Fix for invalid cid while checking directory_is_multihash (#721)

* Fix for invalid cid while checking directory_is_multihash

* Newline

* lint

* print multihash in exception

* Change IPLD in favor if CID

discovery-provider/src/tasks/tracks.py

@@ -246,10 +246,18 @@ def parse_track_event(
         # if cover_art CID is of a dir, store under _sizes field instead
         if track_record.cover_art:
             logger.warning(f"tracks.py | Processing track cover art {track_record.cover_art}")
-            is_directory = update_task.ipfs_client.multihash_is_directory(track_record.cover_art)
-            if is_directory:
-                track_record.cover_art_sizes = track_record.cover_art
-                track_record.cover_art = None
+            try:
+                is_directory = update_task.ipfs_client.multihash_is_directory(track_record.cover_art)
+                if is_directory:
+                    track_record.cover_art_sizes = track_record.cover_art
+                    track_record.cover_art = None
+            except Exception as e:
+                # we are unable to get the cover art
+                if 'invalid multihash' in str(e):
+                    track_record.cover_art_sizes = None
+                    track_record.cover_art = None
+                else:
+                    raise e
 
         update_remixes_table(session, track_record, track_metadata)
 

discovery-provider/src/utils/ipfs_lib.py

@@ -1,6 +1,7 @@
 import logging
 import json
 import time
+import re
 from urllib.parse import urlparse, urljoin
 import requests
 from requests.exceptions import ReadTimeout
@@ -164,6 +165,10 @@ def cat(self, multihash):
             raise  # error is of type ipfshttpclient.exceptions.TimeoutError
 
     def multihash_is_directory(self, multihash):
+        # Check if the multihash is valid
+        if not self.cid_is_valid(multihash):
+            raise Exception(f'invalid multihash {multihash}')
+
         # First attempt to cat multihash locally.
         try:
             # If cat successful, multihash is not directory.
@@ -232,3 +237,9 @@ def update_cnode_urls(self, cnode_endpoints):
 
     def ipfs_id_multiaddr(self):
         return self._multiaddr
+
+    def cid_is_valid(self, cid):
+        if cid and re.match(r"^Qm[a-zA-Z0-9]{44}$", cid):
+            return True
+
+        return False