Logical fixup: drive earliest-crawl discovery from latest URL instead…

… of interim recipe URL; this matches the description of the intended behaviour in the comment

reciperadar/workers/recipes.py

@@ -138,7 +138,7 @@ def crawl_recipe(url):
         return
 
     # Find the first-known crawl for the latest URL, and consider it the origin
-    earliest_crawl = recipe_url.find_earliest_crawl()
+    earliest_crawl = latest_crawl.find_earliest_crawl()
     if not earliest_crawl:
         print(f"Failed to find earliest crawl for url={url}")
         return