From 88f1eb3d037c7214421f9d72940e94f62548a45c Mon Sep 17 00:00:00 2001
From: khannurien <yo@vincentlannurien.com>
Date: Sat, 27 Jun 2026 14:34:47 +0000
Subject: [PATCH] v3: rich content meta regex improvements

---
 api/services/rich-content-service.ts | 57 +++++++++++++---------------
 1 file changed, 26 insertions(+), 31 deletions(-)

diff --git a/api/services/rich-content-service.ts b/api/services/rich-content-service.ts
index f00dca8..977a505 100644
--- a/api/services/rich-content-service.ts
+++ b/api/services/rich-content-service.ts
@@ -139,25 +139,35 @@ function decodeHtmlEntities(str: string): string {
     );
 }
 
+/**
+ * Find the first `<meta>` tag whose `attr` equals `value` and return its
+ * `content` (either attribute ordering). Each tag is matched in isolation so
+ * the `content` capture can never span across tag boundaries.
+ */
+function extractMetaContent(
+  html: string,
+  attr: string,
+  value: string,
+): string | undefined {
+  const escaped = value.replace(/[.*+?^${}()|[\]\\]/g, "\\$&");
+  const keyRe = new RegExp(`\\b${attr}=["']${escaped}["']`, "i");
+  const contentRe = /\bcontent=(["'])([\s\S]*?)\1/i;
+  const tagRe = /<meta\b[^>]*>/gi;
+  let m: RegExpExecArray | null;
+  while ((m = tagRe.exec(html)) !== null) {
+    const tag = m[0];
+    if (!keyRe.test(tag)) continue;
+    const content = contentRe.exec(tag);
+    if (content) return decodeHtmlEntities(content[2]);
+  }
+  return undefined;
+}
+
 export function extractOgTag(
   html: string,
   tag: string,
 ): string | undefined {
-  const patterns = [
-    new RegExp(
-      `<meta[^>]+property=["']og:${tag}["'][^>]+content=(["'])([\\s\\S]*?)\\1`,
-      "i",
-    ),
-    new RegExp(
-      `<meta[^>]+content=(["'])([\\s\\S]*?)\\1[^>]+property=["']og:${tag}["']`,
-      "i",
-    ),
-  ];
-  for (const pattern of patterns) {
-    const match = html.match(pattern);
-    if (match) return decodeHtmlEntities(match[2]);
-  }
-  return undefined;
+  return extractMetaContent(html, "property", `og:${tag}`);
 }
 
 /** Extract content from `<meta name="…" content="…">` (both attribute orderings). */
@@ -165,22 +175,7 @@ export function extractMetaName(
   html: string,
   name: string,
 ): string | undefined {
-  const escaped = name.replace(/[.*+?^${}()|[\]\\]/g, "\\$&");
-  const patterns = [
-    new RegExp(
-      `<meta[^>]+name=["']${escaped}["'][^>]+content=(["'])([\\s\\S]*?)\\1`,
-      "i",
-    ),
-    new RegExp(
-      `<meta[^>]+content=(["'])([\\s\\S]*?)\\1[^>]+name=["']${escaped}["']`,
-      "i",
-    ),
-  ];
-  for (const pattern of patterns) {
-    const match = html.match(pattern);
-    if (match) return decodeHtmlEntities(match[2]);
-  }
-  return undefined;
+  return extractMetaContent(html, "name", name);
 }
 
 /** Extract the text content of the `<title>` element. */