# msghandler - Cross-Platform Communication Layer

A high-performance, **transport-agnostic** communication layer for **Julia**, **JavaScript**, **Python**, and **MicroPython** applications. Implements the Claim-Check pattern for efficient payload transport (direct for small payloads, URL-based for large payloads).

[![License: MIT](https://img.shields.io/badge/License-MIT-blue.svg)](https://opensource.org/licenses/MIT)

---

## Table of Contents

- [Overview](#overview)
- [Cross-Platform Support](#cross-platform-support)
- [Features](#features)
- [Quick Start](#quick-start)
- [API Reference](#api-reference)
- [Cross-Platform Examples](#cross-platform-examples)
- [Testing](#testing)
- [Documentation](#documentation)
- [License](#license)

---

## Quick Data Format Reference

All platforms use the same input format for `smartpack()`:

### Format: `[(dataname, data, type), ...]`

| Element | Type | Description |
|---------|------|-------------|
| `dataname` | String | Name for the payload (e.g., `"message"`, `"config.json"`) |
| `data` | Any | The actual data (type depends on the `type` parameter below) |
| `type` | String | Payload type identifier |

### Supported Payload Types

| Type | Julia Data | JavaScript Data | Python Data | Description |
|------|------------|-----------------|-------------|-------------|
| `"text"` | `String` | `string` | `str` | Plain text |
| `"dictionary"` | `Dict` | `Object` | `dict` | JSON object |
| `"arrowtable"` | `DataFrame` | ❌ | `pandas.DataFrame` | Arrow IPC table (Desktop only) |
| `"jsontable"` | `DataFrame` | `Array<Object>` | `list[dict]` | JSON table |
| `"image"` | `Vector{UInt8}` | `Uint8Array` | `bytes` | Image data |
| `"audio"` | `Vector{UInt8}` | `Uint8Array` | `bytes` | Audio data |
| `"video"` | `Vector{UInt8}` | `Uint8Array` | `bytes` | Video data |
| `"binary"` | `Vector{UInt8}` | `Uint8Array` | `bytes` | Binary data |

### Examples

**Sending multiple payloads:**
```julia
# Julia
data = [
    ("message", "Hello!", "text"),
    ("config", Dict("key" => "value"), "dictionary"),
    ("file", file_bytes, "binary")
]
env, json_str = msghandler.smartpack("subject", data, options...)
```

```python
# Python
data = [
    ("message", "Hello!", "text"),
    ("config", {"key": "value"}, "dictionary"),
    ("file", file_bytes, "binary")
]
env, json_str = await smartpack("subject", data, **options)
```

```javascript
// JavaScript
const data = [
    ["message", "Hello!", "text"],
    ["config", {key: "value"}, "dictionary"],
    ["file", fileBytes, "binary"]
];
const [env, json_str] = await smartpack("subject", data, options);
```

**Important Notes:**
- Always wrap payloads in a list `[]`, even for single payloads
- Type must be one of the supported types above
- Large payloads (≥500KB) automatically use link transport
- Browser only supports: `"text"`, `"dictionary"`, `"jsontable"`, `"image"`, `"audio"`, `"video"`, `"binary"`
  (No Arrow IPC due to browser incompatibility)

---

## Overview

msghandler provides a **transport-agnostic** communication layer with intelligent payload transport selection:

| Transport | Payload Size | Method |
|-----------|--------------|--------|
| **Direct** | < 500KB | Sent directly via chosen transport (Base64 encoded) |
| **Link** | ≥ 500KB | Uploaded to HTTP file server, URL sent via chosen transport |

### Use Cases

- **Chat Applications**: Text, images, audio, video in a single message
- **File Transfer**: Efficient transfer of large files using claim-check pattern
- **IoT/Embedded**: Sensor data, telemetry, and analytics pipelines (MicroPython)
- **Cross-Platform Communication**: Interoperability between Julia, JavaScript, Python, and MicroPython systems
- **Any Transport**: Works with NATS, HTTP, WebSockets, WebRTC, or any custom transport mechanism

### Key Design Principles

- **Transport Agnostic**: Core API (`smartpack`/`smartunpack`) is decoupled from transport
- **Claim-Check Pattern**: Efficient handling of large payloads via URL references
- **Type System**: Rich payload types (text, dictionary, arrowtable, jsontable, binary)
- **Cross-Platform**: Unified API across Julia, JavaScript, Python, and MicroPython

