A Domain-Specific Language (DSL) is a computer language targeted at a particular kind of problem, rather than a general-purpose language aimed at any kind of software problem.

Domain-specific languages have been discussed and used for almost as long as computing has existed. In the data space, they are mostly used with YAML. For example, Dagster uses a DSL to declaratively create a pipeline/graph (docs): ^8853ef

Sometimes, you may want to construct the dependencies of a graph definition from a YAML file or similar. This is useful when migrating to Dagster from other workflow systems.

For example, you can have a YAML like this:

 1 2 3 4 5 6 7 8 9101112131415161718192021

name: some_exampledescription: blah blah blahops:  - def: add_one    alias: A  - def: add_one    alias: B    deps:      num:        op: A  - def: add_two    alias: C    deps:      num:        op: A  - def: subtract    deps:      left:        op: B      right:        op: C

A DSL is a good way to empower less Pythonic users, such as data analysts. It provides a declarative approach to building data pipelines.

A comprehensive write-up by Elliot Gunn and Tim Castillo on Standardize Pipelines with Domain-Specific Languages | Dagster Blog offers valuable insights.

Knotty: A domain-specific language for knitting patterns.

Origin:
References:
Created 2022-09-22