[3] IMPLEMENTATION ACCELERATORS: Automated Code & Catalog Conversion Tools

updated on 22 February 2024

Introduction

Automated code and catalog conversion tools are only one small part of a data platform migration. However, many project team members start here, because they are excited as the prospect of dramatically reducing migration effort and costs. 

Below is a summary of some specialized code conversion tools to consider as part of a wider implementation or migration strategy.

Legacy Catalog Conversion to Databrick's Unity Catalog

T1 has developed a flexible accelerator for migration of legacy tables and catalogs like the Hive Metastore to Unity Catalog. 

One insurance company had the following activities and outcomes when using the accelerator:

  • Converted 200+ jobs, 500+ tables and 20 departments to the most current version of Unity Catalog 
  • Remapped level 2 structure to the new 3 level catalog structure design
  • Automatically mapped all environments to allow for automated deployment and promotion of jobs between development, user acceptance testing and production

SAS to Python Transpiler

Alchemist Transpiler provides automatic conversion of SAS code to python. The tool understands SAS metadata, which allows it to understand the context of what each block of SAS code is intending to do. This allows Alchemist to generate easy to understand and maintain python code. In cases where the transpiler is unsure about the optimal conversion, suggestions are provided. 

The transpiler works with a cohesive testing harness, which works in conjunction with Alchemist "templates". Templates are used in situations where manual code conversion is required. This provides a consistent set of tools and process for testing and re-testing, whether the code being tested was created by Alchemist alone, or with intervention.

Read more