---

## Cross-Platform Support

| Platform | Implementation | Features |
|----------|----------------|----------|
| **Julia** | [`src/msghandler.jl`](src/msghandler.jl) | Full feature set, Arrow IPC, multiple dispatch |
| **JavaScript (Node.js)** | [`src/msghandler_ssr.js`](src/msghandler_ssr.js) | Node.js, async/await, Arrow IPC |
| **JavaScript (Browser)** | [`src/msghandler_csr.js`](src/msghandler_csr.js) | Browser, async/await, JSON table only |
| **Python** | [`src/msghandler.py`](src/msghandler.py) | Desktop Python, asyncio, type hints, Arrow IPC |
| **MicroPython** | [`src/msghandler_mpy.py`](src/msghandler_mpy.py) | Memory-constrained, synchronous API |

### Platform Comparison

| Feature | Julia | JavaScript | JavaScript (Browser) | Python | MicroPython |
|---------|-------|------------|----------------------|--------|-------------|
| Multiple Dispatch | ✅ Native | ❌ | ❌ | ❌ | ❌ |
| Async/Await | ❌ | ✅ Native | ✅ Native | ✅ Native | ⚠️ (uasyncio) |
| Type Safety | ✅ Strong | ⚠️ (TypeScript) | ⚠️ (TypeScript) | ✅ (Type hints) | ❌ |
| Arrow IPC | ✅ Native | ✅ Native | ❌ (Browser incompatible) | ✅ Native | ❌ |
| JSON Table | ✅ | ✅ | ✅ (Only table type) | ✅ | ⚠️ (Limited) |
| Direct Transport | ✅ | ✅ | ✅ | ✅ | ✅ |
| Link Transport | ✅ | ✅ | ✅ | ✅ | ⚠️ (Limited) |
| Handler Functions | ✅ | ✅ | ✅ | ✅ | ✅ |
| Cross-Platform API | ✅ | ✅ | ✅ | ✅ | ✅ |

---

## Features

- ✅ **Transport-agnostic** - Core API works with any communication channel (NATS, HTTP, WebSockets, etc.)
- ✅ **Cross-platform** - Unified API for Julia, JavaScript, Python, and MicroPython
- ✅ **Bi-directional** - Request-reply patterns with `reply_to` support
- ✅ **Multi-payload** - Send multiple payloads of different types in one message
- ✅ **Automatic transport selection** - Direct vs link based on payload size (threshold: 500KB)
- ✅ **Claim-Check pattern** - Efficient large payload handling via URL references
- ✅ **Rich payload types** - Text, dictionary, Arrow IPC, JSON table, image, audio, video, binary
- ✅ **Exponential backoff** - Reliable file server downloads with retry logic
- ✅ **Correlation ID** - End-to-end message tracing
- ✅ **Handler abstraction** - Pluggable file server implementations (Plik, AWS S3, custom)

---

## Quick Start

### Prerequisites

1. **Transport Channel** - Choose your communication channel:
    - **NATS** (recommended): `docker run -p 4222:4222 nats:latest`
    - **HTTP/WebSocket**: Any HTTP server or WebSocket endpoint
    - **Custom**: Implement your own transport mechanism

2. **HTTP File Server** (optional, for payloads ≥ 500KB) - Install and run:
    ```bash
    # Using Plik
    docker run -p 8080:8080 -v /tmp/fileserver:/var/lib/plik -e PLIK_ADMIN_PASSWORD=admin plik/plik
    
    # OR using Python HTTP server
    mkdir -p /tmp/fileserver
    python3 -m http.server 8080 --directory /tmp/fileserver
    ```

### Send Your First Message

#### Julia

```julia
using msghandler, NATS

# Data format: [(dataname, data, type), ...]
# Each tuple contains:
#   1. dataname: String name for the payload (e.g., "message")
#   2. data: The actual data (String, bytes, DataFrame, etc.)
#   3. type: String type indicator (e.g., "text", "binary", "dictionary")
#      Supported types: "text", "dictionary", "arrowtable", "jsontable", 
#                       "image", "audio", "video", "binary"

payload_1 = ("test_message", "Hello World", "text")

file_path_large_image = "./test/large_image.png"
file_data_large_image = read(file_path_large_image)
filename_large_image = basename(file_path_large_image)
payload_2 = (filename_large_image, file_data_large_image, "binary")

payloads = [payload_1, payload_2]  # List of tuples

# Step 1: Create the message envelope (transport-agnostic)
envelope, envelope_json_str = msghandler.smartpack("test.topic", 
                                                  payloads; 
                                                  broker_url="nats.yiem.cc",
                                                  fileserver_url="http://192.168.88.104:8080")

# Step 2: Send via your chosen transport (NATS in this example)
conn = NATS.connect("nats.yiem.cc")
NATS.publish(conn, "test.topic", envelope_json_str; reply_to="test.replytopic")
NATS.drain(conn)

# OR
reply = NATS.request("nats.yiem.cc", "test.topic", env_json_str)
```

#### JavaScript (Node.js)

```javascript
import msghandler from './src/msghandler_ssr.js';

const data = [["message", "Hello World", "text"]];
const [env, env_json_str] = await msghandler.smartpack(
    "/chat/room1",
    data,
    { broker_url: "nats://localhost:4222" }
);
console.log("Message sent!");
```

#### JavaScript (Browser)

```javascript
import msghandler from './src/msghandler_csr.js';

const data = [["message", "Hello World", "text"]];
const [env, env_json_str] = await msghandler.smartpack(
    "/chat/room1",
    data,
    { broker_url: "ws://localhost:4222" }
);
console.log("Message sent!");
```

#### Python

```python
from msghandler import smartpack

# Data format: [(dataname, data, type), ...]
# Each tuple contains:
#   1. dataname: String name for the payload (e.g., "message")
#   2. data: The actual data (str, bytes, dict, list, etc.)
#   3. type: String type indicator (e.g., "text", "binary", "dictionary")
#      Supported types: "text", "dictionary", "jsontable", "image", "audio", "video", "binary"

data = [("message", "Hello World", "text")]
env, env_json_str = await smartpack(
    "/chat/room1",
    data,
    broker_url="nats://localhost:4222",
    fileserver_url="http://localhost:8080"
)
print("Envelope created:", env)

# To send via transport (e.g., NATS, HTTP, WebSocket):
# transport_send_function("/chat/room1", env_json_str)
```

#### MicroPython

```python
from msghandler import smartpack

data = [("message", "Hello World", "text")]
env, env_json_str = smartpack(
    "/chat/room1",
    data,
    broker_url="nats://localhost:4222",
    fileserver_url="http://localhost:8080",
    size_threshold=100000
)
print("Envelope created:", env)

# To send via transport (e.g., HTTP POST, WebSocket):
# transport_send_function("/chat/room1", env_json_str)
```

### Receive Your First Message

#### Julia

```julia
using msghandler, NATS

# Step 1: Receive message from your transport (NATS in this example)
conn = NATS.connect("nats.yiem.cc")
NATS.subscribe(conn, "test.topic") do msg
    println("Received message on $(msg.subject)")
    envelope_json_str = String(msg.payload)
    
    # Step 2: Unpack the envelope (transport-agnostic)
    envelope = msghandler.smartunpack(
        envelope_json_str;
        max_retries = 5,
        base_delay = 100,
        max_delay = 5000
    )
    println(envelope.payloads[1])
end
```

---

## API Reference

### Unified API Standard

All platforms use the same input/output format for payloads.

#### Input Format for `smartpack`

**Format:** `[(dataname1, data1, type1), (dataname2, data2, type2), ...]`

**Each tuple contains 3 elements:**
1. **`dataname`** (String) - Name for the payload (e.g., `"message"`, `"config.json"`, `"avatar.png"`)
2. **`data`** (Any) - The actual data to send (type depends on `type`):
   - `text` → String
   - `dictionary` → Dict/Object
   - `arrowtable` → DataFrame/Arrow.Table
   - `jsontable` → List of dicts/Vector{Dict}
   - `image/audio/video/binary` → bytes/Uint8Array/Vector{UInt8}
3. **`type`** (String) - Payload type (required):
   - `"text"` - Plain text string
   - `"dictionary"` - JSON-serializable dictionary
   - `"arrowtable"` - Apache Arrow IPC table (Desktop only)
   - `"jsontable"` - JSON table (all platforms)
   - `"image"` - Image data (PNG, JPG)
   - `"audio"` - Audio data (WAV, MP3)
   - `"video"` - Video data (MP4, AVI)
   - `"binary"` - Generic binary data

**Important:**
- Always wrap payloads in a list, even for single payloads
- Type must be one of the supported types above
- Large payloads (≥500KB by default) automatically use link transport

