//
//  AtlasExtractor.js
//  Zyquo Atlas
//
//  Author: Simon-Pierre Boucher
//  Mail: contact@spboucher.ai
//
//  Content-extraction driver injected into an isolated WKContentWorld alongside
//  Mozilla Readability.js + Readability-readerable.js. Turns the live page into
//  clean, model-ready markdown with metadata and the user's selection, using
//  the fallback ladder from docs/AI-BROWSER-RESEARCH.md §2:
//    1. Readability (article) → markdown
//    2. visible structured-text walk (app pages)
//    3. title + url + meta description
//  SECURITY: only visible, rendered content is emitted — hidden/off-screen and
//  aria-hidden nodes are dropped so injected instructions in invisible text
//  never reach the model. Every entry point returns a value (never undefined)
//  so the Swift bridge never crashes.
//

(function () {
  "use strict";
  if (window.__zyquoAtlas) return;

  const MAX_BYTES_DEFAULT = 600000;

  function meta(name) {
    const el = document.querySelector(
      'meta[property="' + name + '"],meta[name="' + name + '"]'
    );
    return el ? el.getAttribute("content") : null;
  }

  function canonicalURL() {
    const l = document.querySelector('link[rel="canonical"]');
    return (l && l.href) || location.href;
  }

  function faviconURL() {
    const l = document.querySelector('link[rel~="icon"],link[rel="apple-touch-icon"]');
    try {
      return l ? new URL(l.getAttribute("href"), location.href).href
               : new URL("/favicon.ico", location.origin).href;
    } catch (_) { return null; }
  }

  // Is a node visible to the user (drops the injection surface)?
  function isVisible(el) {
    if (!el || el.nodeType !== 1) return false;
    if (el.getAttribute && el.getAttribute("aria-hidden") === "true") return false;
    const s = window.getComputedStyle(el);
    if (!s || s.display === "none" || s.visibility === "hidden" || parseFloat(s.opacity) === 0) {
      return false;
    }
    const r = el.getBoundingClientRect();
    if (r.width === 0 && r.height === 0) return false;
    return true;
  }

  // Walk visible DOM emitting lightweight markdown (headings, lists, links,
  // code, paragraphs). Used both to render a Readability article DOM and as the
  // app-page fallback.
  function toMarkdown(root) {
    const out = [];
    const SKIP = new Set(["SCRIPT", "STYLE", "NOSCRIPT", "TEMPLATE", "SVG", "CANVAS", "IFRAME"]);
    function walk(node, listDepth) {
      for (const child of node.childNodes) {
        if (child.nodeType === 3) {
          const t = child.textContent.replace(/\s+/g, " ");
          if (t.trim()) out.push(t);
          continue;
        }
        if (child.nodeType !== 1) continue;
        const tag = child.tagName;
        if (SKIP.has(tag)) continue;
        if (!isVisible(child)) continue;
        if (/^H[1-6]$/.test(tag)) {
          const level = "#".repeat(parseInt(tag[1], 10));
          out.push("\n" + level + " " + child.textContent.trim() + "\n");
        } else if (tag === "P") {
          out.push("\n" + child.textContent.trim() + "\n");
        } else if (tag === "LI") {
          out.push("\n" + "  ".repeat(listDepth) + "- " + child.textContent.trim());
        } else if (tag === "PRE") {
          out.push("\n```\n" + child.textContent.replace(/\n+$/, "") + "\n```\n");
        } else if (tag === "A" && child.getAttribute("href")) {
          const txt = child.textContent.trim();
          if (txt) out.push("[" + txt + "](" + child.href + ")");
        } else if (tag === "BR") {
          out.push("\n");
        } else if (tag === "UL" || tag === "OL") {
          walk(child, listDepth + 1);
          out.push("\n");
        } else {
          walk(child, listDepth);
        }
      }
    }
    walk(root, 0);
    return out.join(" ").replace(/[ \t]+/g, " ").replace(/\n{3,}/g, "\n\n").trim();
  }

