drakongen

drakongen generates pseudocode from drakon flowcharts and mind-maps built with DrakonWidget or DrakonHub.

One can use the output pseudocode as a prompt for AI applications like ChatGPT, Grok, or Gemini.

drakongen can also generate a generic AST from drakon flowcharts to generate source code in programming languages, for example, JavaScript. drakongen turns a drakon flowchart into a tree in the JSON format that one can trivially transform into code in any programming language.

How to use drakongen in the browser

Include the drakongen.js script on the web page.

<script src="browser/drakongen.js"></script>

Call drakongen.toPseudocode, drakongen.toMindTree, or drakongen.toTree functions.

<script>
var drakon = ... // Get the drakon chart from DrakonWidget
var mindJson = ... // Graf mind-map as json
var name = "Diagram one"
var filename = "Diagram one.drakon"

var pseudo = drakongen.toPseudocode(drakon, name, filename, "en")
console.log(pseudo)

var mind = drakongen.toMindTree(mindJson, name, filename)
console.log(mind)

var tree = drakongen.toTree(drakon, name, filename, "en")
console.log(tree)

</script>

Available languages:

• en English
• no Norwegian
• ru Russian

How to use drakongen in Node JS

Add the drakongen package to dependencies.

npm i drakongen

• Require the drakongen module.
• Call toPseudocode, toMindTree, or toTree functions.

const {toTree, toPseudocode} = require("drakongen")

var drakon = ... // Get the drakon chart from DrakonWidget
var mindJson = ... // Graf mind-map as json
var name = "Diagram one"
var filename = "Diagram one.drakon"

var pseudo = toPseudocode(drakon, name, filename, "en")
console.log(pseudo)

var mind = toMindTree(mindJson, name, filename)
console.log(mind)

var tree = toTree(drakon, name, filename, "en")
console.log(tree)

How to use drakongen as a command-line tool

Install drakongen globally.

npm i --global drakongen

Run the drakongen utility.

Generate from one file

A .drakon flowchart will be converted to pseudocode.

A .graf mind-map will be converted to tree-like text.

In this example, the language is Norwegian, the path to the output folder is out, the input file is Hello.drakon:

drakongen --language no --output out Hello.drakon

Project mode

Drakongen in project mode generates a single prompt file by processing and concatenating multiple input files, including flowcharts, mind-maps, text files, and directories. It reads a project file specifying the input files, processes them in the defined order, and outputs a consolidated prompt file suitable for AI-driven code generation.

Features

• Processes multiple file types:
  • Plain text files (e.g., .txt) are included as-is.
  • Mind-map files (e.g., .graf) are converted to indented text.
  • Flowchart files (e.g., .drakon) are converted to pseudocode.
  • Directories are recursively processed, including all files within.
• Preserves file order: Files are processed and concatenated in the exact order specified in the project file, which is critical for AI tools relying on sequential context.
• Flexible project file format: A simple text-based project file lists files and directories to process.

Usage

Run the tool with the following command:

drakongen --project <project_file> --output <output_directory>

• --project: Path to the project file (e.g., exampleproject/foo.proj).
• --output: Directory where the output prompt file will be saved (e.g., exampleproject).

Example

Given a project file exampleproject/foo.proj with the following content:

start.txt
math
class Point.graf
foo.drakon

Run:

drakongen --project exampleproject/foo.proj --output exampleproject

This will:

1. Include start.txt as-is.
2. Process all files in the math directory recursively.
3. Convert class Point.graf (mind-map) to indented text.
4. Convert foo.drakon (flowchart) to pseudocode.
5. Concatenate all processed content into exampleproject/foo.txt.

The output file foo.txt will contain the concatenated text in the exact order specified in foo.proj.

Project File Format

The project file (e.g., foo.proj) is a plain text file where each line specifies:

• A file path (relative to the project file) for a text, mind-map, or flowchart file.
• A directory path to process all files within it recursively.