**Example:**
```julia
# Julia
data = [
    ("message", "Hello World", "text"),
    ("config", Dict("key" => "value"), "dictionary"),
    ("file", file_bytes, "binary")
]
```

```python
# Python
data = [
    ("message", "Hello World", "text"),
    ("config", {"key": "value"}, "dictionary"),
    ("file", file_bytes, "binary")
]
```

```javascript
// JavaScript
const data = [
    ["message", "Hello World", "text"],
    ["config", {key: "value"}, "dictionary"],
    ["file", fileBytes, "binary"]
];
```

### Common Payload Examples

| Use Case | Data Format | Type |
|----------|-------------|------|
| Simple text message | `("message", "Hello World", "text")` | `"text"` |
| Configuration JSON | `("config", Dict("key" => "value"), "dictionary")` | `"dictionary"` |
| Table as Arrow | `("data", dataframe, "arrowtable")` | `"arrowtable"` |
| Table as JSON | `("data", [{"col": 1}], "jsontable")` | `"jsontable"` |
| Image file | `("avatar.png", read("file.png"), "image")` | `"image"` |
| Audio file | `("audio.mp3", audio_bytes, "audio")` | `"audio"` |
| Video file | `("video.mp4", video_bytes, "video")` | `"video"` |
| Binary file | `("data.bin", file_bytes, "binary")` | `"binary"` |

#### Output Format for `smartunpack`
```json
{
  "correlation_id": "...",
  "msg_id": "...",
  "timestamp": "...",
  "send_to": "...",
  "msg_purpose": "...",
  "sender_name": "...",
  "sender_id": "...",
  "receiver_name": "...",
  "receiver_id": "...",
  "reply_to": "...",
  "reply_to_msg_id": "...",
  "broker_url": "...",
  "metadata": {...},
  "payloads": [(dataname1, data1, type1), (dataname2, data2, type2), ...]
}
```

### smartpack

Sends data via your chosen transport mechanism (NATS, HTTP, WebSocket, etc.) with intelligent transport selection (direct vs URL-based) based on payload size.

#### Julia

```julia
using msghandler

env, env_json_str = msghandler.smartpack(
    subject::String,
    data::AbstractArray{Tuple{String, Any, String}};
    broker_url::String = "nats://localhost:4222",
    fileserver_url = "http://localhost:8080",
    fileserver_upload_handler::Function = plik_oneshot_upload,
    size_threshold::Int = 500_000,
    correlation_id::String = string(uuid4()),
    msg_purpose::String = "chat",
    sender_name::String = "msghandler",
    receiver_name::String = "",
    receiver_id::String = "",
    reply_to::String = "",
    reply_to_msg_id::String = "",
    is_publish::Bool = true,
    NATS_connection::Union{NATS.Connection, Nothing} = nothing,
    msg_id::String = string(uuid4()),
    sender_id::String = string(uuid4())
)
# Returns: ::Tuple{msg_envelope_v1, String}
```

#### JavaScript (Node.js)

```javascript
import msghandler from './src/msghandler_ssr.js';

const [env, env_json_str] = await msghandler.smartpack(
    subject,
    data,  // Array of [dataname, data, type] tuples
    {
        broker_url: 'nats://localhost:4222',
        fileserver_url: 'http://localhost:8080',
        fileserver_upload_handler: msghandler.plikOneshotUpload,
        size_threshold: 500_000,
        correlation_id: uuidv4(),
        msg_purpose: 'chat',
        sender_name: 'msghandler',
        receiver_name: '',
        receiver_id: '',
        reply_to: '',
        reply_to_msg_id: '',
        is_publish: true,
        nats_connection: null,
        msg_id: uuidv4(),
        sender_id: uuidv4()
    }
);
// Returns: Promise<[env, env_json_str]>
```

#### JavaScript (Browser)

```javascript
import msghandler from './src/msghandler_csr.js';

const [env, env_json_str] = await msghandler.smartpack(
    subject,
    data,
    {
        broker_url: 'ws://localhost:4222',
        fileserver_url: 'http://localhost:8080',
        fileserver_upload_handler: msghandler.plikOneshotUpload,
        size_threshold: 500_000,
        correlation_id: uuidv4(),
        msg_purpose: 'chat',
        sender_name: 'msghandler',
        receiver_name: '',
        receiver_id: '',
        reply_to: '',
        reply_to_msg_id: '',
        is_publish: true,
        nats_connection: null,
        msg_id: uuidv4(),
        sender_id: uuidv4()
    }
);
// Returns: Promise<[env, env_json_str]>
```

