rmi-backend/app/databus/ohlcv_engine.py

396 lines
12 KiB
Python

"""
RugCharts OHLCV Aggregation Engine
===================================
Real-time candle building from trade events.
Produces OHLCV bars at 1m, 5m, 15m, 1h, 4h, 1d timeframes.
Stored in Redis sorted sets for O(log N) range queries.
Wired into DataBus as 'ohlcv' chain.
"""
import json
import logging
import os
import time
from datetime import UTC, datetime
import redis
logger = logging.getLogger("ohlcv_engine")
REDIS_HOST = os.getenv("REDIS_HOST", "rmi-redis")
REDIS_PORT = int(os.getenv("REDIS_PORT", "6379"))
REDIS_PASSWORD = os.getenv("REDIS_PASSWORD", "")
TIMEFRAMES = {
"1m": 60,
"5m": 300,
"15m": 900,
"1h": 3600,
"4h": 14400,
"1d": 86400,
}
CACHE_TTL = {
"1m": 300, # 5 min
"5m": 900, # 15 min
"15m": 1800, # 30 min
"1h": 3600, # 1 hour
"4h": 14400, # 4 hours
"1d": 86400, # 24 hours
}
MAX_CANDLES = 500 # Max candles returned per query
def _redis():
return redis.Redis(
host=REDIS_HOST,
port=REDIS_PORT,
password=REDIS_PASSWORD,
decode_responses=True,
socket_connect_timeout=2,
)
def _frame_key(token: str, chain: str, timeframe: str) -> str:
"""Redis sorted set key for OHLCV candles."""
return f"ohlcv:{chain}:{token}:{timeframe}"
def _candle_cache_key(token: str, chain: str, timeframe: str, limit: int, end_ts: int | None = None) -> str:
"""Cache key for bulk candle queries."""
end = end_ts or int(time.time())
return f"ohlcv_cache:{chain}:{token}:{timeframe}:{limit}:{end}"
class Candle:
"""Single OHLCV candle."""
__slots__ = ("close", "high", "low", "open", "timestamp", "trades", "volume")
def __init__(
self,
ts: int,
open_p: float = 0,
high: float = 0,
low: float = float("inf"),
close: float = 0,
volume: float = 0,
trades: int = 0,
):
self.timestamp = ts
self.open = open_p
self.high = high
self.low = low
self.close = close
self.volume = volume
self.trades = trades
def to_dict(self) -> dict:
return {
"timestamp": self.timestamp,
"datetime": datetime.fromtimestamp(self.timestamp, tz=UTC).isoformat(),
"open": round(self.open, 8),
"high": round(self.high, 8),
"low": round(self.low, 8),
"close": round(self.close, 8),
"volume": round(self.volume, 2),
"trades": self.trades,
}
def to_array(self) -> list:
"""Compact array format: [ts, o, h, l, c, v, n]"""
return [
self.timestamp,
round(self.open, 8),
round(self.high, 8),
round(self.low, 8),
round(self.close, 8),
round(self.volume, 2),
self.trades,
]
class OHLCVEngine:
"""Real-time OHLCV candle builder and query engine."""
def __init__(self):
self._active_candles: dict[str, Candle] = {} # key → current open candle
def _active_key(self, token: str, chain: str, timeframe: str, bucket_ts: int) -> str:
return f"{chain}:{token}:{timeframe}:{bucket_ts}"
def ingest_trade(
self,
token: str,
chain: str,
price: float,
volume: float,
timestamp: int | None = None,
commit: bool = True,
) -> dict[str, Candle]:
"""Ingest a single trade, updating all timeframe candles.
Returns dict of timeframe → updated Candle.
"""
ts = timestamp or int(time.time())
updated = {}
for tf_name, tf_seconds in TIMEFRAMES.items():
bucket_ts = (ts // tf_seconds) * tf_seconds
active_key = self._active_key(token, chain, tf_name, bucket_ts)
candle = self._active_candles.get(active_key)
if candle is None or candle.timestamp != bucket_ts:
# Close old candle and start new
if candle and commit:
self._persist_candle(token, chain, tf_name, candle)
candle = Candle(
ts=bucket_ts,
open_p=price,
high=price,
low=price,
close=price,
volume=volume,
trades=1,
)
self._active_candles[active_key] = candle
else:
# Update open candle
if candle.open == 0:
candle.open = price
candle.high = max(candle.high, price)
candle.low = min(candle.low, price)
candle.close = price
candle.volume += volume
candle.trades += 1
updated[tf_name] = candle
return updated
def _persist_candle(self, token: str, chain: str, timeframe: str, candle: Candle):
"""Write candle to Redis sorted set."""
try:
r = _redis()
key = _frame_key(token, chain, timeframe)
# Store as JSON in sorted set (score = timestamp)
r.zadd(key, {json.dumps(candle.to_dict()): candle.timestamp})
# Trim to MAX_CANDLES
r.zremrangebyrank(key, 0, -(MAX_CANDLES + 1))
r.expire(key, CACHE_TTL.get(timeframe, 3600))
r.close()
except Exception as e:
logger.warning(f"Failed to persist candle: {e}")
def flush_all(self):
"""Persist all active candles to Redis."""
for active_key, candle in list(self._active_candles.items()):
parts = active_key.split(":")
if len(parts) >= 4:
chain, token, tf_name = parts[0], parts[1], parts[2]
self._persist_candle(token, chain, tf_name, candle)
self._active_candles.clear()
def get_candles(
self,
token: str,
chain: str,
timeframe: str = "1h",
limit: int = 100,
end_ts: int | None = None,
) -> list[dict]:
"""Retrieve OHLCV candles for a token.