  // Index of heading offsets in the produced markdown for chunking + citations.
  function headingIndex(markdown) {
    const headings = [];
    const re = /^(#{1,6})\s+(.*)$/gm;
    let m;
    while ((m = re.exec(markdown)) !== null) {
      headings.push({ level: m[1].length, title: m[2].trim(), offset: m.index });
    }
    return headings;
  }

  function currentSelection() {
    const sel = window.getSelection();
    const text = sel ? String(sel).trim() : "";
    if (!text) return null;
    let context = null;
    try {
      const node = sel.anchorNode && sel.anchorNode.parentElement;
      const block = node && node.closest("p,li,section,article,div");
      if (block) context = block.textContent.replace(/\s+/g, " ").trim().slice(0, 800);
    } catch (_) {}
    return { text: text, context: context };
  }

  function truncateAtBoundary(s, maxBytes) {
    if (s.length <= maxBytes) return s;
    const cut = s.lastIndexOf("\n", maxBytes);
    return s.slice(0, cut > maxBytes * 0.5 ? cut : maxBytes) + "\n\n…[truncated]";
  }

  function extract(mode, maxBytes) {
    maxBytes = maxBytes || MAX_BYTES_DEFAULT;
    try {
      let markdown = "";
      let quality = "reader";
      let title = document.title;
      let byline = null;
      let published = null;
      let readerable = false;

      try {
        readerable = typeof isProbablyReaderable === "function" &&
                     isProbablyReaderable(document);
      } catch (_) {}

      if (mode !== "rawText" && readerable && typeof Readability === "function") {
        try {
          const clone = document.cloneNode(true); // never mutate the live page
          const article = new Readability(clone).parse();
          if (article) {
            title = article.title || title;
            byline = article.byline || null;
            published = article.publishedTime || null;
            const holder = document.createElement("div");
            holder.innerHTML = article.content || "";
            markdown = toMarkdown(holder);
          }
        } catch (_) {}
      }

      if (!markdown || markdown.length < 400) {           // fallback ladder
        markdown = toMarkdown(document.body || document.documentElement);
        quality = readerable ? "reader" : "rawText";
      }
      if (!markdown) {                                     // last resort
        markdown = (document.body ? document.body.innerText : "").trim();
        quality = "rawText";
      }

      markdown = truncateAtBoundary(markdown, maxBytes);

      return {
        ok: true,
        context: {
          url: location.href,
          canonical: canonicalURL(),
          title: (title || location.host || "").trim(),
          byline: byline,
          description: meta("description") || meta("og:description"),
          siteName: meta("og:site_name"),
          lang: document.documentElement.lang || null,
          published: published,
          favicon: faviconURL(),
          markdown: markdown,
          quality: quality,
          headings: headingIndex(markdown),
          selection: currentSelection(),
          wordCount: markdown ? markdown.split(/\s+/).length : 0,
          truncated: markdown.indexOf("…[truncated]") !== -1
        }
      };
    } catch (e) {
      return { ok: false, error: String((e && e.stack) || e) };
    }
  }

  window.__zyquoAtlas = { extract: extract };

  // Selection observer → native floating toolbar. Debounced; posts the trimmed
  // selection text + its viewport rect, or an empty text on collapse.
  var selTimer = null, lastSel = "";
  function reportSelection() {
    try {
      var sel = window.getSelection();
      var text = sel ? String(sel).trim() : "";
      if (text === lastSel) return;
      lastSel = text;
      var payload = { text: text };
      if (text && sel.rangeCount) {
        var r = sel.getRangeAt(0).getBoundingClientRect();
        payload.x = r.x; payload.y = r.y; payload.w = r.width; payload.h = r.height;
      }
      if (window.webkit && window.webkit.messageHandlers &&
          window.webkit.messageHandlers.atlasSelection) {
        window.webkit.messageHandlers.atlasSelection.postMessage(payload);
      }
    } catch (e) { /* ignore */ }
  }
  document.addEventListener("selectionchange", function () {
    if (selTimer) clearTimeout(selTimer);
    selTimer = setTimeout(reportSelection, 220);
  });
})();