#### Python

```python
from msghandler import msghandler

env, env_json_str = await msghandler.smartpack(
    subject: str,
    data: List[Tuple[str, Any, str]],
    broker_url: str = "nats://localhost:4222",
    fileserver_url: str = "http://localhost:8080",
    fileserver_upload_handler: Callable = plik_oneshot_upload,
    size_threshold: int = 500_000,
    correlation_id: str = None,
    msg_purpose: str = "chat",
    sender_name: str = "msghandler",
    receiver_name: str = "",
    receiver_id: str = "",
    reply_to: str = "",
    reply_to_msg_id: str = "",
    is_publish: bool = True,
    nats_connection: Any = None,
    msg_id: str = None,
    sender_id: str = None
)
# Returns: Tuple[Dict, str]
```

#### MicroPython

```python
from msghandler import smartpack

# Limited to direct transport (< 100KB threshold)
env, env_json_str = smartpack(
    "/device/config",
    [("config", config, "dictionary")],
    broker_url="nats://localhost:4222",
    size_threshold=100000
)
print("Message sent!")
```

### smartunpack

Receives and processes messages from NATS, handling both direct and link transport.

#### Julia

```julia
using msghandler

env = msghandler.smartunpack(
    msg::NATS.Msg;
    fileserver_download_handler::Function = _fetch_with_backoff,
    max_retries::Int = 5,
    base_delay::Int = 100,
    max_delay::Int = 5000
)
# Returns: ::JSON.Object{String, Any}
```

#### JavaScript (Node.js)

```javascript
import msghandler from './src/msghandler_ssr.js';

const env = await msghandler.smartunpack(
    msg,
    {
        fileserver_download_handler: msghandler.fetchWithBackoff,
        max_retries: 5,
        base_delay: 100,
        max_delay: 5000
    }
);
// Returns: Promise<env_object>
```

#### JavaScript (Browser)

```javascript
import msghandler from './src/msghandler_csr.js';

const env = await msghandler.smartunpack(
    msg,
    {
        fileserver_download_handler: msghandler.fetchWithBackoff,
        max_retries: 5,
        base_delay: 100,
        max_delay: 5000
    }
);
// Returns: Promise<env_object>
```

#### Python

```python
from msghandler import msghandler

env = await msghandler.smartunpack(
    msg,
    fileserver_download_handler=fetch_with_backoff,
    max_retries=5,
    base_delay=100,
    max_delay=5000
)
# Returns: Dict with "payloads" key
```

#### MicroPython

```python
from msghandler import smartunpack

env = smartunpack(
    msg,
    fileserver_download_handler=_sync_fileserver_download,
    max_retries=3,
    base_delay=100,
    max_delay=1000
)
print("Received payloads:", env["payloads"])
```

---

## Cross-Platform Examples

### Example 1: Chat with Mixed Content

Send text, image, and large file in one message.

#### Julia

```julia
using msghandler

data = [
    ("message_text", "Hello!", "text"),
    ("user_avatar", image_data, "image"),
    ("large_document", large_file_data, "binary")
]

env, env_json_str = smartpack("/chat/room1", data; fileserver_url="http://localhost:8080")
```

#### JavaScript (Node.js)

```javascript
import msghandler from './src/msghandler_ssr.js';

const data = [
    ["message_text", "Hello!", "text"],
    ["user_avatar", imageData, "image"],
    ["large_document", largeFileData, "binary"]
];

const [env, env_json_str] = await msghandler.smartpack(
    "/chat/room1",
    data,
    { fileserver_url: 'http://localhost:8080' }
);
```

#### JavaScript (Browser)

```javascript
import msghandler from './src/msghandler_csr.js';

const data = [
    ["message_text", "Hello!", "text"],
    ["user_avatar", image_data, "image"]
];

const [env, env_json_str] = await msghandler.smartpack(
    "/chat/room1",
    data,
    { broker_url: 'ws://localhost:4222' }
);
console.log("Message sent!");
```

#### Python

```python
import asyncio
from msghandler import smartpack

async def main():
    image_data = open("user_avatar.png", "rb").read()
    large_file_data = open("large_document.pdf", "rb").read()

    data = [
        ("message_text", "Hello!", "text"),
        ("user_avatar", image_data, "image"),
        ("large_document", large_file_data, "binary")
    ]

    env, env_json_str = await smartpack(
        "/chat/room1",
        data,
        fileserver_url="http://localhost:8080"
    )
    print("Message sent!")

if __name__ == "__main__":
    asyncio.run(main())
```

