signal-fabric-client 0.1.15

Client library for Signal Fabric

Signal Fabric Client

Official Python client library for Signal Fabric - a lightweight handler-based framework for generating market signals on demand.

Installation

pip install signal-fabric-client

Quick Start

from signal_fabric import GrpcClient, SignalOutcome
import json

with GrpcClient(host='localhost', port=9090,
                ca_cert_path='../../certs/client/test.pem') as client:
    outcome : SignalOutcome = client.process_signal(
        target='spot:BTC/USDT',
        signal_name='binance_rsi',
        signal_op='compute_rsi',
        handler_request={
            "period":  14,
            "timeframe": "1h"
        }
    )
    # Check for errors first
    if outcome.errors:
        print('We got errors:')
        for err_id, err_message in outcome.errors.items():
            print(f"  - {err_id}: {err_message}")
        print(f"\nResult: {outcome.result}")
        print(f"Computation: {outcome.computation}")
    else:
        # No errors, parse the result
        raw_result = outcome.result
        if not raw_result:
            print("Error: Empty result received")
        else:
            try:
                resultObj = json.loads(raw_result)
                print(f"""market: {resultObj['market']}
symbol: {resultObj['symbol']}
latest_rsi: {resultObj['latest_rsi']}
regime: {resultObj['regime']}
computed_at: {outcome.computed_at}
""")
            except json.JSONDecodeError as e:
                print(f"Error parsing result as JSON: {e}")
                print(f"Raw result: {raw_result}")

API Reference

GrpcClient

Constructor

GrpcClient(host: str = 'localhost', port: int = 50051, timeout: int = 30, ca_cert_path: str = None)

Parameters:

• host (str): Server hostname or IP address (default: 'localhost')
• port (int): Server port number (default: 50051)
• timeout (int): Request timeout in seconds (default: 30)
• ca_cert_path (str): Path to CA certificate for server verification (PEM format)

Methods

connect()

Establish connection to the server.

disconnect()

Close the connection to the server.

is_connected() -> bool

Check if client is currently connected.

process_signal(target, signal_name, signal_op, handler_request=None) -> SignalOutcome

Process a signal request.

Parameters:

• target (str): Target for signal computation (e.g., 'BTC', 'ETH', 'AAPL')
• signal_name (str): Signal handler name or profile name
• signal_op (str): Operation to perform (e.g., 'analyze', 'greet')
• handler_request (dict, optional): Request parameters as dictionary

Returns: SignalOutcome object

SignalOutcome

Result object containing the signal computation outcome.

Attributes

• result (str): Signal result value
• computation (str): Description of computation performed
• computed_at (float): Unix timestamp when computed
• errors (Dict[str, str]): Error codes and messages (or empty)
• details (Dict[str, str]): Additional computation details (or empty)

Methods

has_errors() -> bool

Returns True if the outcome contains errors.

is_detailed() -> bool

Returns True if the outcome has errors or additional details.

Requirements

• Python 3.8+
• grpcio >= 1.76.0
• protobuf >= 4.0.0

Server Setup

This client requires a running Signal Fabric server. To set up the server:

1. Clone the Signal Fabric repository:

   git clone https://github.com/phasequant/signal-fabric.git
   cd signal-fabric
   

2. Install dependencies:

   pip install -r requirements.txt
   

3. Start the server:

   python src/server/main.py --config config.yaml
   

Version

Current version: 0.1.15

License

See LICENSE file for details.

Links

• Signal Fabric Repository
• Documentation
• Issue Tracker

Support

For questions, issues, or contributions, please visit the GitHub repository.