@wrongstack/plugins 0.268.0 → 0.270.0

package/dist/index.js

@@ -1345,13 +1345,13 @@ async function duckduckgoSearch(query, numResults) {
   if (!resp.ok) throw new Error(`DuckDuckGo search failed: ${resp.status}`);
   const html = await resp.text();
   const results = [];
-  const resultRe = /<a class="result__a" href="([^"]+)"[^>]*>([\s\S]*?)<\/a>[\s\S]*?<a class="result__snippet"[^>]*>([\s\S]*?)<\/a>/g;
+  const resultRe = /<a\b[^>]*class="[^"]*result__a[^"]*"[^>]*href="([^"]+)"[^>]*>([\s\S]*?)<\/a>([\s\S]*?)(?=<a\b[^>]*class="[^"]*result__a[^"]*"|<\/body>|$)/gi;
   let m;
   while ((m = resultRe.exec(html)) !== null && results.length < numResults) {
-    const url2 = m[1];
+    const url2 = normalizeDuckDuckGoUrl(m[1]);
     if (!url2) continue;
-    const title = (m[2] ?? "").replace(/<[^>]+>/g, "").trim();
-    const snippet = (m[3] ?? "").replace(/<[^>]+>/g, "").trim();
+    const title = htmlToText(m[2] ?? "");
+    const snippet = extractDuckDuckGoSnippet(m[3] ?? "");
     results.push({
       url: url2,
       title,
@@ -1361,8 +1361,33 @@ async function duckduckgoSearch(query, numResults) {
       cached: false
     });
   }
+  if (results.length === 0 && /result__a|result__snippet|result__body|anomaly-modal|captcha/i.test(html)) {
+    throw new Error("DuckDuckGo response format was not recognized or was blocked");
+  }
   return results;
 }
+function normalizeDuckDuckGoUrl(rawUrl) {
+  if (!rawUrl) return null;
+  let url = htmlToText(rawUrl);
+  if (url.startsWith("//duckduckgo.com/l/")) url = `https:${url}`;
+  if (url.startsWith("/l/")) url = new URL(url, "https://duckduckgo.com").toString();
+  try {
+    const parsed = new URL(url);
+    const redirected = parsed.searchParams.get("uddg");
+    if (redirected) return redirected;
+    return parsed.toString();
+  } catch {
+    return null;
+  }
+}
+function htmlToText(html) {
+  return html.replace(/<[^>]+>/g, "").replace(/&amp;/g, "&").replace(/&lt;/g, "<").replace(/&gt;/g, ">").replace(/&quot;/g, '"').replace(/&#39;/g, "'").replace(/\s+/g, " ").trim();
+}
+function extractDuckDuckGoSnippet(html) {
+  const snippetRe = /<(?:a|div)\b[^>]*class="[^"]*(?:result__snippet|result__body)[^"]*"[^>]*>([\s\S]*?)<\/(?:a|div)>/i;
+  const snippetMatch = snippetRe.exec(html);
+  return htmlToText(snippetMatch?.[1] ?? "");
+}
 function assertSafeIp(ip) {
   if (isIPv4(ip) && isPrivateIPv4(ip)) {
     throw new Error(`Blocked private/loopback address: ${ip}`);

package/dist/web-search.js

@@ -14,13 +14,13 @@ async function duckduckgoSearch(query, numResults) {
   if (!resp.ok) throw new Error(`DuckDuckGo search failed: ${resp.status}`);
   const html = await resp.text();
   const results = [];
-  const resultRe = /<a class="result__a" href="([^"]+)"[^>]*>([\s\S]*?)<\/a>[\s\S]*?<a class="result__snippet"[^>]*>([\s\S]*?)<\/a>/g;
+  const resultRe = /<a\b[^>]*class="[^"]*result__a[^"]*"[^>]*href="([^"]+)"[^>]*>([\s\S]*?)<\/a>([\s\S]*?)(?=<a\b[^>]*class="[^"]*result__a[^"]*"|<\/body>|$)/gi;
   let m;
   while ((m = resultRe.exec(html)) !== null && results.length < numResults) {
-    const url2 = m[1];
+    const url2 = normalizeDuckDuckGoUrl(m[1]);
     if (!url2) continue;
-    const title = (m[2] ?? "").replace(/<[^>]+>/g, "").trim();
-    const snippet = (m[3] ?? "").replace(/<[^>]+>/g, "").trim();
+    const title = htmlToText(m[2] ?? "");
+    const snippet = extractDuckDuckGoSnippet(m[3] ?? "");
     results.push({
       url: url2,
       title,
@@ -30,8 +30,33 @@ async function duckduckgoSearch(query, numResults) {
       cached: false
     });
   }
+  if (results.length === 0 && /result__a|result__snippet|result__body|anomaly-modal|captcha/i.test(html)) {
+    throw new Error("DuckDuckGo response format was not recognized or was blocked");
+  }
   return results;
 }
+function normalizeDuckDuckGoUrl(rawUrl) {
+  if (!rawUrl) return null;
+  let url = htmlToText(rawUrl);
+  if (url.startsWith("//duckduckgo.com/l/")) url = `https:${url}`;
+  if (url.startsWith("/l/")) url = new URL(url, "https://duckduckgo.com").toString();
+  try {
+    const parsed = new URL(url);
+    const redirected = parsed.searchParams.get("uddg");
+    if (redirected) return redirected;
+    return parsed.toString();
+  } catch {
+    return null;
+  }
+}
+function htmlToText(html) {
+  return html.replace(/<[^>]+>/g, "").replace(/&amp;/g, "&").replace(/&lt;/g, "<").replace(/&gt;/g, ">").replace(/&quot;/g, '"').replace(/&#39;/g, "'").replace(/\s+/g, " ").trim();
+}
+function extractDuckDuckGoSnippet(html) {
+  const snippetRe = /<(?:a|div)\b[^>]*class="[^"]*(?:result__snippet|result__body)[^"]*"[^>]*>([\s\S]*?)<\/(?:a|div)>/i;
+  const snippetMatch = snippetRe.exec(html);
+  return htmlToText(snippetMatch?.[1] ?? "");
+}
 function assertSafeIp(ip) {
   if (isIPv4(ip) && isPrivateIPv4(ip)) {
     throw new Error(`Blocked private/loopback address: ${ip}`);

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@wrongstack/plugins",
-  "version": "0.268.0",
+  "version": "0.270.0",
   "description": "Official WrongStack plugin collection — auto-doc, git-autocommit, shell-check, cost-tracker, file-watcher, web-search, json-path, cron, template-engine, semver-bump",
   "license": "MIT",
   "author": "ECOSTACK TECHNOLOGY OÜ",
@@ -63,7 +63,7 @@
     "vitest": "^4.1.8"
   },
   "dependencies": {
-    "@wrongstack/core": "0.268.0"
+    "@wrongstack/core": "0.270.0"
   },
   "scripts": {
     "build": "tsup",