#### MicroPython

```python
from msghandler import smartpack

# Note: MicroPython only supports direct transport (< 100KB threshold)
# Large files must be uploaded via file server first
data = [
    ("message_text", "Hello!", "text"),
    ("user_avatar", image_bytes, "image")
]

env, env_json_str = smartpack(
    "/chat/room1",
    data,
    broker_url="nats://localhost:4222",
    size_threshold=100000
)
print("Message sent!")
```

### Example 2: Dictionary Exchange

Send configuration data between platforms.

#### Julia

```julia
using msghandler

config = Dict(
    "wifi_ssid" => "MyNetwork",
    "wifi_password" => "password123",
    "update_interval" => 60
)

data = [("config", config, "dictionary")]
env, env_json_str = smartpack("/device/config", data)
```

#### JavaScript (Node.js)

```javascript
import msghandler from './src/msghandler_ssr.js';

const config = {
    wifi_ssid: "MyNetwork",
    wifi_password: "password123",
    update_interval: 60
};

const [env, env_json_str] = await msghandler.smartpack(
    "/device/config",
    [["config", config, "dictionary"]]
);
console.log("Message sent!");
```

#### Python

```python
import asyncio
from msghandler import smartpack

async def main():
    config = {
        "wifi_ssid": "MyNetwork",
        "wifi_password": "password123",
        "update_interval": 60
    }

    data = [("config", config, "dictionary")]
    env, env_json_str = await smartpack("/device/config", data)
    print("Message sent!")

if __name__ == "__main__":
    asyncio.run(main())
```

#### MicroPython

```python
from msghandler import smartpack

config = {
    "wifi_ssid": "MyNetwork",
    "wifi_password": "password123",
    "update_interval": 60
}

data = [("config", config, "dictionary")]
env, env_json_str = smartpack(
    "/device/config",
    data,
    broker_url="nats://localhost:4222",
    size_threshold=100000
)
print("Message sent!")
```

### Example 3: Table Data (Arrow IPC)

Send tabular data using Apache Arrow IPC format.

#### Julia

```julia
using msghandler
using DataFrames

df = DataFrame(
    id = [1, 2, 3],
    name = ["Alice", "Bob", "Charlie"],
    score = [95, 88, 92]
)

data = [("students", df, "arrowtable")]
env, env_json_str = smartpack("/data/analysis", data)
```

#### JavaScript (Node.js)

```javascript
import msghandler from './src/msghandler_ssr.js';

const df = [
    { id: 1, name: "Alice", score: 95 },
    { id: 2, name: "Bob", score: 88 },
    { id: 3, name: "Charlie", score: 92 }
];

const [env, env_json_str] = await msghandler.smartpack(
    "/data/analysis",
    [["students", df, "arrowtable"]]
);
```

#### Python

```python
from msghandler import msghandler
import pandas as pd

df = pd.DataFrame({
    "id": [1, 2, 3],
    "name": ["Alice", "Bob", "Charlie"],
    "score": [95, 88, 92]
})

data = [("students", df, "arrowtable")]
env, env_json_str = await msghandler.smartpack("/data/analysis", data)
```

#### JavaScript (Browser)

```javascript
import msghandler from './src/msghandler_csr.js';

const df = [
    { id: 1, name: "Alice", score: 95 },
    { id: 2, name: "Bob", score: 88 },
    { id: 3, name: "Charlie", score: 92 }
];

const [env, env_json_str] = await msghandler.smartpack(
    "/data/analysis",
    [["students", df, "jsontable"]],
    { broker_url: 'ws://localhost:4222' }
);
console.log("Message sent!");
```

#### MicroPython

```python
from msghandler import smartpack

# Note: MicroPython only supports direct transport (< 100KB threshold)
# MicroPython doesn't support Arrow IPC, only jsontable
df = [
    {"id": 1, "name": "Alice", "score": 95},
    {"id": 2, "name": "Bob", "score": 88},
    {"id": 3, "name": "Charlie", "score": 92}
]

data = [("students", df, "jsontable")]
env, env_json_str = smartpack(
    "/data/analysis",
    data,
    broker_url="nats://localhost:4222",
    size_threshold=100000
)
print("Message sent!")
```

### Example 4: Request-Response Pattern

Bi-directional communication with reply-to support.

