test(perf): add benchmark tests for core operations

Adds micro-benchmarks for Job.from_dict, JobRequest.to_dict, JSON
serialization round-trips, and JobState.is_terminal to track
deserialization performance regressions.

Author: josedab

tests/test_benchmark.py

@@ -0,0 +1,150 @@
+"""Benchmark tests for OJS Python SDK core operations.
+
+Run with: pytest tests/test_benchmark.py -v
+"""
+
+import json
+import time
+
+import pytest
+
+from ojs.job import Job, JobRequest, JobState
+
+
+MINIMAL_JOB_DICT = {
+    "id": "019539a4-b68c-7def-8000-1a2b3c4d5e6f",
+    "type": "email.send",
+    "state": "available",
+    "queue": "default",
+    "args": [{"to": "user@example.com"}],
+    "attempt": 0,
+    "created_at": "2024-01-15T10:30:00Z",
+}
+
+FULL_JOB_DICT = {
+    "id": "019539a4-b68c-7def-8000-1a2b3c4d5e6f",
+    "type": "email.send",
+    "state": "active",
+    "queue": "email",
+    "args": [{"to": "user@example.com", "subject": "Welcome!"}],
+    "priority": 10,
+    "attempt": 2,
+    "max_retries": 5,
+    "tags": ["onboarding", "email"],
+    "meta": {"campaign_id": "123", "source": "api"},
+    "created_at": "2024-01-15T10:30:00Z",
+    "scheduled_at": "2024-01-15T11:00:00Z",
+}
+
+
+class TestJobDeserialization:
+    """Benchmark Job.from_dict() for various payload sizes."""
+
+    def test_bench_from_dict_minimal(self) -> None:
+        iterations = 10_000
+        start = time.perf_counter()
+        for _ in range(iterations):
+            Job.from_dict(MINIMAL_JOB_DICT)
+        elapsed = time.perf_counter() - start
+        per_op = elapsed / iterations * 1_000_000  # microseconds
+        print(f"\n  Job.from_dict (minimal): {per_op:.2f} µs/op ({iterations} iterations)")
+        assert per_op < 1000, f"Too slow: {per_op:.2f} µs/op"
+
+    def test_bench_from_dict_full(self) -> None:
+        iterations = 10_000
+        start = time.perf_counter()
+        for _ in range(iterations):
+            Job.from_dict(FULL_JOB_DICT)
+        elapsed = time.perf_counter() - start
+        per_op = elapsed / iterations * 1_000_000
+        print(f"\n  Job.from_dict (full): {per_op:.2f} µs/op ({iterations} iterations)")
+        assert per_op < 1000, f"Too slow: {per_op:.2f} µs/op"
+
+
+class TestJobRequestSerialization:
+    """Benchmark JobRequest.to_dict() for various payload sizes."""
+
+    def test_bench_to_dict_minimal(self) -> None:
+        req = JobRequest(type="email.send", args=["user@example.com"])
+        iterations = 10_000
+        start = time.perf_counter()
+        for _ in range(iterations):
+            req.to_dict()
+        elapsed = time.perf_counter() - start
+        per_op = elapsed / iterations * 1_000_000
+        print(f"\n  JobRequest.to_dict (minimal): {per_op:.2f} µs/op ({iterations} iterations)")
+        assert per_op < 1000, f"Too slow: {per_op:.2f} µs/op"
+
+    def test_bench_to_dict_full(self) -> None:
+        req = JobRequest(
+            type="email.send",
+            args=["user@example.com", "Welcome!"],
+            queue="email",
+            priority=10,
+            tags=["onboarding"],
+        )
+        iterations = 10_000
+        start = time.perf_counter()
+        for _ in range(iterations):
+            req.to_dict()
+        elapsed = time.perf_counter() - start
+        per_op = elapsed / iterations * 1_000_000
+        print(f"\n  JobRequest.to_dict (full): {per_op:.2f} µs/op ({iterations} iterations)")
+        assert per_op < 1000, f"Too slow: {per_op:.2f} µs/op"
+
+
+class TestJsonSerialization:
+    """Benchmark raw JSON serialization round-trips."""
+
+    def test_bench_json_dumps_minimal(self) -> None:
+        iterations = 10_000
+        start = time.perf_counter()
+        for _ in range(iterations):
+            json.dumps(MINIMAL_JOB_DICT)
+        elapsed = time.perf_counter() - start
+        per_op = elapsed / iterations * 1_000_000
+        print(f"\n  json.dumps (minimal): {per_op:.2f} µs/op ({iterations} iterations)")
+
+    def test_bench_json_dumps_full(self) -> None:
+        iterations = 10_000
+        start = time.perf_counter()
+        for _ in range(iterations):
+            json.dumps(FULL_JOB_DICT)
+        elapsed = time.perf_counter() - start
+        per_op = elapsed / iterations * 1_000_000
+        print(f"\n  json.dumps (full): {per_op:.2f} µs/op ({iterations} iterations)")
+
+    def test_bench_json_loads_minimal(self) -> None:
+        data = json.dumps(MINIMAL_JOB_DICT)
+        iterations = 10_000
+        start = time.perf_counter()
+        for _ in range(iterations):
+            json.loads(data)
+        elapsed = time.perf_counter() - start
+        per_op = elapsed / iterations * 1_000_000
+        print(f"\n  json.loads (minimal): {per_op:.2f} µs/op ({iterations} iterations)")
+
+    def test_bench_json_loads_full(self) -> None:
+        data = json.dumps(FULL_JOB_DICT)
+        iterations = 10_000
+        start = time.perf_counter()
+        for _ in range(iterations):
+            json.loads(data)
+        elapsed = time.perf_counter() - start
+        per_op = elapsed / iterations * 1_000_000
+        print(f"\n  json.loads (full): {per_op:.2f} µs/op ({iterations} iterations)")
+
+
+class TestJobStateOperations:
+    """Benchmark state-related operations."""
+
+    def test_bench_is_terminal(self) -> None:
+        states = list(JobState)
+        iterations = 100_000
+        start = time.perf_counter()
+        for _ in range(iterations):
+            for s in states:
+                s.is_terminal()
+        elapsed = time.perf_counter() - start
+        per_op = elapsed / (iterations * len(states)) * 1_000_000
+        print(f"\n  JobState.is_terminal: {per_op:.2f} µs/op ({iterations * len(states)} iterations)")