From 12e943044301183871e6ddd7585aba661a944d66 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Rapha=C3=ABl=20Vinot?= <raphael@vinot.info>
Date: Wed, 16 Oct 2024 15:25:12 +0200
Subject: [PATCH] chg: Add logging to debug why some files are not downloadable

---
 lookyloo/lookyloo.py    | 5 +++++
 website/web/__init__.py | 4 +++-
 2 files changed, 8 insertions(+), 1 deletion(-)

diff --git a/lookyloo/lookyloo.py b/lookyloo/lookyloo.py
index 7cb71910..c3601de2 100644
--- a/lookyloo/lookyloo.py
+++ b/lookyloo/lookyloo.py
@@ -1118,12 +1118,15 @@ class Lookyloo():
             url = self.get_urlnode_from_tree(tree_uuid, urlnode_uuid)
         except IndexError:
             # unable to find the uuid, the cache is probably in a weird state.
+            self.logger.info(f'Unable to find node "{urlnode_uuid}" in "{tree_uuid}"')
             return None
         except NoValidHarFile as e:
             # something went poorly when rebuilding the tree (probably a recursive error)
             self.logger.warning(e)
             return None
+
         if url.empty_response:
+            self.logger.info(f'The response for node "{urlnode_uuid}" in "{tree_uuid}" is empty.')
             return None
         if not h or h == url.body_hash:
             # we want the body
@@ -1131,11 +1134,13 @@ class Lookyloo():
 
         # We want an embedded ressource
         if h not in url.resources_hashes:
+            self.logger.info(f'Unable to find "{h}" in capture "{tree_uuid}" - node "{urlnode_uuid}".')
             return None
         for mimetype, blobs in url.embedded_ressources.items():
             for ressource_h, blob in blobs:
                 if ressource_h == h:
                     return 'embedded_ressource.bin', BytesIO(blob.getvalue()), mimetype
+        self.logger.info(f'Unable to find "{h}" in capture "{tree_uuid}" - node "{urlnode_uuid}", but in a weird way.')
         return None
 
     def __misp_add_vt_to_URLObject(self, obj: MISPObject) -> MISPObject | None:
diff --git a/website/web/__init__.py b/website/web/__init__.py
index 3d687a13..d7601abe 100644
--- a/website/web/__init__.py
+++ b/website/web/__init__.py
@@ -1542,14 +1542,16 @@ def recapture(tree_uuid: str) -> str | Response | WerkzeugResponse:
 @app.route('/ressource_by_hash/<string:sha512>', methods=['GET'])
 @file_response  # type: ignore[misc]
 def ressource_by_hash(sha512: str) -> Response:
+    content_fallback = f'Unable to find "{sha512}"'
     if uuids := get_indexing(flask_login.current_user).get_hash_uuids(sha512):
         # got UUIDs for this hash
         capture_uuid, urlnode_uuid = uuids
+        content_fallback += f' in capture "{capture_uuid}" and node "{urlnode_uuid}"'
         if ressource := lookyloo.get_ressource(capture_uuid, urlnode_uuid, sha512):
             filename, body, mimetype = ressource
             return send_file(body, as_attachment=True, download_name=filename)
 
-    return send_file(BytesIO(f'Unable to find {sha512}'.encode()), as_attachment=True, download_name='Unknown_Hash.txt')
+    return send_file(BytesIO(content_fallback.encode()), as_attachment=True, download_name='Unknown_Hash.txt')
 
 
 # ################## Submit existing capture ##################