#### Julia

```julia
using msghandler, NATS

# Requester
env, env_json_str = msghandler.smartpack(
    "/device/command",
    [("command", Dict("action" => "read_sensor"), "dictionary")];
    broker_url="nats://localhost:4222",
    reply_to="/device/response"
)

conn = NATS.connect("nats://localhost:4222")
NATS.publish(conn, "/device/command", env_json_str)
NATS.drain(conn)

# Receiver (in separate application)
conn = NATS.connect("nats://localhost:4222")
NATS.subscribe(conn, "/device/command") do msg
    env = msghandler.smartunpack(msg)
    println("Received command: ", env["payloads"])
    
    result = Dict("value" => 42)
    response_env, response_json = msghandler.smartpack(
        "/device/response",
        [("result", result, "dictionary")],
        reply_to="/device/command",
        reply_to_msg_id=env["msg_id"]
    )
    
    NATS.publish(conn, "/device/response", response_json)
    NATS.drain(conn)
end
```

#### JavaScript (Node.js)

```javascript
import msghandler from './src/msghandler_ssr.js';
import { connect } from 'nats';

// Requester
const [env, env_json_str] = await msghandler.smartpack(
    "/device/command",
    [["command", { action: "read_sensor" }, "dictionary"]],
    { broker_url: 'nats://localhost:4222', reply_to: '/device/response' }
);
const nc = await connect({ port: 4222 });
nc.publish("/device/command", env_json_str);
await nc.flush();

// Receiver (in separate application)
const nc = await connect({ port: 4222 });
const sub = nc.subscribe("/device/command");
for await (const msg of sub) {
    const env = await msghandler.smartunpack(msg);
    console.log("Received command:", env.payloads);
    
    const response_env, response_json = await msghandler.smartpack(
        "/device/response",
        [["result", { value: 42 }, "dictionary"]],
        { reply_to: '/device/command', reply_to_msg_id: env.msg_id }
    );
    
    nc.publish("/device/response", response_json);
    await nc.flush();
}
```

#### Python

```python
import asyncio
from msghandler import smartpack

async def main():
    # Requester
    env, env_json_str = await smartpack(
        "/device/command",
        [("command", {"action": "read_sensor"}, "dictionary")],
        broker_url="nats://localhost:4222",
        reply_to="/device/response"
    )
    print("Request sent!")
    
    # Receiver (in separate application)
    # await nats_consumer.next()
    # env = await smartunpack(msg)
    # Process request and send response
    # response_env, response_json = await smartpack(
    #     "/device/response",
    #     [("result", {"value": 42}, "dictionary")],
    #     reply_to="/device/command",
    #     reply_to_msg_id=env["msg_id"]
    # )

if __name__ == "__main__":
    asyncio.run(main())
```

#### JavaScript (Browser)

```javascript
import msghandler from './src/msghandler_csr.js';

// Requester
const [env, env_json_str] = await msghandler.smartpack(
    "/device/command",
    [["command", { action: "read_sensor" }, "dictionary"]],
    { broker_url: 'ws://localhost:4222', reply_to: '/device/response' }
);
console.log("Request sent!");

// Receiver (in separate application)
// const msg = await natsConsumer.next();
// const env = await msghandler.smartunpack(msg);
// console.log("Received command:", env.payloads);
// const response_env, response_json = await msghandler.smartpack(
//     "/device/response",
//     [["result", { value: 42 }, "dictionary"]],
//     { reply_to: '/device/command', reply_to_msg_id: env.msg_id }
// );
```

#### MicroPython

```python
from msghandler import smartpack

# Requester
env, env_json_str = smartpack(
    "/device/command",
    [("command", {"action": "read_sensor"}, "dictionary")],
    broker_url="nats://localhost:4222",
    reply_to="/device/response",
    size_threshold=100000
)
print("Request sent!")
```

---

## Testing

### Test File Organization

| Platform | Sender Tests | Receiver Tests |
|----------|--------------|----------------|
| **Julia** | `test/test_julia_*_sender.jl` | `test/test_julia_*_receiver.jl` |
| **JavaScript** | `test/test_js_*_sender.js` | `test/test_js_*_receiver.js` |
| **Python** | `test/test_py_*_sender.py` | `test/test_py_*_receiver.py` |

