From d572207122def28b48ec550a466b2c2a8072d704 Mon Sep 17 00:00:00 2001
From: Felipe <41008398+StrawberryMaster@users.noreply.github.com>
Date: Thu, 18 Dec 2025 14:41:35 +0000
Subject: [PATCH] Improve URL detection for timestamps + fix retry download
 links

---
 lib/wayback_machine_downloader.rb | 15 ++++++++++++++-
 1 file changed, 14 insertions(+), 1 deletion(-)

diff --git a/lib/wayback_machine_downloader.rb b/lib/wayback_machine_downloader.rb
index 3c8bf30..4700f2d 100644
--- a/lib/wayback_machine_downloader.rb
+++ b/lib/wayback_machine_downloader.rb
@@ -717,6 +717,19 @@ class WaybackMachineDownloader
         # resolve full URL (handles relative paths like "../img/logo.png")
         resolved_uri = base_uri + asset_rel_url
 
+        # detect if the asset URL is already a Wayback "web/<timestamp>/.../https://..." embed
+        asset_timestamp = parent_timestamp
+        if resolved_uri.path =~ %r{\A/web/([0-9]{4,})[^/]*/(https?://.+)\z}
+          embedded_ts = $1
+          begin
+            orig_uri = URI($2)
+            resolved_uri = orig_uri
+            asset_timestamp = embedded_ts.to_i
+          rescue URI::InvalidURIError
+            # fall back to original resolved_uri and parent timestamp
+          end
+        end
+
         # filter out navigation links (pages) vs assets
         # skip if extension is empty or looks like an HTML page
         path = resolved_uri.path
@@ -753,7 +766,7 @@ class WaybackMachineDownloader
 
         new_file_info = {
           file_url: asset_wbm_url,
-          timestamp: parent_timestamp,
+          timestamp: asset_timestamp,
           file_id: asset_id
         }