Args:
token: Token address
chain: Blockchain ID
timeframe: '1m', '5m', '15m', '1h', '4h', '1d'
limit: Max candles (default 100, max 500)
end_ts: End timestamp (default: now). Returns candles up to this time.
"""
if timeframe not in TIMEFRAMES:
timeframe = "1h"
limit = min(limit, MAX_CANDLES)
end = end_ts or int(time.time())
# Check cache first
cache_key = _candle_cache_key(token, chain, timeframe, limit, end)
try:
r = _redis()
cached = r.get(cache_key)
if cached:
r.close()
return json.loads(cached)
except Exception:
pass
try:
r = _redis() if "r" not in dir() or not r else r
key = _frame_key(token, chain, timeframe)
# Get candles up to end_ts
raw = r.zrangebyscore(key, 0, end, start=0, num=limit)
candles = []
for item in raw:
try:
c = json.loads(item)
candles.append(c)
except Exception:
pass
# Sort by timestamp ascending
candles.sort(key=lambda x: x["timestamp"])
# Include active candle if within range
tf_seconds = TIMEFRAMES[timeframe]
bucket_ts = (end // tf_seconds) * tf_seconds
active_key = self._active_key(token, chain, timeframe, bucket_ts)
active = self._active_candles.get(active_key)
if active and active.timestamp <= end:
# Avoid duplicating if already persisted
if not candles or candles[-1]["timestamp"] != active.timestamp:
candles.append(active.to_dict())
# Cache the result
r.setex(cache_key, 60, json.dumps(candles[-limit:] if len(candles) > limit else candles))
r.close()
return candles[-limit:] if len(candles) > limit else candles
except Exception as e:
logger.warning(f"OHLCV query failed: {e}")
return []
def get_latest_candle(self, token: str, chain: str, timeframe: str = "1h") -> dict | None:
"""Get the most recent candle (may be still-open)."""
candles = self.get_candles(token, chain, timeframe, limit=1)
return candles[0] if candles else None
def get_price_change(self, token: str, chain: str, periods: int = 24, timeframe: str = "1h") -> dict:
"""Calculate price change over N periods."""
candles = self.get_candles(token, chain, timeframe, limit=periods + 1)
if len(candles) < 2:
return {"change_pct": 0, "candles": 0}
first = candles[0]["close"]
last = candles[-1]["close"]
change_pct = ((last - first) / first * 100) if first > 0 else 0
return {
"change_pct": round(change_pct, 2),
"open": first,
"close": last,
"high": max(c["high"] for c in candles),
"low": min(c["low"] for c in candles),
"volume": sum(c["volume"] for c in candles),
"candles": len(candles),
}
def stats(self) -> dict:
"""Engine statistics."""
return {
"active_candles": len(self._active_candles),
"timeframes": list(TIMEFRAMES.keys()),
"max_candles_per_query": MAX_CANDLES,
}
# ── Singleton ──────────────────────────────────────────────────────
ohlcv_engine = OHLCVEngine()
# ── DataBus Provider ────────────────────────────────────────────────
async def fetch_ohlcv(
token: str = "", chain: str = "ethereum", timeframe: str = "1h", limit: int = 100, **kw
) -> dict | None:
"""DataBus provider for OHLCV candle data.
Args:
token: Token address (use mint= or address= as aliases)
chain: Blockchain (ethereum, solana, bsc, base, etc.)
timeframe: 1m, 5m, 15m, 1h, 4h, 1d
limit: Number of candles (default 100, max 500)
"""
address = token or kw.get("mint", "") or kw.get("address", "")
if not address:
return None
candles = ohlcv_engine.get_candles(address, chain, timeframe, limit)
# Calculate summary stats
summary = {}
if candles:
prices = [c["close"] for c in candles]
volumes = [c["volume"] for c in candles]
summary = {
"current_price": prices[-1],
"price_change_pct": round(((prices[-1] - prices[0]) / prices[0] * 100), 2) if prices[0] > 0 else 0,
"high_24h": max(c["high"] for c in candles),
"low_24h": min(c["low"] for c in candles),
"volume_24h": sum(volumes),
"total_trades": sum(c["trades"] for c in candles),
}
# Also compute authenticity if we have volume data
authenticity = None
try:
from app.databus.volume_authenticity import quick_authenticity_score
auth = quick_authenticity_score(
volume_24h=summary.get("volume_24h", 0),
liquidity=float(kw.get("liquidity_usd", 0)),
unique_wallets=int(kw.get("unique_wallets", 0)),
tx_count=summary.get("total_trades", 0),
)
authenticity = {
"fake_volume_pct": auth.get("fake_volume_pct", 0),
"authentic_score": auth.get("authentic_score", 100),
"risk_level": auth.get("risk_level", "UNKNOWN"),
}
except Exception:
pass
return {
"candles": candles,
"summary": summary,
"authenticity": authenticity,
"timeframe": timeframe,
"token": address,
"chain": chain,
"source": "ohlcv_engine",
}
async def ingest_trade_data(
token: str = "",
chain: str = "ethereum",
price: float = 0,
volume: float = 0,
timestamp: int | None = None,
**kw,
) -> dict | None:
"""DataBus provider: Ingest a trade and update all OHLCV candles."""
address = token or kw.get("address", "") or kw.get("token", "")
if not address or price <= 0:
return None
price = float(price)
volume = float(volume)
updated = ohlcv_engine.ingest_trade(address, chain, price, volume, timestamp)
return {
"status": "ingested",
"token": address,
"chain": chain,
"price": price,
"volume": volume,
"updated_timeframes": list(updated.keys()),
"source": "ohlcv_engine",
}