### Run Tests

#### Julia

```bash
# Text message exchange
julia test/test_julia_text_sender.jl
julia test/test_julia_text_receiver.jl

# Dictionary exchange
julia test/test_julia_dict_sender.jl
julia test/test_julia_dict_receiver.jl

# File transfer
julia test/test_julia_file_sender.jl
julia test/test_julia_file_receiver.jl

# Mixed payload types
julia test/test_julia_mix_payloads_sender.jl
julia test/test_julia_mix_payloads_receiver.jl

# Table exchange
julia test/test_julia_table_sender.jl
julia test/test_julia_table_receiver.jl
```

#### JavaScript (Node.js)

```bash
# Text message exchange
node test/test_js_text_sender.js
node test/test_js_text_receiver.js

# Dictionary exchange
node test/test_js_dictionary_sender.js
node test/test_js_dictionary_receiver.js

# Binary transfer
node test/test_js_binary_sender.js
node test/test_js_binary_receiver.js

# Table exchange
node test/test_js_table_sender.js
node test/test_js_table_receiver.js
```

#### Python

```bash
# Text message exchange
python3 test/test_py_text_sender.py
python3 test/test_py_text_receiver.py

# Dictionary exchange
python3 test/test_py_dictionary_sender.py
python3 test/test_py_dictionary_receiver.py

# Binary transfer
python3 test/test_py_binary_sender.py
python3 test/test_py_binary_receiver.py

# Table exchange
python3 test/test_py_table_sender.py
python3 test/test_py_table_receiver.py
```

---

## Browser Deployment

### Using with Node.js Build Tools

The browser implementation (`src/msghandler_csr.js`) can be bundled for production deployment using modern JavaScript build tools.

#### Prerequisites

```bash
# Install the browser-compatible NATS client
npm install nats.ws
```

#### Vite (Recommended)

```bash
npm create vite@latest my-app -- --template vanilla
cd my-app
npm install nats.ws
```

In `vite.config.js`:
```javascript
import { defineConfig } from 'vite';
export default defineConfig({
  resolve: {
    alias: {
      'nats.ws': 'nats.ws/dist/esm/browser.js'
    }
  }
});
```

Build command:
```bash
npm run build  # Outputs to dist/ folder
```

#### Webpack

```bash
npm install webpack webpack-cli --save-dev
npm install nats.ws
```

In `webpack.config.js`:
```javascript
module.exports = {
  entry: './src/index.js',
  output: {
    filename: 'bundle.js',
    path: __dirname + '/dist'
  },
  resolve: {
    alias: {
      'nats.ws': 'nats.ws/dist/esm/browser.js'
    }
  }
};
```

Build command:
```bash
npx webpack
```

#### esbuild (Simple & Fast)

```bash
npm install esbuild nats.ws --save-dev
```

Create `build.js`:
```javascript
import esbuild from 'esbuild';

esbuild.buildSync({
  entryPoints: ['src/msghandler_csr.js'],
  bundle: true,
  outfile: 'dist/msghandler-csr-bundle.js',
  format: 'esm',
  platform: 'browser',
  target: 'es2020'
});
```

Build command:
```bash
node build.js
```

### Using in Your HTML

```html
<!DOCTYPE html>
<html>
<head>
  <title>My App</title>
</head>
<body>
  <script type="module" src="dist/msghandler-csr-bundle.js"></script>
  <script type="module">
    import msghandlerCSR from './dist/msghandler-csr-bundle.js';
    
    // Use the library
    const [env, envJson] = await msghandlerCSR.smartpack(
      "/chat/user/v1/message",
      [["msg", "Hello", "text"]],
      { broker_url: "wss://nats.example.com" }
    );
  </script>
</body>
</html>
```

---

## Documentation

For detailed architecture and implementation information, see:

- [`docs/architecture.md`](docs/architecture.md) - Cross-platform architecture, API parity, platform-specific patterns
- [`docs/requirements.md`](docs/requirements.md) - Business requirements and user stories
- [`docs/spec.md`](docs/spec.md) - Technical specification and contracts
- [`docs/walkthrough.md`](docs/walkthrough.md) - Real-world application building guides

---

## License

MIT License

Copyright (c) 2026 msghandler Contributors

Permission is hereby granted, free of charge, to any person obtaining a copy
of this software and associated documentation files (the "Software"), to deal
in the Software without restriction, including without limitation the rights
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
copies of the Software, and to permit persons to whom the Software is
furnished to do so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all
copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
SOFTWARE.