fix(ocr-pipeline): preserve sub-column data in vocab table display

Three fixes for sub-columns disappearing at end of streaming:

1. Backend: add column_marker mapping in _cells_to_vocab_entries()
   so marker text is included in vocab entries (not silently dropped)

2. Frontend types: add source_page and bbox_ref to WordEntry interface

3. Frontend table: show page_ref column (Seite) in vocab table when
   entries have source_page data, instead of only EN/DE/Example

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

admin-lehrer/app/(admin)/ai/ocr-pipeline/types.ts

@@ -180,11 +180,13 @@ export interface WordEntry {
   english: string
   german: string
   example: string
+  source_page?: string
   confidence: number
   bbox: WordBbox
   bbox_en: WordBbox | null
   bbox_de: WordBbox | null
   bbox_ex: WordBbox | null
+  bbox_ref?: WordBbox | null
   status?: 'pending' | 'confirmed' | 'edited' | 'skipped'
 }
 

admin-lehrer/components/ocr-pipeline/StepWordRecognition.tsx

@@ -518,11 +518,15 @@ export function StepWordRecognition({ sessionId, onNext, goToStep }: StepWordRec
               {/* Entry/Cell table */}
               <div className="max-h-80 overflow-y-auto">
                 {isVocab ? (
-                  /* Vocab table: EN/DE/Example columns */
+                  /* Vocab table: EN/DE/Example columns + optional page_ref */
+                  (() => {
+                    const hasPageRef = editedEntries.some(e => e.source_page)
+                    return (
                       <table className="w-full text-xs">
                         <thead className="sticky top-0 bg-white dark:bg-gray-800">
                           <tr className="text-left text-gray-500 dark:text-gray-400 border-b dark:border-gray-700">
                             <th className="py-1 pr-2 w-8">#</th>
+                            {hasPageRef && <th className="py-1 pr-2 w-12 text-gray-400">Seite</th>}
                             <th className="py-1 pr-2">English</th>
                             <th className="py-1 pr-2">Deutsch</th>
                             <th className="py-1 pr-2">Example</th>
@@ -539,6 +543,11 @@ export function StepWordRecognition({ sessionId, onNext, goToStep }: StepWordRec
                               onClick={() => { setActiveIndex(idx); setMode('labeling') }}
                             >
                               <td className="py-1 pr-2 text-gray-400">{idx + 1}</td>
+                              {hasPageRef && (
+                                <td className="py-1 pr-2 font-mono text-gray-400 dark:text-gray-500">
+                                  {entry.source_page || ''}
+                                </td>
+                              )}
                               <td className="py-1 pr-2 font-mono text-gray-700 dark:text-gray-300 cursor-pointer">
                                 <MultilineText text={entry.english} />
                               </td>
@@ -555,6 +564,8 @@ export function StepWordRecognition({ sessionId, onNext, goToStep }: StepWordRec
                           ))}
                         </tbody>
                       </table>
+                    )
+                  })()
                 ) : (
                   /* Generic table: dynamic columns from columns_used */
                   <table className="w-full text-xs">

klausur-service/backend/cv_vocab_pipeline.py

@@ -4215,12 +4215,14 @@ def _cells_to_vocab_entries(
         'column_de': 'german',
         'column_example': 'example',
         'page_ref': 'source_page',
+        'column_marker': 'marker',
     }
     bbox_key_map = {
         'column_en': 'bbox_en',
         'column_de': 'bbox_de',
         'column_example': 'bbox_ex',
         'page_ref': 'bbox_ref',
+        'column_marker': 'bbox_marker',
     }
 
     # Group cells by row_index
@@ -4238,12 +4240,14 @@ def _cells_to_vocab_entries(
             'german': '',
             'example': '',
             'source_page': '',
+            'marker': '',
             'confidence': 0.0,
             'bbox': None,
             'bbox_en': None,
             'bbox_de': None,
             'bbox_ex': None,
             'bbox_ref': None,
+            'bbox_marker': None,
             'ocr_engine': row_cells[0].get('ocr_engine', '') if row_cells else '',
         }