Principle:Sdv dev SDV Metadata Loading
| Knowledge Sources | |
|---|---|
| Domains | Data_Science, Schema_Management |
| Last Updated | 2026-02-14 00:00 GMT |
Overview
A schema deserialization mechanism that reconstructs a Metadata object from a previously saved JSON file for reproducible data synthesis workflows.
Description
Metadata loading provides the ability to reconstruct a Metadata object from a saved JSON file. This is essential for reproducibility: once a metadata schema has been defined and validated, it can be saved to JSON and loaded in future sessions to ensure consistent synthesis behavior. The loader handles both current multi-table format and legacy single-table format with automatic conversion.
Usage
Use metadata loading when working with previously defined schemas, when sharing metadata across team members, or when deploying synthesis pipelines that require a fixed schema definition.
Theoretical Basis
- JSON deserialization: Read the metadata dictionary from a JSON file
- Version detection: Check METADATA_SPEC_VERSION to determine format
- Format migration: Convert legacy single-table format to unified format if needed
- Object construction: Build Metadata instance from the dictionary
Related Pages
Implemented By
Page Connections
Double-click a node to navigate. Hold to expand connections.
Principle
Implementation
Heuristic
Environment