Here's an example of what the DSL should look like:
graph = buildGraph { bgColor: Yellow } [
node "a" { color: Red },
node "b" { color: Green },
node "c" {},
edge "a" "b" { color: Red, style: Dotted },
edge "b" "c" { color: Blue },
edge "a" "c" { color: Green },
]This code should build a graph with a yellow background, and three nodes, "a", "b", and "c", respectively colored red, green, and black (which is the default color). They should be connected by 3 edges of different colors and styles: the edge between "a" and "b" should be red and dotted, and the edges between "b" and "c" and between "a" and "c" should be solid (which is the default style) and green.
The node function should simply create an AddNode value with the arguments as payload. This is a DSL command used only by the buildGraph function. For example, node "a" { color: Red } should return AddNode "a" { color: Red } . If an attribute is missing, its default value should be used. For example, node "c" {} should return AddNode "c" { color: Black }.
Similarly, the edge function should create an AddEdge value. For example, edge "a" "b" {} should return AddEdge "a" "b" {color: Default, style: Solid}.
These two simple functions make the DSL code much more pleasant to read & write.
Once you have implemented the node and edge functions (they should be easy), your main goal is to write the buildGraph function: it must go through the list of DSL commands and produce the desired graph, represented as a record { bgColor: ..., nodes: ..., edges: ...}.
To double the fun, you can optionally try to implement a toDot function that converts the graph to a Str with using the Dot format!