Fix false header detection: skip continuation lines and mid-column cells

Single-cell rows were incorrectly detected as headings when they were
actually continuation lines. Two new guards:
1. Text starting with "(" is a continuation (e.g. "(usw.)", "(TV-Serie)")
2. Single cells beyond the first two content columns are overflow lines,
   not headings. Real headings appear in the first columns.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

klausur-service/backend/grid_editor_helpers.py

@@ -1053,6 +1053,16 @@ def _detect_heading_rows_by_single_cell(
             text = (cell.get("text") or "").strip()
             if not text or text.startswith("["):
                 continue
+            # Continuation lines start with "(" — e.g. "(usw.)", "(TV-Serie)"
+            if text.startswith("("):
+                continue
+            # Single cell NOT in the first content column is likely a
+            # continuation/overflow line, not a heading.  Real headings
+            # ("Theme 1", "Unit 3: ...") appear in the first or second
+            # content column.
+            first_content_col = col_indices[0] if col_indices else 0
+            if cell.get("col_index", 0) > first_content_col + 1:
+                continue
             # Skip garbled IPA without brackets (e.g. "ska:f – ska:vz")
             # but NOT text with real IPA symbols (e.g. "Theme [θˈiːm]")
             _REAL_IPA_CHARS = set("ˈˌəɪɛɒʊʌæɑɔʃʒθðŋ")