geminispace.info

commit a8a1abd487ad5dd0587d680a136dda0a8b1fd7a7
parent 0a0d4669dea4557b24f35fdba4177203e782018a
Author: Natalie Pendragon <natpen@natpen.net>
Date:   Wed, 15 Jul 2020 08:20:03 -0400

[serve] Get backlinks from db instead of index

Diffstat:
M
serve/models.py
 | 
23
++++++++++++-----------

1 file changed, 12 insertions(+), 11 deletions(-)
diff --git a/serve/models.py b/serve/models.py
@@ -56,7 +56,7 @@ class GUS():
                 "prompt"       : result["prompt"] if "prompt" in result else "",
                 "highlights"   : self.gemini_highlighter.highlight_hit(result, "content", top=1) if "content" in result and result["content_type"] in ["text/plain", "text/gemini", "text/markdown"] else "",
                 "link_text"    : GUS._get_link_text(result),
-                "backlinks"    : result["backlinks"].split(" ") if "backlinks" in result and result["backlinks"] != "" else [],
+                "backlinks"    : list(set(result["backlinks"].split(" "))) if "backlinks" in result and result["backlinks"] != "" else [],
             } for result in results],
         )
 
@@ -65,17 +65,18 @@ class GUS():
         resource = GeminiResource(url)
         if not resource.is_valid:
             return []
-        backlinks = []
 
-        result = self.searcher.document(url_id=resource.indexable_url.rstrip("/"))
-        if result and "backlinks" in result and result["backlinks"] != "":
-            backlinks.extend(result["backlinks"].split(" "))
-
-        result = self.searcher.document(url_id=resource.indexable_url.rstrip("/") + "/")
-        if result and "backlinks" in result and result["backlinks"] != "":
-            backlinks.extend(result["backlinks"].split(" "))
-
-        return backlinks
+        u = resource.indexable_url.rstrip("/")
+        from_page = Page.alias()
+        to_page = Page.alias()
+        backlinks = (Link
+                     .select(from_page)
+                     .join(from_page, on=(from_page.id == Link.from_page_id))
+                     .join(to_page, on=(to_page.id == Link.to_page_id))
+                     .where(to_page.url << [u, f"{u}/"])
+                     .dicts())
+        backlink_urls = [b["url"] for b in backlinks]
+        return list(set(backlink_urls))
 
 
     def _get_link_text(result):