blun-king-cli 5.4.0 → 6.1.0

package/reference-inspector.js

@@ -1,256 +0,0 @@
-const path = require("path");
-const { execFile } = require("child_process");
-const { storeReference } = require("./local-data");
-
-let chromiumLoader = null;
-
-function normalizeDetectedUrl(input) {
-  const raw = String(input || "").trim();
-  if (!raw) return null;
-
-  let cleaned = raw
-    .replace(/^[<\s"'`(]+/, "")
-    .replace(/[>\s"'`),.;:!?]+$/, "");
-
-  if (!cleaned) return null;
-  if (/^www\./i.test(cleaned)) cleaned = `https://${cleaned}`;
-  if (/^[a-z0-9.-]+\.[a-z]{2,}(\/|$)/i.test(cleaned) && !/^[a-z][a-z0-9+.-]*:/i.test(cleaned)) {
-    cleaned = `https://${cleaned}`;
-  }
-
-  try {
-    const parsed = new URL(cleaned);
-    if (!/^https?:$/i.test(parsed.protocol) && !/^data:$/i.test(parsed.protocol)) return null;
-    return parsed.toString();
-  } catch {
-    return null;
-  }
-}
-
-function extractUrls(text) {
-  const matches = [
-    ...Array.from(String(text || "").matchAll(/https?:\/\/[^\s)>"']+/gi)).map((match) => match[0]),
-    ...Array.from(String(text || "").matchAll(/(?:^|[\s<(])((?:www\.)?[a-z0-9.-]+\.[a-z]{2,}(?:\/[^\s)>"']*)?)/gi)).map((match) => match[1])
-  ];
-
-  return [...new Set(matches.map(normalizeDetectedUrl).filter(Boolean))];
-}
-
-function summarizeHtml(source) {
-  const title = (source.match(/<title[^>]*>([\s\S]*?)<\/title>/i) || [, ""])[1].trim();
-  const h1s = [...source.matchAll(/<h1[^>]*>([\s\S]*?)<\/h1>/ig)]
-    .map((match) => match[1].replace(/<[^>]+>/g, "").trim())
-    .filter(Boolean);
-  const links = [...source.matchAll(/<a\b/ig)].length;
-  const images = [...source.matchAll(/<img\b/ig)].length;
-  const text = cleanText((source.match(/<body[^>]*>([\s\S]*?)<\/body>/i) || [, source])[1]).slice(0, 3000);
-
-  return {
-    title,
-    h1s,
-    links,
-    images,
-    text,
-    responsive: /<meta[^>]+name=["']viewport["']/i.test(source)
-  };
-}
-
-function cleanText(source) {
-  return String(source || "")
-    .replace(/<script[\s\S]*?<\/script>/gi, " ")
-    .replace(/<style[\s\S]*?<\/style>/gi, " ")
-    .replace(/<[^>]+>/g, " ")
-    .replace(/\s+/g, " ")
-    .trim();
-}
-
-function shouldAutoReference(input, task = {}) {
-  const text = String(input || "");
-  if (!extractUrls(text).length) return false;
-  if (["installation", "browser_capture"].includes(task?.type)) return false;
-  if (task?.type === "website_builder" || task?.type === "analysis") return true;
-  return /\b(referenz|vorlage|inspiriert|angelehnt|wie diese|wie die|schau dir|analysier.*https?:\/\/|bau.*https?:\/\/)\b/i.test(text);
-}
-
-function buildReferencePromptBlock(references = []) {
-  if (!Array.isArray(references) || references.length === 0) return "";
-
-  const lines = ["[Auto Reference Chain]"];
-  references.forEach((ref, index) => {
-    const summary = ref.summary || ref;
-    lines.push(`Referenz ${index + 1}: ${ref.url || "-"}`);
-    if (summary.title) lines.push(`Titel: ${summary.title}`);
-    if (Array.isArray(summary.h1s) && summary.h1s.length) lines.push(`H1: ${summary.h1s.join(" | ")}`);
-    if (summary.text) lines.push(`Text: ${String(summary.text).slice(0, 500)}`);
-    if (typeof summary.links === "number") lines.push(`Links: ${summary.links}`);
-    if (typeof summary.images === "number") lines.push(`Bilder: ${summary.images}`);
-    if (ref.screenshotPath) lines.push(`Screenshot: ${ref.screenshotPath}`);
-    if (ref.error) lines.push(`Fehler: ${ref.error}`);
-  });
-
-  return lines.join("\n");
-}
-
-async function loadChromium() {
-  if (chromiumLoader !== null) return chromiumLoader;
-
-  try {
-    const playwright = require("playwright");
-    chromiumLoader = playwright.chromium;
-  } catch {
-    chromiumLoader = false;
-  }
-
-  return chromiumLoader;
-}
-
-function execFileAsync(command, args) {
-  return new Promise((resolve, reject) => {
-    execFile(command, args, { maxBuffer: 20 * 1024 * 1024 }, (error, stdout, stderr) => {
-      if (error) reject(new Error(stderr || error.message));
-      else resolve(stdout);
-    });
