feat(ebook-search): add load-test CLI for the search service

Add a Typer CLI script that drives POST /search on a running server at a
configurable concurrency and reports latency percentiles (p50/p90/p95/p99),
throughput, and HTTP status distribution. Queries are drawn from the shared
eval JSONL set so load testing and evaluation exercise the same questions.

tests/ebook_search/test_loadtest.py

@@ -0,0 +1,79 @@
+"""Tests for the load-test runner and its statistics helpers."""
+
+from __future__ import annotations
+
+import asyncio
+from typing import TYPE_CHECKING
+
+import pytest
+
+from python.ebook_search.loadtest import RequestResult, load_queries, percentile, run_load, summarize
+
+if TYPE_CHECKING:
+    from pytest_mock import MockerFixture
+
+
+def test_load_queries_reads_shared_set() -> None:
+    queries = load_queries(None)
+    assert len(queries) == 70
+    assert all(isinstance(query, str) and query for query in queries)
+
+
+def test_percentile_interpolates() -> None:
+    values = [10.0, 20.0, 30.0, 40.0]
+    assert percentile(values, 50) == pytest.approx(25.0)
+    assert percentile(values, 90) == pytest.approx(37.0)
+    assert percentile(values, 0) == 10.0
+    assert percentile(values, 100) == 40.0
+    assert percentile([], 95) == 0.0
+
+
+def test_summarize_counts_and_throughput() -> None:
+    results = [
+        RequestResult(status_code=200, latency_ms=10.0, ok=True),
+        RequestResult(status_code=200, latency_ms=20.0, ok=True),
+        RequestResult(status_code=200, latency_ms=30.0, ok=True),
+        RequestResult(status_code=500, latency_ms=40.0, ok=False),
+    ]
+    summary = summarize(results, wall_seconds=2.0)
+
+    assert summary.total == 4
+    assert summary.successes == 3
+    assert summary.failures == 1
+    assert summary.throughput_rps == pytest.approx(2.0)
+    assert summary.latency_max_ms == 40.0
+    assert summary.status_counts == {200: 3, 500: 1}
+
+
+def test_summarize_handles_empty() -> None:
+    summary = summarize([], wall_seconds=0.0)
+    assert summary.total == 0
+    assert summary.throughput_rps == 0.0
+    assert summary.latency_p95_ms == 0.0
+
+
+def test_run_load_aggregates_mocked_responses(mocker: MockerFixture) -> None:
+    response = mocker.Mock(status_code=200, is_success=True)
+    client = mocker.MagicMock()
+    client.__aenter__.return_value = client
+    client.post = mocker.AsyncMock(return_value=response)
+    mocker.patch("python.ebook_search.loadtest.httpx.AsyncClient", return_value=client)
+
+    summary = asyncio.run(
+        run_load(
+            base_url="http://test",
+            queries=["q1", "q2"],
+            request_count=4,
+            concurrency=2,
+            rerank=False,
+            warmup=1,
+            timeout_seconds=1.0,
+        )
+    )
+
+    assert summary.total == 4
+    assert summary.successes == 4
+    assert summary.failures == 0
+    assert summary.status_counts == {200: 4}
+    # 1 warmup request (not measured) plus 4 measured requests.
+    assert client.post.await_count == 5