structurizr-parser

Parses Structurizr DSL files to support workspace population, code validation and other uses

This is BETA software and should be used with caution as it is not feature complete or fully tested

The library is broken up into three parts. Each should be executed in turn depending on the outcome you are expecting.

Lexer

This converts the input file into a set of tokens. Errors in tokenization are emitted alongside the tokens found.

let lexingResult = StructurizrLexer.tokenize(dsl);

• lexingResult.errors is an array of errors
• lexingResult.tokens is an array of discovered tokens

Parser

This takes the output of the Lexer (the set of tokens) and applies the structurizr DSL rules. This will result in a syntax tree that represents what was extracted from the tokens using the structurizr grammer asa guide. Any errors are also emitted.

StructurizrParser.input = lexingResult.tokens;
const cst = StructurizrParser.workspaceWrapper();

• StructurizrParser.errors is an array of errors
• cst is the syntax tree generated by the parser

Interpreter

This last step will take the syntax tree produced by the Parser and execute some operations to produce a given desired outcome. At present the Structurizr Interpreter tries to convert the syntax tree in structurzr class. The new C4 Interpreter produces at the moment a light weight C4 class. It is possible that the better approach would be for the interpreter to directly generate on output like a draw.io (MX) file for graphical rendering.

Produce a structurizr class

const wspace = StructurizrInterpreter.visit(cst) as Workspace;

Produce a light weight C4 Class

const c4wspace = C4Interpreter.visit(cst) as C4Workspace;

Future Work

It is for future work to see if we will continue to support multiple interpreters designed to produce final outputs like the Structurizr class, the new C4 class, or an MX file or a PlantUML file or whatever.