-  });
-}
-
-async function fetchHtml(url, options = {}) {
-  try {
-    const resp = await fetch(url, {
-      headers: {
-        "User-Agent": options.userAgent || "BLUN-King/5.0 (+reference-inspector)"
-      }
-    });
-    if (!resp.ok) throw new Error(`Reference fetch failed with HTTP ${resp.status}`);
-    return await resp.text();
-  } catch {
-    try {
-      const curlBinary = process.platform === "win32" ? "curl.exe" : "curl";
-      return await execFileAsync(curlBinary, [
-        "-L",
-        "-A",
-        options.userAgent || "BLUN-King/5.0 (+reference-inspector)",
-        url
-      ]);
-    } catch {
-      if (process.platform === "win32") {
-        const script = [
-          `$ProgressPreference='SilentlyContinue'`,
-          `$resp = Invoke-WebRequest -UseBasicParsing -Headers @{'User-Agent'='${options.userAgent || "BLUN-King/5.0 (+reference-inspector)"}'} -Uri '${url}'`,
-          `[Console]::Out.Write($resp.Content)`
-        ].join("; ");
-        return await execFileAsync("powershell.exe", ["-NoProfile", "-Command", script]);
-      }
-
-      throw new Error("Reference fetch failed");
-    }
-  }
-}
-
-async function inspectReference(url, options = {}) {
-  let html = "";
-  let summary = {
-    title: "",
-    h1s: [],
-    links: 0,
-    images: 0,
-    responsive: false
-  };
-  const result = {
-    url,
-    fetchedAt: Date.now(),
-    summary,
-    html,
-    screenshotPath: null,
-    screenshotAvailable: false,
-    error: null
-  };
-
-  try {
-    html = await fetchHtml(url, options);
-    summary = summarizeHtml(html);
-    result.html = html;
-    result.summary = summary;
-  } catch (error) {
-    result.error = error.message;
-  }
-
-  if (!result.error && options.screenshot !== false) {
-    const chromium = await loadChromium();
-    if (chromium) {
-      try {
-        const browser = await chromium.launch({ headless: true });
-        try {
-          const page = await browser.newPage({ viewport: { width: 1440, height: 1024 } });
-          await page.goto(url, { waitUntil: "networkidle", timeout: options.timeout || 30000 });
-          const stored = storeReference(url, {
-            type: "reference",
-            mode: "html+screenshot",
-            summary
-          });
-          const screenshotPath = path.join(path.dirname(stored.filePath), `${stored.id}.png`);
-          await page.screenshot({ path: screenshotPath, fullPage: true });
-          result.screenshotPath = screenshotPath;
-          result.screenshotAvailable = true;
-          result.storage = stored.filePath;
-        } finally {
-          await browser.close();
-        }
-      } catch {
-        result.screenshotAvailable = false;
-      }
-    }
-  }
-
-  if (!result.storage) {
-    const stored = storeReference(url, {
-      type: "reference",
-      mode: result.error ? "error" : "html",
-      summary,
-      error: result.error
-    });
-    result.storage = stored.filePath;
-  }
-
-  return result;
-}
-
-async function screenshotUrl(url, options = {}) {
-  const result = await inspectReference(url, { ...options, screenshot: true });
-  return {
-    url: result.url,
-    screenshotPath: result.screenshotPath,
-    screenshotAvailable: result.screenshotAvailable,
-    summary: result.summary,
-    storage: result.storage,
-    error: result.error
-  };
-}
-
-async function autoReferenceChain(input, options = {}) {
-  const urls = extractUrls(input).slice(0, options.limit || 2);
-  const references = [];
-
-  for (const url of urls) {
-    try {
-      references.push(await inspectReference(url, { screenshot: options.screenshot !== false }));
-    } catch (error) {
-      references.push({
-        url,
-        error: error.message,
-        summary: { title: "", h1s: [], links: 0, images: 0, text: "", responsive: false }
-      });
-    }
-  }
-
-  return references;
-}
-
-module.exports = {
-  extractUrls,
-  inspectReference,
-  summarizeHtml,
-  shouldAutoReference,
-  buildReferencePromptBlock,
-  screenshotUrl,
-  autoReferenceChain
-};