camoufox-mcp-server/dist/extractors/links.js

MCP server for browser automation using Camoufox - a privacy-focused Firefox fork with advanced anti-detection features

import { redactUrl } from "../utils.js";
export async function extractLinks(page, maxLinks, selector) {
    return page.evaluate(({ maxItems, cssSelector }) => {
        const root = cssSelector
            ? document.querySelector(cssSelector)
            : document.body ?? document.documentElement;
        if (!root) {
            return { links: [], truncated: false, found: false };
        }
        function cssIdent(value) {
            if (typeof CSS !== "undefined" && typeof CSS.escape === "function") {
                return CSS.escape(value);
            }
            return value.replace(/[^a-zA-Z0-9_-]/g, "\\$&");
        }
        function selectorFor(element) {
            if (element.id) {
                return `#${cssIdent(element.id)}`;
            }
            const path = [];
            let current = element;
            while (current && current !== document.documentElement && path.length < 8) {
                let part = current.tagName.toLowerCase();
                if (current.classList.length > 0) {
                    part += `.${Array.from(current.classList).slice(0, 2).map(cssIdent).join(".")}`;
                }
                const parent = current.parentElement;
                if (parent) {
                    const sameTagSiblings = Array.from(parent.children).filter((child) => child.tagName === current?.tagName);
                    if (sameTagSiblings.length > 1) {
                        part += `:nth-of-type(${sameTagSiblings.indexOf(current) + 1})`;
                    }
                }
                path.unshift(part);
                current = parent;
            }
            return path.join(" > ");
        }
        function textOf(element) {
            return (element.textContent ?? element.getAttribute("aria-label") ?? element.getAttribute("title") ?? "")
                .replace(/\s+/g, " ")
                .trim()
                .slice(0, 500);
        }
        function isVisible(element) {
            const rect = element.getBoundingClientRect();
            const style = window.getComputedStyle(element);
            return rect.width > 0 && rect.height > 0 && style.display !== "none" && style.visibility !== "hidden" && style.visibility !== "collapse";
        }
        const candidates = [
            ...(root.matches("a[href]") ? [root] : []),
            ...Array.from(root.querySelectorAll("a[href]")),
        ];
        const links = [];
        let truncated = false;
        const seen = new Set();
        for (const link of candidates) {
            const href = link.href;
            if (!href || seen.has(href)) {
                continue;
            }
            const visible = isVisible(link);
            const text = textOf(link);
            if (!text && !visible) {
                continue;
            }
            if (links.length >= maxItems) {
                truncated = true;
                break;
            }
            seen.add(href);
            links.push({
                text,
                href,
                selector: selectorFor(link),
                visible,
                confidence: visible && text ? 0.95 : visible || text ? 0.75 : 0.5,
            });
        }
        return { links, truncated, found: true };
    }, { maxItems: maxLinks, cssSelector: selector });
}
export async function buildLinksPayload(page, response, maxLinks, selector) {
    const extracted = await extractLinks(page, maxLinks, selector);
    return {
        url: redactUrl(page.url()),
        title: await page.title(),
        status: response?.status(),
        contentType: response?.headers()["content-type"],
        selector,
        selectorFound: extracted.found,
        links: extracted.links,
        truncated: extracted.truncated,
        maxLinks,
    };
}