fix(teams): resolve dividers from tab name + full title, not last notice row

Section titles are stacked single-cell rows: the "TipiLAN 2026 CS2" title
sits above notice rows ("If a team withdraws..."), and parse_team_sections
kept only the LAST one, so the notice clobbered the title and resolve_divider
saw no game/year keywords -> None -> teams never positioned.

- parse_team_sections now accumulates all single-cell rows above a header, so
  the game/year title survives alongside the notices.
- _refresh_teams_sync resolves the divider from "<tab name> <section title>",
  so the game is taken reliably from the CS2/LoL worksheet name while the year
  still comes from the title, keeping year-scoped TEAM_DIVIDER_*_2026 vars.
- regression test: title survives notice rows and still resolves.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01XPRsW4tazVtYi2jNzVQkre

core/sheets.py

@@ -362,7 +362,11 @@ def parse_team_sections(rows: list[list]) -> list[TeamSection]:
         lineup_col = _find_col(rows[i], lambda c: c.strip().lower().startswith(_LINEUP_HEADER_PREFIX))
         if name_col is None or lineup_col is None:
             if (text := _merged_title(rows[i])) is not None:
-                title = text
+                # Accumulate every single-cell row above the header, not just the
+                # last one: sheets stack the "TipiLAN 2026 CS2" title above notice
+                # rows ("If a team withdraws..."), and the later notices must not
+                # clobber the title whose keywords resolve_divider needs.
+                title = f"{title} {text}".strip() if title else text
             i += 1
             continue
         rosters: dict[str, list[str]] = {}
@@ -444,7 +448,10 @@ def _refresh_teams_sync() -> dict[str, list[str]]:
     dividers: dict[str, int] = {}
     for ws in spreadsheet.worksheets():
         for section in parse_team_sections(ws.get_all_values()):
-            divider = resolve_divider(section.title)
+            # Match on the tab name too ("CS2"/"LoL"): it names the game reliably
+            # even when the game/year title row is shadowed by notice rows, while
+            # the section title still supplies the year for year-scoped dividers.
+            divider = resolve_divider(f"{ws.title} {section.title}")
             for team, players in section.rosters.items():
                 rosters.setdefault(team, []).extend(players)
                 if divider:

tests/test_team_sync.py

@@ -142,6 +142,22 @@ def test_resolve_divider_no_config_or_no_title_is_none():
     assert sheets.resolve_divider("", DIVIDERS) is None
 
 
+def test_title_survives_notice_rows_below_it():
+    # Real sheets stack the game/year title above single-cell notice rows; the
+    # title's keywords must survive so resolve_divider still matches (regression:
+    # notice rows used to overwrite the title, yielding a None divider).
+    rows = [
+        ["TipiLAN 2026 CS2 Registration Log"] + [""] * 4,
+        ["This log is updated automatically."] + [""] * 4,
+        ["If a team from the Top 32 withdraws, ..."] + [""] * 4,
+        ["No", "Team Name", "Lineup (nickname, citizenship)", "", ""],
+        ["1", "GENESIS", "kapa (EST)", "", ""],
+    ]
+    section = sheets.parse_team_sections(rows)[0]
+    assert "CS2" in section.title and "2026" in section.title
+    assert sheets.resolve_divider(section.title, DIVIDERS) == 1498736834656604251
+
+
 # --- plan_team_positions (pure role-ordering maths) ------------------------
 
 def test_plan_moves_teams_directly_under_their_divider():