View source for Martin Fowler (software engineer)

== Domain-specific languages ==
Fowler’s ''Domain-specific languages'' discusses small, composable programming languages focused on an individual domain.<ref name=":0">{{Cite book|last=Fowler|first=Martin|title=Domain-specific languages|date=2011|publisher=Addison-Wesley|isbn=978-0-13-210754-9|location=Upper Saddle River, N.J.|oclc=686709295}}</ref>{{Rp|27}}  He argues that domain-specific languages increase productivity by removing the need for the programmer to understand a full programming language, by improving communication with [[domain expert]]s, and by separating the execution of a task from its definition.<ref name=":0" />{{Rp|33}} These benefits are set against the cost of learning a new language and building the tools for this language.<ref name=":0" />{{Rp|39}}

Fowler introduces the concept of '''internal''' (or '''embedded''') and '''external''' domain-specific languages. An internal language is a subset of another language and can be executed by the tools for this outer language. [[Ruby (programming language)|Ruby]] and [[Lisp (programming language)|Lisp]] are given as an example of languages where internal domain-specific languages are common. Fowler also introduces the idea of '''Semantic Model'''.<ref name=":0" />{{Rp|14}} Various domain-specific languages are presented including [[graphviz]], a language for specifying [[graph (discrete mathematics)|graphs]] to be rendered; JMock, a java mocking framework; [[CSS]], a language to specify stylistic elements of a website; [[Hibernate (framework)#Hibernate Query Language (HQL)|HQL]], an [[object relational mapper]]; [[XAML]], a language used to specify and change [[graphical user interface]]s; FIT, a language to express testing scenarios; and [[Make (software)|make]], a tool to build software<ref name=":0" />{{Rp|147}}

The book discusses implementing an external domain-specific language using [[Parsing|parsers]], [[lexers]], [[abstract syntax tree]]s and [[Code generation (compiler)|code generation]] referred to as "syntax-driven translation"<ref name=":0" />{{Rp|219}} This is contrasted with "delimiter-driven translation" which is simpler but less powerful. Here the language is simple enough to be [[Interpreter (computing)|interpreted]] by splitting on delimiters and switching logic based on individual entries.<ref name=":0" />{{Rp|201}}

Ways of implementing internal domain-specific languages are discussed, with attention to nested function calls,<ref name=":0" />{{Rp|357}} sequences of function calls,<ref name=":0" />{{Rp|351}} or [[method chaining]]<ref name=":0" />{{Rp|373}}.{{fact|date=March 2025}}