#890: strip leading track-number prefix from filename-derived titles

Files named '01 - Sun It Rises.flac' with no embedded title tag leaked the stem,
number and all, into tracks.title as '01 - Sun It Rises' — which never matches the
canonical 'Sun It Rises', so the real track reads as a false 'missing' and albums
sort wrong.

New conservative strip_leading_track_number (paths.py): removes a clear track-number
prefix (zero-padded number, OR a number followed by a real separator+space) while
leaving titles that merely start with a number untouched — '7 Rings', '99 Luftballons',
'50 Ways to Leave Your Lover', '1-800-273-8255', '1979' all preserved. Never reduces
to empty/bare-number/punctuation.

Applied at:
- get_import_clean_title (context.py) — the universal resolver every import path funnels
  through, so the DB title AND the re-written embedded tag come out clean.
- album_matching scorer — so '01 - Sun It Rises' scores against 'Sun It Rises' and the
  file matches its real track (inheriting the clean canonical name).

27 targeted tests + 772 imports/matching green.

core/imports/album_matching.py

@@ -156,8 +156,11 @@ def score_file_against_track(
     score = 0.0
 
     # Title similarity (TITLE_WEIGHT). Falls back to filename stem when
-    # the file has no title tag.
+    # the file has no title tag — strip a leading track-number prefix off that
+    # stem (#890) so "01 - Sun It Rises" scores against "Sun It Rises".
     title = file_tags.get('title') or os.path.splitext(os.path.basename(file_path))[0]
+    from core.imports.paths import strip_leading_track_number
+    title = strip_leading_track_number(title)
     track_name = track.get('name', '')
     score += similarity(title, track_name) * TITLE_WEIGHT
 

core/imports/context.py

@@ -183,7 +183,12 @@ def get_import_clean_title(
     if not title:
         track_info = get_import_track_info(context)
         title = _first_value(track_info, "name", "title", default="")
-    return str(title or default)
+    title = str(title or default)
+    # #890: strip a leading track-number prefix that leaked from a filename stem
+    # (e.g. "01 - Sun It Rises" → "Sun It Rises") so it matches the canonical title.
+    # Conservative — clean source titles ("7 Rings" etc.) pass through untouched.
+    from core.imports.paths import strip_leading_track_number
+    return strip_leading_track_number(title)
 
 
 def get_import_clean_album(

core/imports/paths.py

@@ -158,6 +158,33 @@ def clean_track_title(track_title: str, artist_name: str) -> str:
     return cleaned if cleaned else original
 
 
+# A leading track-number prefix is EITHER a zero-padded number (01, 04, 099 — no
+# real song title starts with one), OR a plain number followed by a real separator
+# AND a space ("3 - ", "12. "). Deliberately NOT a bare "number space word", so it
+# leaves "7 Rings", "99 Luftballons", "50 Ways to Leave Your Lover" and
+# "1-800-273-8255" untouched.
+_TRACK_NUM_PREFIX_RE = re.compile(r"^\s*(?:0\d{1,2}[\s._)\-]*|\d{1,3}\s*[._)\-]\s+)(?=\S)")
+
+
+def strip_leading_track_number(title: str) -> str:
+    """Conservatively remove a leading track-number prefix from a track title.
+
+    Fixes #890 — files named ``01 - Sun It Rises.flac`` whose stem leaks into the
+    title as ``01 - Sun It Rises``, which then never matches the canonical
+    ``Sun It Rises`` (false "missing"). Only strips an unambiguous track-number
+    prefix; a coincidental leading number that's part of the title is preserved, and
+    it never reduces a title to empty or a bare number."""
+    s = (title or "").strip()
+    if not s:
+        return title or ""
+    stripped = _TRACK_NUM_PREFIX_RE.sub("", s, count=1).strip()
+    # Keep the original if stripping left nothing real — empty, a bare number, or
+    # only punctuation (e.g. "01 - " → "-"). A real title has a letter/digit.
+    if stripped.isdigit() or not re.search(r"[^\W_]", stripped):
+        return s
+    return stripped
+
+
 
 
 def get_album_type_display(raw_type, track_count) -> str:

tests/imports/test_track_number_strip.py

@@ -0,0 +1,77 @@
+"""#890: a leading track number leaking from a filename stem into the title
+("01 - Sun It Rises") makes the track never match the canonical "Sun It Rises",
+so it reads as a false "missing". strip_leading_track_number removes the prefix —
+conservatively, so titles that merely START with a number are left alone.
+"""
+
+from __future__ import annotations
+
+import pytest
+
+from core.imports.context import get_import_clean_title
+from core.imports.paths import strip_leading_track_number
+
+
+# ── the bug: track-number prefixes get stripped ───────────────────────────────
+@pytest.mark.parametrize("dirty,clean", [
+    ("01 - Sun It Rises", "Sun It Rises"),          # the screenshot
+    ("04 - Tiger Mountain Peasant Song", "Tiger Mountain Peasant Song"),
+    ("05 - Quiet Houses", "Quiet Houses"),
+    ("07 - Heard Them Stirring", "Heard Them Stirring"),
+    ("01 Sun It Rises", "Sun It Rises"),            # zero-padded, no separator
+    ("3 - Title", "Title"),                          # plain number + separator + space
+    ("12. Some Song", "Some Song"),                  # dot separator
+    ("10 - Track Ten", "Track Ten"),
+    ("09) Closing Time", "Closing Time"),            # paren separator
+    ("  02 -  Spaced Out ", "Spaced Out"),           # messy whitespace
+])
+def test_strips_track_number_prefix(dirty, clean):
+    assert strip_leading_track_number(dirty) == clean
+
+
+# ── the guard: titles that legitimately start with a number are UNTOUCHED ──────
+@pytest.mark.parametrize("title", [
+    "7 Rings",
+    "99 Luftballons",
+    "50 Ways to Leave Your Lover",
+    "1-800-273-8255",          # number-with-dashes is part of the title
+    "1979",
+    "9 to 5",
+    "4 Minutes",
+    "8 Mile",
+    "21 Guns",
+    "24 Hour Party People",     # no separator → not a track number
+    "0 to 100",
+    "Sun It Rises",             # no leading number at all
+])
+def test_preserves_real_titles(title):
+    assert strip_leading_track_number(title) == title
+
+
+# ── degenerate inputs ─────────────────────────────────────────────────────────
+def test_never_reduces_to_empty_or_bare_number():
+    assert strip_leading_track_number("01") == "01"      # bare number → keep
+    assert strip_leading_track_number("01 - ") == "01 -"  # nothing left → keep original (trimmed)
+    assert strip_leading_track_number("") == ""
+    assert strip_leading_track_number(None) == ""
+
+
+def test_only_strips_one_prefix():
+    # A title that legitimately follows the number keeps its own leading number.
+    assert strip_leading_track_number("01 - 24 Hour Party People") == "24 Hour Party People"
+
+
+# ── the chokepoint: every import path resolves its title through here ──────────
+def test_get_import_clean_title_strips_filename_leak():
+    # original_search['title'] came from the filename stem (no embedded tag).
+    ctx = {"original_search_result": {"title": "01 - Sun It Rises"}}
+    assert get_import_clean_title(ctx) == "Sun It Rises"
+
+
+def test_get_import_clean_title_leaves_clean_source_title():
+    ctx = {"original_search_result": {"title": "7 Rings"}}
+    assert get_import_clean_title(ctx) == "7 Rings"
+
+
+def test_get_import_clean_title_default_untouched():
+    assert get_import_clean_title({}, default="Unknown Track") == "Unknown Track"