Client | OHLC Aggregator

The client_ohlc (Open-High-Low-Close) module listens to a continuous stream of executions and market trades, converting them into discrete, time-bucketed candlesticks optimized for technical analysis and charting.

Overview

Live market data arrives as a firehose of individual ticks. To display a readable financial chart, client_app relies entirely on this module to mathematically slice the data into defined intervals (e.g., 1m, 5m, 1H).

Key Responsibilities

• Process rapid MarketDataIncrementalRefresh (35=X) or Trade sequences.
• Aggregate ticks into sequential time intervals based on configured granularities.
• Maintain the current 'live' unfinalized candle.
• Format output vectors specifically for graphing libraries (like ImPlot).

Architecture

graph TD
    subgraph "FIX Data Feed"
        FEED(client_fix: Market Data)
    end
 
    subgraph "client_ohlc Module"
        ROUTER[TickRouter]
        1M[1M Aggregator]
        H1[1 Hour Aggregator]
    end
 
    FEED --> ROUTER
    ROUTER -->|Tick Price/Qty| 1M
    ROUTER -->|Tick Price/Qty| H1
    
    1M -.->|Candle Vector| UI(client_app Charts)

Class Diagram

classDiagram
    class OHLCBuilder {
        +addTrade(Price, Quantity, Timestamp)
        +getCandles(Interval) : vector~Candlestick~
        +getCurrentCandle() : Candlestick
        -finalizeCandle()
        -m_currentCandle: Candlestick
        -m_history: vector~Candlestick~
        -m_intervalMs: uint64_t
    }
    
    class Candlestick {
        +startTime: Timestamp
        +open: double
        +high: double
        +low: double
        +close: double
        +volume: uint64_t
    }
 
    OHLCBuilder *--> Candlestick : Generates

Component Responsibilities

Component	Description
**OHLCBuilder**	State machine assigned to a specific string of intervals and a specific symbol. Dictates when a candle closes.
**Candlestick**	Mathematical container. The core unit recognized universally by charting tools.
**addTrade()**	Immediately updates the high, low, and close properties of the active candle. Pushes it to m_history if the tick's timestamp crosses the m_intervalMs threshold.

Critical Design Conventions

• Memory Contiguous Output: The m_history vector must be memory-contiguous so that visualizers like ImPlot can map array pointers directly to GPU buffers without iterating through objects.
• No UI Dependency: This module is completely detached from the renderer; it merely holds the arrays.