feat: bump to 0.16.0, add checkpoint result, callbacks, and binance example

Cargo.toml

@@ -1,6 +1,6 @@
 [package]
 name = "laminardb"
-version = "0.15.1"
+version = "0.16.0"
 edition = "2024"
 rust-version = "1.85"
 license = "MIT"

examples/binance_signals.py

@@ -0,0 +1,59 @@
+"""Binance BTC trade signals in 30 lines of Python.
+
+WebSocket → Stream → Signal — zero infrastructure, pure SQL.
+
+Run: python examples/binance_signals.py
+"""
+
+import laminardb
+
+db = laminardb.open(":memory:")
+
+# 1. Ingest live BTC trades straight from Binance WebSocket
+db.execute("""
+    CREATE SOURCE trades (
+        s VARCHAR, p DOUBLE, q DOUBLE, "T" BIGINT,
+        WATERMARK FOR "T" AS "T" - INTERVAL '0' SECOND
+    ) FROM WEBSOCKET (
+        url    = 'wss://stream.binance.com:9443/ws/btcusdt@trade',
+        format = 'json'
+    )
+""")
+
+# 2. 10-second VWAP windows with BUY/SELL/HOLD signal
+db.execute("""
+    CREATE STREAM signals AS
+    SELECT
+        s                            AS symbol,
+        SUM(p * q) / SUM(q)          AS vwap,
+        AVG(p)                       AS avg_price,
+        COUNT(*)                     AS trades,
+        CASE
+            WHEN AVG(p) > SUM(p * q) / SUM(q) * 1.001 THEN 'SELL'
+            WHEN AVG(p) < SUM(p * q) / SUM(q) * 0.999 THEN 'BUY'
+            ELSE 'HOLD'
+        END AS signal
+    FROM trades
+    GROUP BY s, TUMBLE("T", INTERVAL '10' SECOND)
+    EMIT ON WINDOW CLOSE
+""")
+
+# 3. Start pipeline and print signals as they arrive
+db.start()
+
+print("Listening for BTC signals...\n")
+sub = db.subscribe_stream("signals")
+
+try:
+    while True:
+        batch = sub.next()
+        if batch:
+            for row in batch.fetchall():
+                symbol, vwap, avg_price, num_trades, signal = row
+                print(f"  {signal:4s}  {symbol}  vwap=${vwap:,.2f}  avg=${avg_price:,.2f}  ({num_trades} trades)")
+except KeyboardInterrupt:
+    print("\nStopped.")
+finally:
+    sub.cancel()
+    db.shutdown()
+    db.close()