Open-source platform for 3D AI agents. Turn text or a photo into a rigged, animated GLB avatar, give it an LLM brain, memory and a wallet, and embed it anywhere with one web component. Remote MCP server, x402 payments, Three.js, browser-native.
JavaScript 3d-generationai-agentsanimationavatardigital-humangenerative-aiglbgltfimage-to-3dllmmcpmcp-servermodel-context-protocolsolanatext-to-3dthreejsweb-componentswebglwebxrx402
export function parseXML(xmlString) {
// Regular expression to match the opening tag
const openingTagRegex = /<([^\s>]+)([^>]*)>/;
const match = xmlString.match(openingTagRegex);
if (!match) {
throw new Error("Invalid XML: No opening tag found");
}
// Extract element name (first capture group)
const element = match[1];
// Extract attributes string (second capture group) and trim whitespace
const attributesString = match[2].trim();
// Parse attributes
const attributes = {};
if (attributesString) {
// Pattern to match attribute name and optional value
// Handles: name="value", name='value', name=value, and standalone attributes
const pattern = /([^\s=]+)(?:=(?:"([^"]*)"|'([^']*)'|([^\s>]+)))?/g;
let attrMatch;
while ((attrMatch = pattern.exec(attributesString)) !== null) {
const [, name, doubleQuoted, singleQuoted, unquoted] = attrMatch;
// Determine the value - use the first defined value or true for standalone attributes
const value = doubleQuoted !== undefined ? doubleQuoted :
singleQuoted !== undefined ? singleQuoted :
unquoted !== undefined ? unquoted : true;
attributes[name] = value;
}
}
// Extract content between opening and closing tags
const fullOpeningTag = xmlString.match(new RegExp(`<${element}[^>]*>`))[0];
const closingTag = `</${element}>`;
const startPos = xmlString.indexOf(fullOpeningTag) + fullOpeningTag.length;
const endPos = xmlString.lastIndexOf(closingTag);
let content = "";
if (startPos > 0 && endPos > startPos) {
content = xmlString.substring(startPos, endPos).trim();
}
return {
element,
attributes,
content
};
}