From f93ccc344358009dd42d6f040aed25c33d6be773 Mon Sep 17 00:00:00 2001
From: jnuyao <jnuyao@gmail.com>
Date: Sun, 29 Mar 2026 09:49:29 +0800
Subject: [PATCH] fix(ui): prevent marked from auto-linking adjacent CJK
 characters (openclaw#48410)

Verified:
- ui: pnpm test -- --run src/ui/markdown.test.ts
- local full gate relaxed for this run; no required GitHub checks reported on the branch

Co-authored-by: jnuyao <2928523+jnuyao@users.noreply.github.com>
Co-authored-by: Tak Hoffman <781889+Takhoffman@users.noreply.github.com>
---
 CHANGELOG.md               |  1 +
 ui/src/ui/markdown.test.ts | 18 +++++++++++++++
 ui/src/ui/markdown.ts      | 45 ++++++++++++++++++++++++++++++++++++++
 3 files changed, 64 insertions(+)

diff --git a/CHANGELOG.md b/CHANGELOG.md
index fec2bdfd69f..c798c658527 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -14,6 +14,7 @@ Docs: https://docs.openclaw.ai
 - Gateway/health: carry webhook-vs-polling account mode from channel descriptors into runtime snapshots so passive channels like LINE and BlueBubbles skip false stale-socket health failures. (#47488) Thanks @karesansui-u.
 - Memory/QMD: honor `memory.qmd.update.embedInterval` even when regular QMD update cadence is disabled or slower by arming a dedicated embed-cadence maintenance timer, while avoiding redundant timers when regular updates are already frequent enough. (#37326) Thanks @barronlroth.
 - Agents/memory flush: keep daily memory flush files append-only during embedded attempts so compaction writes do not overwrite earlier notes. (#53725) Thanks @HPluseven.
+- Web UI/markdown: stop bare auto-links from swallowing adjacent CJK text while preserving valid mixed-script path and query characters in rendered links. (#48410) Thanks @jnuyao.
 
 ## 2026.3.28
 
diff --git a/ui/src/ui/markdown.test.ts b/ui/src/ui/markdown.test.ts
index 8c2f37cbea4..e27faf8fbaa 100644
--- a/ui/src/ui/markdown.test.ts
+++ b/ui/src/ui/markdown.test.ts
@@ -162,4 +162,22 @@ describe("toSanitizedMarkdownHtml", () => {
       warnSpy.mockRestore();
     }
   });
+
+  it("keeps adjacent trailing CJK text outside bare auto-links", () => {
+    const html = toSanitizedMarkdownHtml("https://example.com重新解读");
+    expect(html).toContain('<a href="https://example.com"');
+    expect(html).toContain(">https://example.com</a>重新解读");
+  });
+
+  it("preserves valid mixed-script query parameters inside auto-links", () => {
+    const html = toSanitizedMarkdownHtml("https://api.example.com?q=重新&lang=en");
+    expect(html).toContain('href="https://api.example.com?q=%E9%87%8D%E6%96%B0&amp;lang=en"');
+    expect(html).toContain(">https://api.example.com?q=重新&amp;lang=en</a>");
+  });
+
+  it("preserves valid mixed-script path segments inside auto-links", () => {
+    const html = toSanitizedMarkdownHtml("https://example.com/path/重新/file");
+    expect(html).toContain('href="https://example.com/path/%E9%87%8D%E6%96%B0/file"');
+    expect(html).toContain(">https://example.com/path/重新/file</a>");
+  });
 });
diff --git a/ui/src/ui/markdown.ts b/ui/src/ui/markdown.ts
index 160d0e96399..6f5c69b2b30 100644
--- a/ui/src/ui/markdown.ts
+++ b/ui/src/ui/markdown.ts
@@ -63,6 +63,7 @@ const MARKDOWN_CACHE_MAX_CHARS = 50_000;
 const INLINE_DATA_IMAGE_RE = /^data:image\/[a-z0-9.+-]+;base64,/i;
 const markdownCache = new Map<string, string>();
 const TAIL_LINK_BLUR_CLASS = "chat-link-tail-blur";
+const TRAILING_CJK_TAIL_RE = /([\u4E00-\u9FFF\u3000-\u303F\uFF01-\uFF5E\s]+)$/;
 
 function getCachedMarkdown(key: string): string | null {
   const cached = markdownCache.get(key);
@@ -121,6 +122,50 @@ function installHooks() {
   });
 }
 
+// Extension to prevent auto-linking algorithms from swallowing adjacent CJK characters.
+const cjkAutoLinkExtension = {
+  name: "url",
+  level: "inline",
+  // Indicate where an auto-link might start
+  start(src: string) {
+    const match = src.match(/https?:\/\//i);
+    return match ? match.index! : -1;
+  },
+  tokenizer(src: string) {
+    // GFM standard regex for auto-links
+    const rule = /^https?:\/\/[^\s<]+[^<.,:;"')\]\s]/i;
+    const match = rule.exec(src);
+    if (match) {
+      let urlText = match[0];
+
+      // Stop before any CJK character or typical punctuation following CJK
+      // This stops link boundaries from bleeding into mixed-language paragraphs.
+      const cjkMatch = urlText.match(TRAILING_CJK_TAIL_RE);
+      if (cjkMatch) {
+        urlText = urlText.substring(0, urlText.length - cjkMatch[1].length);
+      }
+
+      return {
+        type: "link",
+        raw: urlText,
+        text: urlText,
+        href: urlText,
+        tokens: [
+          {
+            type: "text",
+            raw: urlText,
+            text: urlText,
+          },
+        ],
+      };
+    }
+  },
+};
+
+marked.use({
+  extensions: [cjkAutoLinkExtension as unknown as import("marked").TokenizerAndRendererExtension],
+});
+
 export function toSanitizedMarkdownHtml(markdown: string): string {
   const input = markdown.trim();
   if (!input) {