Overview

The DataPacket class (/inc/Core/DataPacket.php) provides standardized data packet creation and management across the entire pipeline system. Introduced in version 0.2.1, it replaces scattered array construction with centralized, consistent data packet handling.

Architecture

Location: /inc/Core/DataPacket.php Since: 0.2.1

Purpose

Data packets standardize the data contract between pipeline steps and the execution engine. They ensure consistent data structure and workflow processing across all handlers and steps.

Constructor

public function __construct(array $data, array $metadata, string $type)

Parameters

• $data (array): Content data (title, body, file_info, etc.)
• $metadata (array): Metadata (source_type, timestamps, etc.)
• $type (string): Packet type (fetch, ai_response, tool_result, etc.)

addTo() Method

Adds the packet to the data packets array, maintaining pipeline workflow order:

public function addTo(array $dataPackets): array

Workflow Processing

The method adds packets to the front of the array so the most recent contributions are easily accessible by subsequent steps:

// Processing flow: Step 1 → Step 2 → Step 3
$dataPackets = [];
$dataPackets = $step1_packet->addTo($dataPackets); // [packet1]
$dataPackets = $step2_packet->addTo($dataPackets); // [packet2, packet1]
$dataPackets = $step3_packet->addTo($dataPackets); // [packet3, packet2, packet1]

// Next step sees packet3 first (most recent)

Data Structure

Complete Packet Structure

[
    'type' => 'fetch',        // Packet type
    'timestamp' => 1234567890, // Unix timestamp
    'data' => [
        'content_string' => 'Content text...',
        'file_info' => [
            'filename' => 'image.jpg',
            'path' => '/path/to/file',
            'url' => 'https://example.com/file.jpg'
        ]
    ],
    'metadata' => [
        'source_type' => 'wordpress',
        'item_identifier_to_log' => 'post_123',
        'original_id' => '123',
        'original_title' => 'Post Title',
        'original_date_gmt' => '2024-01-01 12:00:00'
    ]
]

Usage Examples

Fetch Handler

use DataMachineCoreDataPacket;

$dataPacket = new DataPacket(
    ['content_string' => $content, 'file_info' => null],
    [
        'source_type' => 'rss',
        'item_identifier_to_log' => $item_id,
        'original_title' => $title,
        'original_date_gmt' => $date
    ],
    'fetch'
);

return $this->successResponse([$dataPacket->addTo([])]);

AI Step

$dataPacket = new DataPacket(
    ['content_string' => $ai_generated_content],
    ['source_type' => 'ai_response'],
    'ai_response'
);

return $dataPacket->addTo($this->dataPackets);

Tool Result

$dataPacket = new DataPacket(
    ['tool_result' => $tool_output],
    ['tool_name' => 'web_search', 'query' => $query],
    'tool_result'
);

return $dataPacket->addTo($this->dataPackets);

Data vs Metadata

Data Section

Contains the actual content that AI agents and handlers work with:

• content_string: Clean text content (no URLs)
• file_info: File metadata when applicable
• tool_result: Tool execution results
• Custom handler-specific data

Metadata Section

Contains contextual information for logging and processing:

• source_type: Handler type (rss, wordpress, twitter, etc.)
• item_identifier_to_log: Unique identifier for deduplication
• original_*: Original source data (title, date, id, etc.)
• Handler-specific metadata

Engine Data Separation

Data packets contain clean data for AI processing, while engine parameters are stored separately in the database:

// Clean data in packet (AI-visible)
'data' => [
    'content_string' => 'Clean content without URLs'
]

// Engine data in database (handler-accessible)
[
    'source_url' => 'https://source.com/post/123',
    'image_url' => 'https://source.com/image.jpg'
]

Benefits

• Consistency: Standardized data structure across all pipeline components
• Type Safety: Defined packet types and data structure
• Workflow Clarity: Clear separation of data and metadata
• Maintainability: Centralized packet creation logic
• Debugging: Structured data for logging and inspection

Usage Examples

DataPacket is used throughout the codebase for standardized packet creation:

Fetch Handlers: See FetchHandler implementations:

• RSS Handler
• Reddit Handler
• Files Handler

AI Steps: For conversation result packaging

Any Step: Any component creating or modifying data packets

All handlers extending FetchHandler automatically use DataPacket for packet creation.