70 %
Chris Biscardi

Rendering Trees with SlateJS

SlateJS can render arbitrary elements as long as the leaf nodes are text. This means we can construct arbitrary JSON trees of to represent our content and let Slate handle the rest.

JS
import React, { useMemo } from "react";
import ReactDOM from "react-dom";
// Import the Slate editor factory.
import { createEditor } from "slate";
// Import the Slate components and React plugin.
import { Slate, Editable, withReact } from "slate-react";
const defaultValue = [
{
type: "stuff",
children: [
{
text: "A line of text in a paragraph.",
marks: []
}
]
}
];
const App = () => {
const editor = useMemo(
() => withReact(createEditor()),
[]
);
return (
// Add the default value as a prop to the editor context.
<Slate editor={editor} defaultValue={defaultValue}>
<Editable />
</Slate>
);
};
const rootElement = document.getElementById("root");
ReactDOM.render(<App />, rootElement);

This works because the default React element used to render a Slate element is a div. Note the contents of the data-slate-node attributes in the resulting HTML.

html
<div
data-gramm="false"
role="textbox"
data-slate-editor="true"
data-slate-node="value"
contenteditable="true"
style="outline: none; white-space: pre-wrap; overflow-wrap: break-word;"
>
<div
data-slate-node="element"
style="position: relative;"
>
<span data-slate-node="text"
><span data-slate-leaf="true"
><span data-slate-string="true"
>A line of text in a paragraph.</span
></span
></span
>
</div>
</div>

The type field isn't special, we could name it whatever we wanted. The children field is important and needs to exist. This is part of the Node interfaces.

the tldr; is that your elements need to have children and your Text nodes need to have text and marks. Everything else is up to you from how to structure your tree to anything else relating to the keys that make up that tree.

GraphQL
interface Element {
children: Node[]
}
interface Text {
text: string
marks: Mark[]
}