Labs API upgrade fixes

Some uses of the labs api endpoints upgraded recently went un-noticed. Fixing
those.

listenbrainz/db/playlist.py

@@ -788,8 +788,8 @@ def get_playlist_recordings_metadata(mb_curs, ts_curs, playlist: Playlist) -> Pl
 
     for rec, row in zip(playlist.recordings, rows):
         rec.artist_credit = row.get("artist_credit_name", "")
-        if "[artist_credit_mbids]" in row and row["[artist_credit_mbids]"] is not None:
-            rec.artist_mbids = [UUID(mbid) for mbid in row["[artist_credit_mbids]"]]
+        if "artist_credit_mbids" in row and row["artist_credit_mbids"] is not None:
+            rec.artist_mbids = [UUID(mbid) for mbid in row["artist_credit_mbids"]]
         rec.title = row.get("recording_name", "")
         rec.release_name = row.get("release_name", "")
         rec.duration_ms = row.get("length", "")

listenbrainz/db/popularity.py

@@ -219,7 +219,7 @@ def get_top_recordings_for_artist(db_conn, ts_conn, artist_mbid, count=None):
             data.pop("original_recording_mbid", None)
             data.update({
                 "artist_name": data.pop("artist_credit_name"),
-                "artist_mbids": data.pop("[artist_credit_mbids]"),
+                "artist_mbids": data.pop("artist_credit_mbids"),
                 "total_listen_count": recording["total_listen_count"],
                 "total_user_count": recording["total_user_count"],
                 "release_color": releases_color.get(str(data["release_mbid"]), {})

listenbrainz/labs_api/labs/api/base_mbid_mapping.py

@@ -18,6 +18,7 @@ class BaseMBIDMappingOutput(BaseModel):
     artist_mbids: Optional[list[UUID]]
     release_mbid: Optional[UUID]
     recording_mbid: Optional[UUID]
+    match_type: int
 
 
 class BaseMBIDMappingQuery(Query):

listenbrainz/labs_api/labs/tests/test_mbid_mapping.py

@@ -98,6 +98,7 @@
         "artist_mbids": ["a3cb23fc-acd3-4ce0-8f36-1e5aa6a18432"],
         "artist_credit_name": "U2",
         "index": 0,
+        "match_type": 4,
         "recording_arg": "gloria",
         "recording_mbid": "398a5f12-80ba-4d29-8b6b-bfe2176341a6",
         "recording_name": "Gloria",
@@ -110,6 +111,7 @@
         "artist_mbids": ["8f6bd1e4-fbe1-4f50-aa9b-94c450ec0f11"],
         "artist_credit_name": "Portishead",
         "index": 1,
+        "match_type": 4,
         "recording_arg": "strangers",
         "recording_mbid": "e97f805a-ab48-4c52-855e-07049142113d",
         "recording_name": "Strangers",
@@ -122,6 +124,7 @@
         "artist_mbids": ["8f6bd1e4-fbe1-4f50-aa9b-94c450ec0f11"],
         "artist_credit_name": "Portishead",
         "index": 2,
+        "match_type": 1,
         "recording_arg": "glory box (feat. your mom)",
         "recording_mbid": "145f5c43-0ac2-4886-8b09-63d0e92ded5d",
         "recording_name": "Glory Box",
@@ -137,6 +140,7 @@
         "artist_mbids": ["8f6bd1e4-fbe1-4f50-aa9b-94c450ec0f11"],
         "artist_credit_name": "Portishead",
         "index": 0,
+        "match_type": 3,
         "recording_arg": "strangers a",
         "recording_mbid": "e97f805a-ab48-4c52-855e-07049142113d",
         "recording_name": "Strangers",
@@ -171,7 +175,7 @@ def test_basics(self):
             q.outputs().__fields__.keys(),
             ['index', 'artist_credit_arg', 'recording_arg',  'artist_credit_name',
              'artist_mbids', 'release_name', 'recording_name', 'release_mbid', 'recording_mbid',
-             'artist_credit_id'])
+             'artist_credit_id', 'match_type'])
 
     @patch('typesense.documents.Documents.search')
     def test_fetch(self, search):
@@ -180,7 +184,6 @@ def test_fetch(self, search):
         q = MBIDMappingQuery()
         resp = q.fetch(json_request_0, RequestSource.json_post)
         self.assertEqual(len(resp), 3)
-        self.maxDiff = None
         self.assertDictEqual(json.loads(resp[0].json()), json_response_0[0])
         self.assertDictEqual(json.loads(resp[1].json()), json_response_0[1])
         self.assertDictEqual(json.loads(resp[2].json()), json_response_0[2])

listenbrainz/mbid_mapping_writer/matcher.py

@@ -1,13 +1,12 @@
-import uuid
-from operator import itemgetter
+from operator import attrgetter
 
 import sqlalchemy
 import psycopg2
-from flask import current_app
-from psycopg2.extras import execute_values
-from listenbrainz.labs_api.labs.api.mbid_mapping import MBIDMappingQuery
-from listenbrainz.mbid_mapping_writer.mbid_mapper import MATCH_TYPES, MATCH_TYPE_NO_MATCH, MATCH_TYPE_EXACT_MATCH
-from listenbrainz.labs_api.labs.api.artist_credit_recording_lookup import ArtistCreditRecordingLookupQuery
+from datasethoster import RequestSource
+from listenbrainz.labs_api.labs.api.mbid_mapping import MBIDMappingQuery, MBIDMappingInput
+from listenbrainz.mbid_mapping_writer.mbid_mapper import MATCH_TYPES, MATCH_TYPE_EXACT_MATCH
+from listenbrainz.labs_api.labs.api.artist_credit_recording_lookup import ArtistCreditRecordingLookupQuery, \
+    ArtistCreditRecordingLookupInput
 from listenbrainz.db import timescale
 
 
@@ -69,7 +68,7 @@ def process_listens(app, listens, priority):
     for typ in MATCH_TYPES:
         stats[typ] = 0
 
-    msids = {str(listen['recording_msid']): listen for listen in listens}
+    msids = {str(listen["recording_msid"]): listen for listen in listens}
     stats["total"] = len(msids)
     if priority == NEW_LISTEN:
         stats["listen_count"] += len(msids)
@@ -108,8 +107,8 @@ def process_listens(app, listens, priority):
             # For all listens that are not matched, enter a no match entry, so we don't
             # keep attempting to look up more listens.
             for listen in remaining_listens:
-                matches.append((listen['recording_msid'], None, None, None, None, None, None, None, MATCH_TYPES[0]))
-                stats['no_match'] += 1
+                matches.append((listen["recording_msid"], None, None, None, None, None, None, None, MATCH_TYPES[0]))
+                stats["no_match"] += 1
 
             stats["processed"] += len(matches)
 
@@ -199,40 +198,46 @@ def lookup_listens(app, listens, stats, exact, debug):
         return [], [], stats
 
     if debug:
-        app.logger.info(f"""Lookup (exact {exact}) '{listens[0]["data"]["artist_name"]}', '{listens[0]["data"]["track_name"]}'""")
+        app.logger.info(f"""Lookup (exact {exact}) "{listens[0]["data"]["artist_name"]}", "{listens[0]["data"]["track_name"]}" """)
 
     if exact:
         q = ArtistCreditRecordingLookupQuery(debug=debug)
+        ModelT = ArtistCreditRecordingLookupInput
     else:
         q = MBIDMappingQuery(timeout=SEARCH_TIMEOUT, remove_stop_words=True, debug=debug)
+        ModelT = MBIDMappingInput
 
     params = []
     for listen in listens:
-        params.append({'[artist_credit_name]': listen["track_metadata"]["artist_name"],
-                       '[recording_name]': listen["track_metadata"]["track_name"]})
+        params.append(ModelT(
+            artist_credit_name=listen["track_metadata"]["artist_name"],
+            recording_name=listen["track_metadata"]["track_name"]
+        ))
 
     rows = []
-    hits = q.fetch(params)
-    for hit in sorted(hits, key=itemgetter("index"), reverse=True):
-        listen = listens[hit["index"]]
+    hits = q.fetch(params, RequestSource.json_post)
+    for hit in sorted(hits, key=attrgetter("index"), reverse=True):
+        listen = listens[hit.index]
 
         if exact:
-            hit["match_type"] = MATCH_TYPE_EXACT_MATCH
-        stats[MATCH_TYPES[hit["match_type"]]] += 1
-        rows.append((listen['recording_msid'],
-                     hit["recording_mbid"],
-                     hit["release_mbid"],
-                     hit["release_name"],
-                     hit["artist_mbids"],
-                     hit["artist_credit_id"],
-                     hit["artist_credit_name"],
-                     hit["recording_name"],
-                     MATCH_TYPES[hit["match_type"]]))
+            match_type = MATCH_TYPE_EXACT_MATCH
+        else:
+            match_type = hit.match_type
+        stats[MATCH_TYPES[match_type]] += 1
+        rows.append((listen["recording_msid"],
+                     str(hit.recording_mbid),
+                     str(hit.release_mbid),
+                     hit.release_name,
+                     [str(artist_mbid) for artist_mbid in hit.artist_mbids],
+                     hit.artist_credit_id,
+                     hit.artist_credit_name,
+                     hit.recording_name,
+                     MATCH_TYPES[match_type]))
 
         if debug:
             app.logger.info("\n".join(q.get_debug_log_lines()))
 
-        listens.pop(hit["index"])
+        listens.pop(hit.index)
         if len(listens) == 0:
             break
 

listenbrainz/webserver/views/metadata_api.py

@@ -1,12 +1,13 @@
 from brainzutils.ratelimit import ratelimit
+from datasethoster import RequestSource
 from flask import Blueprint, request, jsonify, current_app
 
 from listenbrainz.db.mbid_manual_mapping import create_mbid_manual_mapping, get_mbid_manual_mapping
 from listenbrainz.db.metadata import get_metadata_for_recording, get_metadata_for_artist, get_metadata_for_release_group
 from listenbrainz.db.model.mbid_manual_mapping import MbidManualMapping
 from listenbrainz.labs_api.labs.api.artist_credit_recording_lookup import ArtistCreditRecordingLookupQuery
 from listenbrainz.labs_api.labs.api.artist_credit_recording_release_lookup import \
-    ArtistCreditRecordingReleaseLookupQuery
+    ArtistCreditRecordingReleaseLookupQuery, ArtistCreditRecordingReleaseLookupInput
 from listenbrainz.mbid_mapping_writer.mbid_mapper import MBIDMapper
 from listenbrainz.webserver import ts_conn
 from listenbrainz.webserver.decorators import crossdomain
@@ -210,11 +211,11 @@ def metadata_release_group():
 
 
 def process_results(match, metadata, incs):
-    recording_mbid = match["recording_mbid"]
+    recording_mbid = str(match["recording_mbid"])
     result = {
         "recording_mbid": recording_mbid,
-        "release_mbid": match["release_mbid"],
-        "artist_mbids": match["artist_mbids"],
+        "release_mbid": str(match["release_mbid"]),
+        "artist_mbids": [str(artist_mbid) for artist_mbid in match["artist_mbids"]],
         "recording_name": match["recording_name"],
         "release_name": match["release_name"],
         "artist_credit_name": match["artist_credit_name"]
@@ -263,24 +264,24 @@ def get_mbid_mapping():
     incs = parse_incs(request.args.get("inc")) if metadata else []
 
     params = [
-        {
-            "[artist_credit_name]": artist_name,
-            "[recording_name]": recording_name,
-            "[release_name]": release_name
-        }
+        ArtistCreditRecordingReleaseLookupInput(
+            artist_credit_name=artist_name,
+            recording_name=recording_name,
+            release_name=release_name or "",
+        )
     ]
 
     try:
         if release_name:
             q = ArtistCreditRecordingReleaseLookupQuery(debug=False)
-            exact_results = q.fetch(params)
+            exact_results = q.fetch(params, RequestSource.json_post)
             if exact_results:
-                return process_results(exact_results[0], metadata, incs)
+                return process_results(exact_results[0].dict(), metadata, incs)
 
         q = ArtistCreditRecordingLookupQuery(debug=False)
-        exact_results = q.fetch(params)
+        exact_results = q.fetch(params, RequestSource.json_post)
         if exact_results:
-            return process_results(exact_results[0], metadata, incs)
+            return process_results(exact_results[0].dict(), metadata, incs)
 
         q = MBIDMapper(timeout=10, remove_stop_words=True, debug=False, retry_on_timeout=False)
         fuzzy_result = q.search(artist_name, recording_name, release_name)