From 843de8a45ef3fe150e7e816fd9f200ab281ada18 Mon Sep 17 00:00:00 2001
From: BoulderBadgeDad <batmanesignal@gmail.com>
Date: Wed, 3 Jun 2026 16:26:29 -0700
Subject: [PATCH] Similar Artists worker: guarantee every stored similar has a
 source id

Verified against live data: 1312/1313 stored similars carry a metadata source id,
but 1 slipped through name-only (a match on a source with no id column, e.g.
discogs). Enforce the standard: process_artist now SKIPS any similar whose match
doesn't map to a storable id column (spotify/itunes/deezer/musicbrainz) instead
of writing a useless name-only row. Regression test covers discogs-match + no-id
cases. Now 100% of newly-stored similars are actionable.
---
 core/similar_artists_worker.py       |  7 +++++++
 tests/test_similar_artists_worker.py | 16 ++++++++++++++++
 2 files changed, 23 insertions(+)

diff --git a/core/similar_artists_worker.py b/core/similar_artists_worker.py
index 283e05a5..282ae360 100644
--- a/core/similar_artists_worker.py
+++ b/core/similar_artists_worker.py
@@ -108,6 +108,13 @@ def process_artist(
         if not name:
             continue
         kwargs = map_payload_to_store_kwargs(s)
+        if not kwargs:
+            # The match resolved to a source with no id column in similar_artists
+            # (e.g. discogs). Storing it name-only would be useless — you can't
+            # navigate/explore/download it. Enforce the standard: every stored
+            # similar carries a metadata source id, or we skip it.
+            logger.debug("Skipping similar '%s' (matched %s — no storable source id)", name, s.get('source'))
+            continue
         try:
             ok = store_similar(
                 source_artist_id=source_artist_id,
diff --git a/tests/test_similar_artists_worker.py b/tests/test_similar_artists_worker.py
index 364f0f14..3f496041 100644
--- a/tests/test_similar_artists_worker.py
+++ b/tests/test_similar_artists_worker.py
@@ -115,6 +115,22 @@ def test_process_artist_error_when_fetch_raises_carries_detail():
     assert status == 'error' and count == 0 and 'musicmap down' in detail
 
 
+def test_process_artist_skips_similars_without_a_storable_source_id():
+    # Every stored similar MUST carry a metadata source id (spotify/itunes/deezer/
+    # musicbrainz) — otherwise it's not actionable. A match on a source with no id
+    # column (e.g. discogs) is skipped, never stored name-only.
+    store, calls = _capture_store()
+    payload = {'success': True, 'similar_artists': [
+        {'name': 'KeepMe', 'id': 'sp1', 'source': 'spotify'},
+        {'name': 'DropMe', 'id': 'dg1', 'source': 'discogs'},   # no id column → must skip
+        {'name': 'NoId', 'source': 'spotify'},                  # no id at all → must skip
+    ]}
+    status, count, _ = w.process_artist('S', 'A', lambda n, l: payload, store)
+    assert count == 1 and len(calls) == 1
+    assert calls[0]['similar_artist_name'] == 'KeepMe'
+    assert calls[0]['similar_artist_spotify_id'] == 'sp1'
+
+
 def test_process_artist_skips_unstorable_but_counts_real():
     # A store() that fails for one row shouldn't abort the rest.
     calls = []