feat(dict): high-speed offline dictionary with wordforms and zero daemon SQLite mmap

app/database.py

@@ -1,5 +1,6 @@
 import sqlite3
 import contextlib
+from pathlib import Path
 from typing import Generator
 from app import config
 
@@ -174,6 +175,59 @@ def _migrate(conn: sqlite3.Connection):
 
     conn.execute("CREATE INDEX IF NOT EXISTS idx_sessions_user ON sessions(user_id);")
 
+    # Ensure offline_dictionary table exists
+    conn.execute("""
+    CREATE TABLE IF NOT EXISTS offline_dictionary (
+        term TEXT PRIMARY KEY COLLATE NOCASE,
+        translation TEXT NOT NULL,
+        phonetic TEXT DEFAULT '',
+        pos TEXT DEFAULT ''
+    );
+    """)
+    conn.execute("CREATE INDEX IF NOT EXISTS idx_offline_dict_term ON offline_dictionary(term COLLATE NOCASE);")
+
+    # Auto-seed offline dictionary if empty
+    _seed_offline_dictionary(conn)
+
+
+def _seed_offline_dictionary(conn: sqlite3.Connection):
+    """Seeds the offline dictionary from data/offline_dict.tsv.gz if the table is empty."""
+    try:
+        count = conn.execute("SELECT COUNT(*) FROM offline_dictionary").fetchone()[0]
+        if count > 0:
+            return
+    except Exception:
+        return
+
+    archive_candidates = [
+        config.BASE_DIR / "data" / "offline_dict.tsv.gz",
+        config.BASE_DIR / "app" / "data" / "offline_dict.tsv.gz",
+        Path("/app/data/offline_dict.tsv.gz"),
+    ]
+    archive_path = next((p for p in archive_candidates if p.is_file()), None)
+    if not archive_path:
+        return
+
+    import gzip
+
+    batch = []
+    with gzip.open(archive_path, "rt", encoding="utf-8") as f:
+        for line in f:
+            parts = line.rstrip("\r\n").split("\t")
+            if len(parts) == 4:
+                batch.append(parts)
+                if len(batch) >= 10000:
+                    conn.executemany(
+                        "INSERT OR IGNORE INTO offline_dictionary (term, translation, phonetic, pos) VALUES (?, ?, ?, ?)",
+                        batch,
+                    )
+                    batch.clear()
+        if batch:
+            conn.executemany(
+                "INSERT OR IGNORE INTO offline_dictionary (term, translation, phonetic, pos) VALUES (?, ?, ?, ?)",
+                batch,
+            )
+
 
 def init_db():
     config.DB_PATH.parent.mkdir(parents=True, exist_ok=True)
@@ -271,6 +325,14 @@ def init_db():
         );
         CREATE INDEX IF NOT EXISTS idx_books_user ON books(user_id);
         CREATE INDEX IF NOT EXISTS idx_books_user_updated ON books(user_id, updated_at DESC);
+
+        CREATE TABLE IF NOT EXISTS offline_dictionary (
+            term TEXT PRIMARY KEY COLLATE NOCASE,
+            translation TEXT NOT NULL,
+            phonetic TEXT DEFAULT '',
+            pos TEXT DEFAULT ''
+        );
+        CREATE INDEX IF NOT EXISTS idx_offline_dict_term ON offline_dictionary(term COLLATE NOCASE);
         """)
         # Apply migrations for pre-existing databases
         _migrate(conn)

app/dictionary.py

@@ -240,8 +240,40 @@ async def lookup_term(query: str) -> Dict[str, Any]:
             except Exception:
                 pass
 
-    words = query_clean.split()
-    is_sentence = len(words) > 4
+        # Check offline dictionary first (Zero Daemon, < 0.1ms via SQLite mmap)
+        words = query_clean.split()
+        is_sentence = len(words) > 4
+        if not is_sentence:
+            try:
+                off_row = conn.execute(
+                    "SELECT term, translation, phonetic, pos FROM offline_dictionary WHERE term = ? COLLATE NOCASE",
+                    (query_clean,)
+                ).fetchone()
+                if off_row:
+                    raw_trans = off_row["translation"] or ""
+                    tr_list = [t.strip() for t in raw_trans.split(",") if t.strip()]
+                    res = {
+                        "query": query_clean,
+                        "is_sentence": False,
+                        "translation": raw_trans,
+                        "translations_list": tr_list,
+                        "translation_ru": raw_trans,
+                        "phonetic": off_row["phonetic"] or "",
+                        "audio_file": "",
+                        "audio_url": "",
+                        "part_of_speech": off_row["pos"] or "",
+                        "definitions": [],
+                        "examples": [],
+                        "source": "offline",
+                    }
+                    conn.execute(
+                        "INSERT OR REPLACE INTO dictionary_cache (query, response_json) VALUES (?, ?)",
+                        (query_clean, json.dumps(res, ensure_ascii=False))
+                    )
+                    return res
+            except Exception:
+                pass
+
 
     result: Dict[str, Any] = {
         "query": query_clean,