fix(dictionary): strip Latin romanization debris from Russian translations and sanitize dictionary cache

app/cefr.py

@@ -15,6 +15,7 @@ import sqlite3
 import random
 from typing import Dict, Any, List, Optional
 from datetime import datetime
+from app.dictionary import clean_russian_translation
 
 CEFR_LEVELS = ["A1", "A2", "B1", "B2"]
 
@@ -2190,7 +2191,7 @@ def process_cefr_answer(
             "SELECT term, translation, phonetic, pos FROM offline_dictionary WHERE term = ? COLLATE NOCASE",
             (selected_clean,),
         ).fetchone()
-        sel_trans = sel_row["translation"] if sel_row else selected_clean
+        sel_trans = clean_russian_translation(sel_row["translation"]) if sel_row else selected_clean
         sel_phon = sel_row["phonetic"] if sel_row else ""
         sel_pos = sel_row["pos"] if sel_row else ""
         sel_ctx = resolve_context_sentence_for_term(conn, selected_clean)
@@ -2244,7 +2245,7 @@ def process_cefr_answer(
                 "SELECT term, translation, phonetic, pos FROM offline_dictionary WHERE term = ? COLLATE NOCASE",
                 (opt_clean,),
             ).fetchone()
-            opt_trans = opt_row["translation"] if opt_row else opt_clean
+            opt_trans = clean_russian_translation(opt_row["translation"]) if opt_row else opt_clean
             opt_phon = opt_row["phonetic"] if opt_row else ""
             opt_pos = opt_row["pos"] if opt_row else ""
 
@@ -2413,7 +2414,7 @@ def ingest_cefr_word(
         "SELECT term, translation, phonetic, pos FROM offline_dictionary WHERE term = ? COLLATE NOCASE",
         (clean_term,),
     ).fetchone()
-    trans = row["translation"] if row else clean_term
+    trans = clean_russian_translation(row["translation"]) if row else clean_term
     phon = row["phonetic"] if row else ""
     pos = row["pos"] if row else ""