pkbrokers 0.1.20251228.48

A Python-based stock broker module

PKBrokers

Platforms
Package / Docs
Tests/Code-Quality

---

Table of Contents

• What is PKBrokers?
• Installation
• Quick Start
• Architecture Overview
• Core Modules
  • In-Memory Candle Store
  • Data Manager
  • Kite Instruments
  • Tick Watcher
  • Local Candle Database
  • Telegram Bots
  • Authentication
• API Reference
• Environment Variables
• Contributing
• Related Projects

---

What is PKBrokers?

PKBrokers is a high-performance Python library for connecting to stock brokers (primarily Zerodha's Kite Connect) to fetch real-time market data, instruments, and ticks. Key features include:

• 🚀 High-Performance Candle Store - O(1) access to OHLCV candles across 10 timeframes
• 📊 Real-Time Tick Processing - WebSocket-based tick aggregation
• 💾 Multi-Source Data Management - SQLite, Turso, pickle files, and Kite API
• 🤖 Telegram Bot Integration - Distribute tick data via Telegram
• 🔐 Automated Authentication - TOTP-based Kite login
• 📦 24/7 Data Availability - GitHub-based data persistence

This library is part of the PKScreener ecosystem.

---

Installation

From PyPI

pip install pkbrokers

From Source

git clone https://github.com/pkjmesra/pkbrokers.git
cd pkbrokers
pip install -r requirements.txt
pip install -e .

Requirements

• Python 3.9+
• Zerodha Kite Connect account (for real-time data)
• See requirements.txt for dependencies

---

Quick Start

High-Performance Data Provider

from pkbrokers.kite import get_candle_store, HighPerformanceDataProvider

# Get singleton candle store
store = get_candle_store()

# Or use high-level data provider
provider = HighPerformanceDataProvider()

# Get 5-minute candles for any stock
df = provider.get_stock_data("RELIANCE", interval="5m", count=50)

# Get current day's OHLCV
ohlcv = provider.get_current_ohlcv("TCS")
print(f"Open: {ohlcv['open']}, High: {ohlcv['high']}, Low: {ohlcv['low']}, Close: {ohlcv['close']}")

Data Manager (Multi-Source)

from pkbrokers.kite.datamanager import InstrumentDataManager

# Initialize manager
manager = InstrumentDataManager()

# Execute data synchronization
success = manager.execute()

if success:
    # Access stock data
    reliance = manager.pickle_data["RELIANCE"]
    df = pd.DataFrame(
        data=reliance['data'],
        columns=reliance['columns'],
        index=reliance['index']
    )
    print(f"Shape: {df.shape}")

Kite Authentication

from pkbrokers.kite.examples.externals import kite_auth

# Authenticate and get access token
# Requires KUSER, KPWD, KTOTP environment variables
kite_auth()

# Token is now available as KTOKEN
from PKDevTools.classes.Environment import PKEnvironment
token = PKEnvironment().KTOKEN

---

Architecture Overview

┌─────────────────────────────────────────────────────────────────────────────┐
│                            PKBrokers Architecture                           │
├─────────────────────────────────────────────────────────────────────────────┤
│                                                                             │
│  ┌────────────────────────────────────────────────────────────────┐         │
│  │                     Application Layer                          │         │
│  │            PKScreener | Custom Applications                    │         │
│  └─────────────────────────────┬──────────────────────────────────┘         │
│                                │                                            │
│  ┌─────────────────────────────▼──────────────────────────────────┐         │
│  │                      Data Provider API                         │         │
│  │   HighPerformanceDataProvider | InstrumentDataManager          │         │
│  └─────────────────────────────┬──────────────────────────────────┘         │
│                                │                                            │
│       ┌────────────────────────┼────────────────────────┐                   │
│       │                        │                        │                   │
│  ┌────▼────┐           ┌───────▼───────┐        ┌───────▼───────┐           │
│  │InMemory │           │  Local SQLite │        │ Remote Data   │           │
│  │Candle   │           │  Database     │        │ (GitHub/Turso)│           │
│  │Store    │           │               │        │               │           │
│  └────┬────┘           └───────────────┘        └───────────────┘           │
│       │                                                                     │
│  ┌────▼──────────────────────────────────────────────────────────-┐         │
│  │                    Tick Processing Layer                       │         │
│  │   KiteTokenWatcher | CandleAggregator | TickProcessor          │         │
│  └────────────────────────────┬───────────────────────────────────┘         │
│                               │                                             │
│  ┌────────────────────────────▼───────────────────────────────────┐         │
│  │                     WebSocket Layer                            │         │
│  │          ZerodhaWebSocketClient | KiteTicker                   │         │
│  └────────────────────────────┬───────────────────────────────────┘         │
│                               │                                             │
│  ┌────────────────────────────▼───────────────────────────────────┐         │
│  │                  Kite Connect API / Authentication             │         │
│  │            Authenticator | KiteInstruments                     │         │
│  └────────────────────────────────────────────────────────────────┘         │
│                                                                             │
│  ┌────────────────────────────────────────────────────────────────┐         │
│  │                      Bot Layer (Telegram)                      │         │
│  │         PKTickBot | Orchestrator | Consumer                    │         │
│  └────────────────────────────────────────────────────────────────┘         │
│                                                                             │
└─────────────────────────────────────────────────────────────────────────────┘

See more details on Architecture

Core Modules

1. In-Memory Candle Store

High-performance, in-memory OHLCV storage with O(1) access to all timeframes.

from pkbrokers.kite.inMemoryCandleStore import InMemoryCandleStore, get_candle_store

# Get singleton instance
store = get_candle_store()

# Process incoming tick
store.process_tick({
    'instrument_token': 256265,
    'last_price': 21500.50,
    'volume': 1000000,
    'timestamp': datetime.now()
})

# Get completed candles
candles = store.get_candles(
    instrument_token=256265,
    interval='5m',
    count=50
)

# Get current forming candle
current = store.get_current_candle(
    instrument_token=256265,
    interval='5m'
)

# Export to ticks.json
store.save_ticks_json("/path/to/ticks.json")

# Get statistics
stats = store.get_stats()
print(f"Instruments: {stats['instrument_count']}")
print(f"Ticks processed: {stats['ticks_processed']}")

Supported Timeframes

Interval	Description	Max Candles Stored
1m	1 minute	375 (full day)
2m	2 minutes	188
3m	3 minutes	125
4m	4 minutes	94
5m	5 minutes	75
10m	10 minutes	38
15m	15 minutes	25
30m	30 minutes	13
60m	60 minutes	7
day	Daily	1

Features

• O(1) Access: Instant lookup via hash-based indexing
• No Rate Limits: Unlike Yahoo Finance
• Auto-Persistence: Saves to disk every 5 minutes
• Memory Efficient: ~100MB for 2000 instruments
• Thread-Safe: Lock-protected operations

---

2. Data Manager

Comprehensive data synchronization from multiple sources.

from pkbrokers.kite.datamanager import InstrumentDataManager

manager = InstrumentDataManager()

# Set specific stocks (optional)
manager.list_stock_codes = ["RELIANCE", "TCS", "INFY"]

# Execute synchronization
# Priority: SQLite → InMemoryCandleStore → Kite API → Pickle files
success = manager.execute()

# Access data
if success:
    for symbol, data in manager.pickle_data.items():
        df = pd.DataFrame(
            data=data['data'],
            columns=data['columns'],
            index=data['index']
        )
        print(f"{symbol}: {len(df)} rows")

Data Source Priority

1. During Market Hours:

   • Local SQLite database
   • InMemoryCandleStore (real-time ticks)
   • Kite API (authenticated)
   • GitHub ticks.json

2. After Market Hours:

   • Local pickle files
   • Remote GitHub pickle files

---

3. Kite Instruments

Manage instrument data from Kite Connect API.

from pkbrokers.kite.instruments import KiteInstruments, Instrument

# Initialize with credentials
kite = KiteInstruments(
    api_key="your_api_key",
    access_token="your_access_token"
)

# Sync instruments from Kite API
kite.sync_instruments(force_fetch=True)

# Get instrument count
count = kite.get_instrument_count()
print(f"Total instruments: {count}")

# Get NSE stocks only
equities = kite.get_equities(only_nse_stocks=True)

# Get instrument tokens for subscription
tokens = kite.get_instrument_tokens(equities)

# Fetch instrument by token
instrument = kite.get_instrument(256265)  # NIFTY 50
print(f"Symbol: {instrument.tradingsymbol}")

Instrument Model

@dataclass
class Instrument:
    instrument_token: int      # Unique identifier
    exchange_token: str        # Exchange-specific token
    tradingsymbol: str         # Trading symbol (e.g., 'RELIANCE')
    name: Optional[str]        # Full name
    last_price: Optional[float]
    expiry: Optional[str]      # For derivatives
    strike: Optional[float]    # For options
    tick_size: float
    lot_size: int
    instrument_type: str       # EQ, FUT, OPT, INDEX
    segment: str               # NSE, BSE
    exchange: str
    last_updated: str
    nse_stock: bool

---

4. Tick Watcher

WebSocket-based real-time tick processing.

from pkbrokers.kite.kiteTokenWatcher import KiteTokenWatcher

# Initialize watcher
watcher = KiteTokenWatcher()

# Start watching (blocking)
try:
    watcher.watch(test_mode=False)
except KeyboardInterrupt:
    watcher.stop()

Command-Line Usage

# Start tick watcher
pkkite --ticks

# Test mode (3 minutes)
pkkite --ticks --test

# Authenticate first
pkkite --auth

# Fetch historical data
pkkite --history=5minute

---

5. Local Candle Database

SQLite-based candle storage for persistence.

from pkbrokers.kite.localCandleDatabase import LocalCandleDatabase

# Initialize database
db = LocalCandleDatabase()

# Save daily candle
db.save_daily_candle(
    symbol="RELIANCE",
    date=date.today(),
    open_price=2500.0,
    high_price=2550.0,
    low_price=2480.0,
    close_price=2530.0,
    volume=1000000
)

# Load candles
candles = db.load_daily_candles("RELIANCE", days=30)

# Save intraday candles
db.save_intraday_candle(
    symbol="RELIANCE",
    timestamp=datetime.now(),
    interval="5m",
    open_price=2500.0,
    high_price=2510.0,
    low_price=2495.0,
    close_price=2505.0,
    volume=50000
)

---

6. Telegram Bots

PKTickBot

Telegram bot for distributing tick data.

from pkbrokers.bot.tickbot import PKTickBot

bot = PKTickBot(
    bot_token="your_bot_token",
    ticks_file_path="/path/to/ticks.json",
    chat_id="-1001234567890"
)

# Start bot (blocking)
bot.run()

Available Commands:

Command	Description
/ticks	Get zipped ticks.json file
/db	Get local SQLite database
/status	Check bot and data status
/top	Get top 20 ticking symbols
/token	Get current KTOKEN
/refresh_token	Generate new KTOKEN
/restart	Refresh token and restart watcher
/test_ticks	Start 3-minute tick test
/help	Show help message

Orchestrator

Multi-process orchestrator for bot and data management.

from pkbrokers.bot.orchestrator import Orchestrator

orchestrator = Orchestrator()

# Check if market is open
if orchestrator.should_run_kite_process():
    orchestrator.start_kite_process()

---

7. Authentication

Automated Kite Connect authentication using TOTP.

from pkbrokers.kite.authenticator import KiteAuthenticator

auth = KiteAuthenticator(
    user_id="your_user_id",
    password="your_password",
    totp_secret="your_totp_secret",
    api_key="your_api_key"
)

# Get access token
access_token = auth.authenticate()

# Token is automatically saved to environment

Environment Variables Required:

• KUSER: Kite user ID
• KPWD: Kite password
• KTOTP: TOTP secret key
• KAPI: Kite API key

---

API Reference

Main Exports

from pkbrokers.kite import (
    # Candle Store
    InMemoryCandleStore,
    get_candle_store,
    
    # Data Providers
    HighPerformanceDataProvider,
    InstrumentDataManager,
    
    # Instruments
    KiteInstruments,
    Instrument,
    
    # Tick Processing
    KiteTokenWatcher,
    CandleAggregator,
    
    # Database
    LocalCandleDatabase,
    
    # Authentication
    KiteAuthenticator,
)

from pkbrokers.bot import (
    PKTickBot,
    Orchestrator,
)

Module Structure

pkbrokers/
├── __init__.py
├── bot/
│   ├── __init__.py
│   ├── consumer.py          # Data consumer
│   ├── orchestrator.py      # Multi-process orchestrator
│   └── tickbot.py           # Telegram tick bot
├── kite/
│   ├── __init__.py
│   ├── authenticator.py     # Kite authentication
│   ├── candleAggregator.py  # Tick → Candle aggregation
│   ├── datamanager.py       # Multi-source data manager
│   ├── databasewriter.py    # Database writer
│   ├── inMemoryCandleStore.py  # In-memory candle store
│   ├── instrumentHistory.py # Historical data
│   ├── instruments.py       # Instrument management
│   ├── kiteTokenWatcher.py  # WebSocket tick watcher
│   ├── localCandleDatabase.py  # SQLite candle storage
│   ├── tickProcessor.py     # Tick processing
│   ├── ticks.py             # Tick utilities
│   ├── trader.py            # Trading operations
│   ├── zerodhaWebSocketClient.py  # WebSocket client
│   └── examples/
│       ├── externals.py     # External helpers
│       └── pkkite.py        # CLI entry point
└── scripts/
    └── publish_candle_data.py  # Data publishing

---

Environment Variables

Variable	Required	Description
KUSER	Yes*	Kite user ID
KPWD	Yes*	Kite password
KTOTP	Yes*	TOTP secret for 2FA
KAPI	Yes*	Kite API key
KTOKEN	Auto	Access token (auto-generated)
TOKEN	Yes**	Telegram bot token
CHAT_ID	Yes**	Default Telegram chat ID
TURSO_DB_URL	No	Turso database URL
TURSO_DB_AUTH_TOKEN	No	Turso auth token

*Required for Kite Connect features
**Required for Telegram bot features

---

Contributing

Development Setup

git clone https://github.com/pkjmesra/pkbrokers.git
cd pkbrokers
python -m venv venv
source venv/bin/activate
pip install -r requirements.txt
pip install -e .

Running Tests

pytest test/
pytest --cov=pkbrokers test/

Code Style

ruff check pkbrokers/
ruff format pkbrokers/

---

Related Projects

• PKScreener - Stock screening application
• PKDevTools - Common development tools
• PKNSETools - NSE market data tools

---

License

MIT License - see LICENSE file.

---