fix: border ghost filter + row overlap fix for box zones

1. Add _filter_border_ghosts() to grid editor - removes OCR artefacts
   like | sitting on box borders before row/column clustering.
   The tall | (h=55) was inflating row 0's y_max, causing row overlap.

2. Fix _assign_word_to_row() to prefer closest y_center when rows
   overlap, instead of always returning the first matching row.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

klausur-service/backend/cv_words_first.py

@@ -134,12 +134,16 @@ def _assign_word_to_column(word: Dict, columns: List[Dict]) -> int:
 
 
 def _assign_word_to_row(word: Dict, rows: List[Dict]) -> int:
-    """Return row index for a word based on its Y-center."""
+    """Return row index for a word based on its Y-center.
+
+    When rows overlap (e.g. due to tall border-ghost characters inflating
+    a row's y_max), prefer the row whose y_center is closest.
+    """
     y_center = word['top'] + word['height'] / 2
-    # Find the row whose y_range contains this word's center
-    for row in rows:
-        if row['y_min'] <= y_center <= row['y_max']:
-            return row['index']
+    # Find all rows whose y_range contains this word's center
+    matching = [r for r in rows if r['y_min'] <= y_center <= r['y_max']]
+    if matching:
+        return min(matching, key=lambda r: abs(r['y_center'] - y_center))['index']
     # Fallback: nearest row by Y-center
     return min(rows, key=lambda r: abs(r['y_center'] - y_center))['index']