Polyglot Integration

Polyglot Integration

Overview

Integrate code written in different programming languages to leverage their unique strengths and ecosystems.

When to Use

• Performance-critical code in C/C++/Rust
• ML models in Python from other languages
• Legacy system integration
• Leveraging language-specific libraries
• Microservices polyglot architecture

Implementation Examples

1. Node.js Native Addons (C++)

// addon.cc
#include <node.h>

namespace demo {

using v8::FunctionCallbackInfo;
using v8::Isolate;
using v8::Local;
using v8::Object;
using v8::String;
using v8::Value;
using v8::Number;

void Add(const FunctionCallbackInfo<Value>& args) {
  Isolate* isolate = args.GetIsolate();

  if (args.Length() < 2) {
    isolate->ThrowException(v8::Exception::TypeError(
        String::NewFromUtf8(isolate, "Wrong number of arguments")));
    return;
  }

  if (!args[0]->IsNumber() || !args[1]->IsNumber()) {
    isolate->ThrowException(v8::Exception::TypeError(
        String::NewFromUtf8(isolate, "Arguments must be numbers")));
    return;
  }

  double value = args[0]->NumberValue() + args[1]->NumberValue();
  Local<Number> num = Number::New(isolate, value);

  args.GetReturnValue().Set(num);
}

void Initialize(Local<Object> exports) {
  NODE_SET_METHOD(exports, "add", Add);
}

NODE_MODULE(NODE_GYP_MODULE_NAME, Initialize)

}

// Usage in Node.js
const addon = require('./build/Release/addon');
console.log(addon.add(3, 5)); // 8

2. Python from Node.js

// Using child_process
import { spawn } from 'child_process';

function callPython(script: string, args: any[] = []): Promise<any> {
  return new Promise((resolve, reject) => {
    const python = spawn('python3', [script, ...args.map(String)]);

    let stdout = '';
    let stderr = '';

    python.stdout.on('data', (data) => {
      stdout += data.toString();
    });

    python.stderr.on('data', (data) => {
      stderr += data.toString();
    });

    python.on('close', (code) => {
      if (code !== 0) {
        reject(new Error(stderr));
      } else {
        try {
          resolve(JSON.parse(stdout));
        } catch (error) {
          resolve(stdout);
        }
      }
    });
  });
}

// Usage
const result = await callPython('./ml_model.py', [100, 200]);
console.log('Python result:', result);

# ml_model.py
import sys
import json

def predict(x, y):
    # ML model logic
    return x * 2 + y

if __name__ == '__main__':
    x = int(sys.argv[1])
    y = int(sys.argv[2])
    result = predict(x, y)
    print(json.dumps({'prediction': result}))

3. Rust from Python (PyO3)

// lib.rs
use pyo3::prelude::*;

#[pyfunction]
fn fibonacci(n: u64) -> PyResult<u64> {
    let mut a = 0;
    let mut b = 1;

    for _ in 0..n {
        let temp = a;
        a = b;
        b = temp + b;
    }

    Ok(a)
}

#[pyfunction]
fn process_large_array(arr: Vec<f64>) -> PyResult<f64> {
    Ok(arr.iter().sum())
}

#[pymodule]
fn rust_extension(_py: Python, m: &PyModule) -> PyResult<()> {
    m.add_function(wrap_pyfunction!(fibonacci, m)?)?;
    m.add_function(wrap_pyfunction!(process_large_array, m)?)?;
    Ok(())
}

# Usage in Python
import rust_extension

result = rust_extension.fibonacci(100)
print(f"Fibonacci: {result}")

arr = list(range(1000000))
total = rust_extension.process_large_array(arr)
print(f"Sum: {total}")

4. gRPC Polyglot Communication

// service.proto
syntax = "proto3";

service Calculator {
  rpc Add (NumberPair) returns (Result);
  rpc Multiply (NumberPair) returns (Result);
}

message NumberPair {
  double a = 1;
  double b = 2;
}

message Result {
  double value = 1;
}

# Python gRPC Server
import grpc
from concurrent import futures
import service_pb2
import service_pb2_grpc

class Calculator(service_pb2_grpc.CalculatorServicer):
    def Add(self, request, context):
        return service_pb2.Result(value=request.a + request.b)

    def Multiply(self, request, context):
        return service_pb2.Result(value=request.a * request.b)

def serve():
    server = grpc.server(futures.ThreadPoolExecutor(max_workers=10))
    service_pb2_grpc.add_CalculatorServicer_to_server(Calculator(), server)
    server.add_insecure_port('[::]:50051')
    server.start()
    server.wait_for_termination()

if __name__ == '__main__':
    serve()

// Node.js gRPC Client
import * as grpc from '@grpc/grpc-js';
import * as protoLoader from '@grpc/proto-loader';

const packageDefinition = protoLoader.loadSync('service.proto');
const proto = grpc.loadPackageDefinition(packageDefinition);

const client = new proto.Calculator(
  'localhost:50051',
  grpc.credentials.createInsecure()
);

client.Add({ a: 10, b: 20 }, (error, response) => {
  if (error) {
    console.error(error);
  } else {
    console.log('Result:', response.value);
  }
});

5. Java from Python (Py4J)

// JavaApp.java
public class JavaApp {
    public int add(int a, int b) {
        return a + b;
    }

    public String processData(String data) {
        return data.toUpperCase();
    }

    public static void main(String[] args) {
        JavaApp app = new JavaApp();
        GatewayServer server = new GatewayServer(app);
        server.start();
    }
}

# Python client
from py4j.java_gateway import JavaGateway

gateway = JavaGateway()
app = gateway.entry_point

result = app.add(10, 20)
print(f"Result: {result}")

processed = app.processData("hello world")
print(f"Processed: {processed}")

Best Practices

✅ DO

• Use appropriate IPC mechanism
• Handle serialization carefully
• Implement proper error handling
• Consider performance overhead
• Use type-safe interfaces
• Document integration points

❌ DON'T

• Pass complex objects across boundaries
• Ignore memory management
• Skip error handling
• Use blocking calls in async code

Integration Methods

Resources

• Node.js C++ Addons
• PyO3
• gRPC
• Py4J