From 176a008ee354ae794afcc676f25d157bc6cc8c44 Mon Sep 17 00:00:00 2001
From: Claude <noreply@anthropic.com>
Date: Sat, 4 Apr 2026 16:25:54 +0000
Subject: [PATCH] Fix Full Article extraction for Naver Blog

Naver Blog's desktop site (blog.naver.com) is a JavaScript-heavy SPA
that Mercury/Feedbin's extractor cannot parse, returning no content.
The mobile site (m.blog.naver.com) renders as static HTML and extracts
successfully.

Add extractionURL(for:) to ArticleExtractor that transforms
blog.naver.com URLs to m.blog.naver.com before sending to the extractor
service, also stripping the ?fromRss=true&trackingCode=rss tracking
parameters that Naver appends to RSS feed links.

https://claude.ai/code/session_01DWJjZT2o7Cir7tmGYR3Z2U
---
 .../Article Extractor/ArticleExtractor.swift  | 28 +++++++++++++++++--
 1 file changed, 25 insertions(+), 3 deletions(-)

diff --git a/Shared/Article Extractor/ArticleExtractor.swift b/Shared/Article Extractor/ArticleExtractor.swift
index a0ab41b5ff..c30ca26678 100644
--- a/Shared/Article Extractor/ArticleExtractor.swift	
+++ b/Shared/Article Extractor/ArticleExtractor.swift	
@@ -38,9 +38,10 @@ public enum ArticleExtractorState: Sendable {
 
 		let clientURL = "https://extract.feedbin.com/parser"
 		let username = SecretKey.mercuryClientID
-		let signature = articleLink.hmacUsingSHA1(key: SecretKey.mercuryClientSecret)
+		let extractionLink = ArticleExtractor.extractionURL(for: articleLink)
+		let signature = extractionLink.hmacUsingSHA1(key: SecretKey.mercuryClientSecret)
 
-		if let base64URL = articleLink.data(using: .utf8)?.base64EncodedString() {
+		if let base64URL = extractionLink.data(using: .utf8)?.base64EncodedString() {
 			let fullURL = "\(clientURL)/\(username)/\(signature)?base64_url=\(base64URL)"
 			if let url = URL(string: fullURL) {
 				self.url = url
@@ -51,7 +52,28 @@ public enum ArticleExtractorState: Sendable {
 		return nil
     }
 
-    public func process() {
+	/// Returns a URL string optimized for extraction, applying site-specific transformations where needed.
+	static func extractionURL(for articleLink: String) -> String {
+		guard let url = URL(string: articleLink),
+			  let host = url.host()?.lowercased() else {
+			return articleLink
+		}
+
+		// Naver Blog desktop URLs use a JavaScript-heavy SPA that extractors can't parse.
+		// The mobile site (m.blog.naver.com) renders as static HTML and works correctly.
+		if host == "blog.naver.com" {
+			var components = URLComponents(url: url, resolvingAgainstBaseURL: false)
+			components?.host = "m.blog.naver.com"
+			components?.query = nil
+			if let mobileURL = components?.url {
+				return mobileURL.absoluteString
+			}
+		}
+
+		return articleLink
+	}
+
+	public func process() {
 
         state = .processing