15.1.4 Core Concepts

SDK objects map one-to-one with IDMP product concepts. Understanding these mappings helps you quickly locate the API you need.

Object Mapping

SDK Class / Module	IDMP Concept	Description
ApiClient	—	SDK entry point; manages connection, authentication, and request dispatch
ElementResourceApi	Element	Nodes in the asset tree: equipment, systems, areas
AttributeResourceApi	Attribute	Named properties of an element; can be bound to time-series data or static values
MetricResourceApi	Metric	Time-series data stream from an attribute's historical and real-time data
EventResourceApi	Event	Alarm or state-change records triggered by analysis rules
PanelResourceApi	Panel	Visualization chart associated with an element
UserResourceApi	User	User management and authentication
UomResourceApi	UOM	Unit of measurement classes and conversions

Data Access Hierarchy

IDMP SDK data access follows this hierarchy:

Element
  └─ Attribute
       └─ Time-Series Data (Metric)

Typical data read flow:

1. Use ElementResourceApi to find the target element (by name, path, or ID).
2. Use the element ID to query its attribute list (AttributeResourceApi).
3. Use the attribute ID to query time-series data (MetricResourceApi).

Pagination

All list endpoints support pagination. The response format is:

{
  "data": [...],        // records on the current page
  "total": 100,         // total record count
  "pageNum": 1,         // current page number (1-based)
  "pageSize": 20        // records per page
}

Control pagination with the pageNum and pageSize query parameters:

# Python example
result = element_api.api_v1_elements_get(page_num=1, page_size=50)

Request and Response Structure

All API responses follow a consistent format:

{
  "code": 0,            // 0 = success; non-zero = error
  "message": "success", // status description
  "data": { ... }       // actual response payload
}

When code is non-zero, the SDK raises an ApiException. See Error Handling for details.

Time Format

All time parameters and return values use Unix timestamps in milliseconds (UTC).

import time

# Query the last 1 hour of data
now_ms = int(time.time() * 1000)
one_hour_ago_ms = now_ms - 3600 * 1000