Skip to content

📝 Streaming Quote History (Example)

Request: Stream historical price bars (OHLC) using safe defaults. This demo wrapper prints batches of candles and exits after a short timeout.

Code Example

// Quick use: last 90 days of H1 candles for a symbol,
// delivered in weekly chunks. Demo loop times out after ~30s.
svc.StreamQuoteHistoryExample(ctx, "EURUSD")

Method Signature

// Demo wrapper: safe defaults + console printing
func (s *MT4Service) StreamQuoteHistoryExample(ctx context.Context, symbol string)

Defaults used inside the example wrapper:

  • Date range: from = now - 90 days, to = now
  • Timeframe: QH_PERIOD_H1
  • Chunk duration: 7 * 24h (weekly)
  • Demo timeout: \~30 seconds (select-case exit)

Under the hood, it calls the service API:

func (s *MT4Service) StreamQuoteHistory(
    ctx context.Context,
    symbol string,
    timeframe pb.ENUM_QUOTE_HISTORY_TIMEFRAME,
    from, to time.Time,
    chunk time.Duration,
) (<-chan *pb.QuoteHistoryData, <-chan error)

🔽 Input

Field Type Required Description
ctx context.Context ✅ Controls lifetime (cancel/timeout) of stream
symbol string ✅ Trading symbol, e.g. "EURUSD"

The Example wrapper fixes timeframe/chunk/range to sensible defaults. Use the service API for full control.

⬆️ Output

Console output of OHLC bars from each *pb.QuoteHistoryData batch. For each candle, the example prints:

  • Time (UTC),
  • Open, Close (and can be extended to High, Low).

Sample line format:

[YYYY-mm-dd HH:MM:SS] O: <open> C: <close>

The proto may expose additional fields (e.g., High, Low, Volume). Print whatever you need.

ENUM: pb.ENUM_QUOTE_HISTORY_TIMEFRAME

Common values (check your generated pb for exact names):

  • QH_PERIOD_M1, QH_PERIOD_M5, QH_PERIOD_M15, QH_PERIOD_M30
  • QH_PERIOD_H1, QH_PERIOD_H4
  • QH_PERIOD_D1, QH_PERIOD_W1, QH_PERIOD_MN1

🧩 Notes & Tips

  • Connection required: Ensure ConnectByServerName or ConnectByHostPort succeeded before streaming.
  • Chunking: Larger chunk ⇒ fewer round-trips (better throughput). Smaller chunk ⇒ lower memory spikes.
  • Auto-reconnect: The underlying stream retries on transient gRPC/API errors with exponential backoff + jitter.
  • Back-pressure: If you do heavy work per batch, offload to a worker so the reader loop never blocks.

⚠️ Pitfalls

  • Ignoring errCh: Always select on the error channel to catch terminal failures.
  • Huge ranges: Very wide ranges on small timeframes can generate large data; consider paging by time (chunks) appropriately.
  • Timezone confusion: Timestamps are UTC in the proto; format/convert on display as needed.

🧪 Testing

  • Happy path: Expect multiple batches for the last 90 days of H1 data.
  • Network hiccups: Brief outages should be retried automatically; expect a short pause.
  • Cancel path: Cancel the passed ctx and verify both channels close and the goroutine exits.