope="module")
 def static_file_server():
     """Fixture that starts a server with web_root configured for static file tests."""
-    project_dir = os.path.dirname(os.path.dirname(os.path.abspath(__file__)))
-    build_script = os.path.join(project_dir, "build.sh")
-    subprocess.check_call([build_script, "--fake"], cwd=project_dir)
-
     work_dir = tempfile.mkdtemp(prefix="pipeline-static-test-")
 
     web_root = os.path.join(work_dir, "web")
@@ -88,7 +87,7 @@ def static_file_server():
 class TestStaticFiles:
     def test_serve_html(self, static_file_server):
         server, _ = static_file_server
-        resp = requests.get(f"{server.base_url}/index.html")
+        resp = requests.get(f"{server.base_url}/index.html", timeout=REQUEST_TIMEOUT)
         assert resp.status_code == 200
         assert "text/html" in resp.headers.get("Content-Type", "")
         assert "<h1>Hello Static</h1>" in resp.text
@@ -96,28 +95,32 @@ class TestStaticFiles:
 
     def test_serve_css(self, static_file_server):
         server, _ = static_file_server
-        resp = requests.get(f"{server.base_url}/style.css")
+        resp = requests.get(f"{server.base_url}/style.css", timeout=REQUEST_TIMEOUT)
         assert resp.status_code == 200
         assert "text/css" in resp.headers.get("Content-Type", "")
         assert resp.text == "body { color: red; }"
 
     def test_serve_js(self, static_file_server):
         server, _ = static_file_server
-        resp = requests.get(f"{server.base_url}/app.js")
+        resp = requests.get(f"{server.base_url}/app.js", timeout=REQUEST_TIMEOUT)
         assert resp.status_code == 200
         assert "text/javascript" in resp.headers.get("Content-Type", "")
         assert resp.text == "console.log('hello');"
 
     def test_serve_nested_file(self, static_file_server):
         server, _ = static_file_server
-        resp = requests.get(f"{server.base_url}/subdir/nested.txt")
+        resp = requests.get(
+            f"{server.base_url}/subdir/nested.txt", timeout=REQUEST_TIMEOUT
+        )
         assert resp.status_code == 200
         assert "text/plain" in resp.headers.get("Content-Type", "")
         assert resp.text == "nested file content"
 
     def test_etag_304(self, static_file_server):
         server, _ = static_file_server
-        resp1 = requests.get(f"{server.base_url}/index.html")
+        resp1 = requests.get(
+            f"{server.base_url}/index.html", timeout=REQUEST_TIMEOUT
+        )
         assert resp1.status_code == 200
         etag = resp1.headers.get("ETag")
         assert etag is not None
@@ -125,17 +128,22 @@ class TestStaticFiles:
         resp2 = requests.get(
             f"{server.base_url}/index.html",
             headers={"If-None-Match": etag},
+            timeout=REQUEST_TIMEOUT,
         )
         assert resp2.status_code == 304
         assert resp2.text == ""
 
     def test_cache_control_header(self, static_file_server):
         server, _ = static_file_server
-        resp = requests.get(f"{server.base_url}/index.html")
+        resp = requests.get(
+            f"{server.base_url}/index.html", timeout=REQUEST_TIMEOUT
+        )
         assert resp.status_code == 200
         assert "max-age" in resp.headers.get("Cache-Control", "")
 
     def test_static_404(self, static_file_server):
         server, _ = static_file_server
-        resp = requests.get(f"{server.base_url}/nonexistent.html")
+        resp = requests.get(
+            f"{server.base_url}/nonexistent.html", timeout=REQUEST_TIMEOUT
+        )
         assert resp.status_code == 404