feat(data): OHLCV loader via ccxt with parquet cache

Adjust .gitignore to keep src/multi_swarm/data/ tracked (only
top-level /data/ cache directory remains ignored).

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

tests/unit/test_ohlcv_loader.py

@@ -0,0 +1,64 @@
+from datetime import UTC, datetime
+from pathlib import Path
+
+import pandas as pd
+import pytest
+
+from multi_swarm.data.ohlcv_loader import OHLCVLoader, OHLCVRequest
+
+
+@pytest.fixture
+def sample_ohlcv_rows():
+    base_ts = int(datetime(2024, 1, 1, tzinfo=UTC).timestamp() * 1000)
+    rows = []
+    for i in range(48):
+        rows.append(
+            [base_ts + i * 3600 * 1000, 40000 + i, 40100 + i, 39900 + i, 40050 + i, 100.0 + i]
+        )
+    return rows
+
+
+def test_loader_fetches_and_caches(tmp_path: Path, mocker, sample_ohlcv_rows):
+    fake_exchange = mocker.MagicMock()
+    fake_exchange.fetch_ohlcv.return_value = sample_ohlcv_rows
+    mocker.patch("multi_swarm.data.ohlcv_loader.ccxt.binance", return_value=fake_exchange)
+
+    loader = OHLCVLoader(cache_dir=tmp_path)
+    req = OHLCVRequest(
+        symbol="BTC/USDT",
+        timeframe="1h",
+        start=datetime(2024, 1, 1, tzinfo=UTC),
+        end=datetime(2024, 1, 3, tzinfo=UTC),
+    )
+    df = loader.load(req)
+
+    assert isinstance(df, pd.DataFrame)
+    assert list(df.columns) == ["open", "high", "low", "close", "volume"]
+    assert len(df) == 48
+    assert df.index.is_monotonic_increasing
+    cache_files = list(tmp_path.glob("*.parquet"))
+    assert len(cache_files) == 1
+
+
+def test_loader_uses_cache_on_second_call(tmp_path: Path, mocker, sample_ohlcv_rows):
+    fake_exchange = mocker.MagicMock()
+    fake_exchange.fetch_ohlcv.return_value = sample_ohlcv_rows
+    mocker.patch("multi_swarm.data.ohlcv_loader.ccxt.binance", return_value=fake_exchange)
+
+    loader = OHLCVLoader(cache_dir=tmp_path)
+    req = OHLCVRequest(
+        symbol="BTC/USDT",
+        timeframe="1h",
+        start=datetime(2024, 1, 1, tzinfo=UTC),
+        end=datetime(2024, 1, 3, tzinfo=UTC),
+    )
+    df1 = loader.load(req)
+    df2 = loader.load(req)
+
+    assert fake_exchange.fetch_ohlcv.call_count == 2  # paginazione interna, non caching
+    pd.testing.assert_frame_equal(df1, df2)
+    # Seconda chiamata legge da cache, non chiama exchange
+    fake_exchange.fetch_ohlcv.reset_mock()
+    df3 = loader.load(req)
+    assert fake_exchange.fetch_ohlcv.call_count == 0
+    pd.testing.assert_frame_equal(df1, df3)