The RAML community has announced the release of RAML 1.0, API Workbench and a JavaScript plus a Java RAML parser.
RAML 1.0 is a Release Candidate but it is close to final, the last touches being done this month. The Specification defines the language and its features. Compared to 0.8, the new features are:
- Data types – the ability to describe the data. It supports built-in types –string, number, integer, boolean, date, file - or inherited ones. It understands enums, arrays, maps, unions, and complex data types can be built based on these.
- Libraries – the ability to include predefined resources, data types, assets, and others, using namespaces
- Overlays – including descriptions, annotations, and examples from other RAML files
- Annotations – supporting third-party metadata beyond what’s specified in RAML 1.0
- Improved security schemas
RAML 1.0 is accompanied by an Atom-based IDE called API Workbench. The IDE can be used to design, build, test, document and share APIs defined with RAML. It works both with RAML 0.8 and 1.0. API Workbench supports code completion, suggesting keywords, references, paths, common values, type properties, etc. The IDE performs automatic validation at several levels: structure, reference, resource, schema, or line. New elements can be created with the help of wizards.
MuleSoft has also contributed two parsers, one for JavaScript/TypeScript and another for Java. They are currently in beta. Other parsers including – .NET, Go, JavaScript, PHP, Python, Ruby, Scala – and other tools can be found at various stages at development in the RAML Projects Library.
RAML is being backed by a number of vendors involved in the creation, utilization and management of APIs, such as AWS, CA Technologies, Cisco, Software AG, VMware, and, of course, MuleSoft.