Make the artist a gate, not a quarter of the score

Requesting "Michael Jackson - Chicago" filed The Fall's "Chicago, Now!" from
Extricate. The chain: the artist name is filtered by the Soulseek server, so
the ladder correctly fell back to the bare title -- at which point the result
set is every song anyone has ever called Chicago, and the artist is the only
thing telling them apart.

match_ratio weights the artist at 0.25, so a perfect title with no artist
anywhere scores 0.75. That clears MIN_MATCH, and the ~22 point gap it leaves
against a genuine match is nothing next to the 340 points score() awards for a
free upload slot and a short queue. An idle peer holding the wrong song
therefore beat a busy peer holding the right one, every time.

That is the failure the confidence score was explicitly designed to prevent --
"a fast peer must not be able to raise the confidence of a wrong match" -- and
rewriting score() to rank availability first reintroduced it on the other side
of the wall. Availability is a tiebreak among plausible files; it is never a
reason to take a different song.

Candidates whose path names the artist are now kept outright and the rest
discarded, falling back to title-only matches only when there are none. The
downstream confidence gate still ran correctly here: it scored 0.75, said
"artist not found in path", and held the file for review.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

app.py

@@ -481,6 +481,30 @@ def similarity(want: str, got: str) -> float:
     return max(seq, len(a & b) / len(a) if a else 0.0)
 
 
+def artist_confirmed(remote_path: str, artist: str) -> bool:
+    """Is the requested artist actually named somewhere in this path?
+
+    This is a gate rather than a score because of what the query ladder does.
+    Once a filtered artist name forces the search down to the bare title, the
+    result set stops being "this song from various peers" and becomes "every
+    song anyone ever called Chicago" -- The Fall, Groove Armada, Sufjan
+    Stevens. The artist is then the only thing distinguishing the right
+    recording from a completely different song, and a weighted contribution is
+    not enough: 0.75 for a perfect title with no artist still clears MIN_MATCH,
+    and an available stranger will out-score a queued match every time.
+    """
+    na = norm(artist)
+    if not na:
+        return False
+    path = norm(remote_path.replace("\\", "/"))
+    if na in path:
+        return True
+    # Every distinctive word, in any order: catches "Jackson, Michael" and
+    # "MJ - Chicago (Michael Jackson)" without matching on "the" or "and".
+    words = [w for w in na.split() if len(w) > 3]
+    return bool(words) and all(w in path for w in words)
+
+
 def match_ratio(remote_path: str, artist: str, title: str) -> float:
     """How well a remote path matches the track asked for, 0..1.
 
@@ -1078,6 +1102,20 @@ async def pick_and_enqueue(job_id: str, artist: str, title: str,
         set_state(job_id, "failed", no_candidate_detail(seen, stats, queries))
         return
 
+    # Prefer candidates whose path actually names the artist, and only settle
+    # for ones that do not when there is nothing else. Availability is a
+    # tiebreak among plausible files, never a reason to take the wrong song.
+    named = {k: v for k, v in pool.items()
+             if artist_confirmed(v[2]["filename"], artist)}
+    if named:
+        if len(named) < len(pool):
+            log.info("job %s keeping %d/%d candidates that name %r",
+                     job_id[:8], len(named), len(pool), artist)
+        pool = named
+    else:
+        log.info("job %s no candidate names %r; falling back to title-only "
+                 "matches, which review will have to judge", job_id[:8], artist)
+
     ranked = sorted(pool.values(), key=lambda r: -r[0])
     cands = spread_by_peer(ranked)[:CANDIDATES]
     with db() as c: