Fix colspan text + box row fields for GridTable compatibility

Colspan: use original word-block text instead of split cell texts.
Prevents "euros a nd cents" from split_cross_column_words.

Box rows: add is_header field (was undefined, causing GridTable
rendering issues). Add y_min_px/y_max_px to header_only rows.
These missing fields caused empty rows with only row numbers visible.

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

klausur-service/backend/grid_editor_helpers.py

@@ -1296,22 +1296,11 @@ def _detect_colspan_cells(
                 is_part_of_span = True
                 # Only emit the merged cell for the FIRST column in the span
                 if ci == span["cols"][0]:
-                    # Collect all cells in this span
-                    span_cells = [c for c in cells
-                                  if c.get("row_index") == ri
-                                  and c.get("col_index") in span["cols"]]
-                    # Merge texts (skip if same text repeated)
-                    texts = []
-                    for sc in sorted(span_cells, key=lambda c: c.get("col_index", 0)):
-                        t = sc.get("text", "").strip()
-                        if t and t not in texts:
-                            texts.append(t)
-                    merged_text = " ".join(texts)
-
-                    # Collect all word_boxes
-                    all_wb = []
-                    for sc in span_cells:
-                        all_wb.extend(sc.get("word_boxes", []))
+                    # Use the ORIGINAL word-block text (not the split cell texts
+                    # which may have broken words like "euros a" + "nd cents")
+                    orig_word = span["word"]
+                    merged_text = orig_word.get("text", "").strip()
+                    all_wb = [orig_word]
 
                     # Compute merged bbox
                     if all_wb: