Your lightweight open-source personal agent.
Python agent-frameworkai-agentai-agentschatbotchatopsdiscord-botllm-agentsllmslocal-llmmcpmodel-context-protocolmulti-agentopenai-compatibleopenclawpersonal-ai-assistantpythonself-hostedtelegram-bot-ai-assistantwebuiworkflow-automation
import { useCallback, useLayoutEffect, useRef, useState, type ReactNode } from "react";
import "./DisplayMath.css";
export function DisplayMath({ children }: { children: ReactNode }) {
const frame = useRef<HTMLSpanElement>(null);
const content = useRef<HTMLSpanElement>(null);
const [layout, setLayout] = useState({ scale: 1, height: 0 });
const update = useCallback(() => {
if (!frame.current || !content.current) return;
const tag = content.current.querySelector<HTMLElement>(".katex-html > .tag");
const tagWidth = `${tag?.offsetWidth ?? 0}px`;
if (content.current.style.getPropertyValue("--math-tag-width") !== tagWidth) {
content.current.style.setProperty("--math-tag-width", tagWidth);
}
const width = content.current.offsetWidth;
const scale = width ? Math.min(1, frame.current.clientWidth / width) : 1;
const height = content.current.offsetHeight * scale;
setLayout(previous => previous.scale === scale && previous.height === height ? previous : { scale, height });
}, []);
useLayoutEffect(() => {
const observer = new ResizeObserver(update);
observer.observe(frame.current!);
observer.observe(content.current!);
let cancelled = false;
void document.fonts?.ready.then(() => { if (!cancelled) update(); });
return () => { cancelled = true; observer.disconnect(); };
}, [update]);
useLayoutEffect(update, [children, update]);
return <span ref={frame} className="math-fit" style={{ height: layout.height || undefined }}
data-math-scale={layout.scale}>
<span ref={content} className="math-fit-content" style={{ transform: `translateX(-50%) scale(${layout.scale})` }}>
<span className="katex-display">{children}</span>
</span>
</span>;
}