feat: initial trading platform implementation

Binance spot crypto trading platform with microservices architecture:
- shared: Pydantic models, Redis Streams broker, asyncpg DB layer
- data-collector: Binance WebSocket/REST market data collection
- strategy-engine: Plugin-based strategy execution (RSI, Grid)
- order-executor: Order execution with risk management
- portfolio-manager: Position tracking and PnL calculation
- backtester: Historical strategy testing with simulator
- cli: Click-based CLI for all operations
- Docker Compose orchestration with Redis and PostgreSQL
- 24 test files covering all modules

1 files changed, 53 insertions, 0 deletions

diff --git a/services/data-collector/src/data_collector/binance_rest.py b/services/data-collector/src/data_collector/binance_rest.py
new file mode 100644
index 0000000..af0eb77
--- /dev/null
+++ b/services/data-collector/src/data_collector/binance_rest.py
@@ -0,0 +1,53 @@
+"""Binance REST API helpers for fetching historical candle data."""
+from datetime import datetime, timezone
+from decimal import Decimal
+
+from shared.models import Candle
+
+
+def _normalize_symbol(symbol: str) -> str:
+    """Convert 'BTC/USDT' to 'BTCUSDT'."""
+    return symbol.replace("/", "")
+
+
+async def fetch_historical_candles(
+    exchange,
+    symbol: str,
+    timeframe: str,
+    since: int,
+    limit: int = 500,
+) -> list[Candle]:
+    """Fetch historical OHLCV candles from the exchange and return Candle models.
+
+    Args:
+        exchange: An async ccxt exchange instance.
+        symbol: Market symbol, e.g. 'BTC/USDT'.
+        timeframe: Candle timeframe, e.g. '1m'.
+        since: Start timestamp in milliseconds.
+        limit: Maximum number of candles to fetch.
+
+    Returns:
+        A list of Candle model instances.
+    """
+    rows = await exchange.fetch_ohlcv(symbol, timeframe, since=since, limit=limit)
+
+    normalized = _normalize_symbol(symbol)
+    candles: list[Candle] = []
+
+    for row in rows:
+        ts_ms, o, h, l, c, v = row
+        open_time = datetime.fromtimestamp(ts_ms / 1000, tz=timezone.utc)
+        candles.append(
+            Candle(
+                symbol=normalized,
+                timeframe=timeframe,
+                open_time=open_time,
+                open=Decimal(str(o)),
+                high=Decimal(str(h)),
+                low=Decimal(str(l)),
+                close=Decimal(str(c)),
+                volume=Decimal(str(v)),
+            )
+        )
+
+    return candles