feat: prevent duplicate cards, merge contexts, and display star badge when word is in library

app/database.py

@@ -83,6 +83,75 @@ def _migrate(conn: sqlite3.Connection):
     """)
     conn.execute("CREATE INDEX IF NOT EXISTS idx_books_user ON books(user_id);")
 
+    # --- Deduplicate vocabulary_cards (case-insensitive) and enforce UNIQUE index ---
+    dup_rows = conn.execute("""
+        SELECT user_id, lower(trim(term)) as clean_term, COUNT(*) as cnt
+        FROM vocabulary_cards
+        GROUP BY user_id, lower(trim(term))
+        HAVING cnt > 1
+    """).fetchall()
+
+    for d in dup_rows:
+        u_id = d["user_id"]
+        c_term = d["clean_term"]
+        cards = conn.execute(
+            "SELECT * FROM vocabulary_cards WHERE user_id = ? AND lower(trim(term)) = ? ORDER BY id ASC",
+            (u_id, c_term),
+        ).fetchall()
+        if not cards:
+            continue
+        primary = cards[0]
+        primary_id = primary["id"]
+        merged_contexts = [primary["context_sentence"]] if primary["context_sentence"] else []
+        merged_tags = set([t.strip() for t in (primary["tags"] or "").split(",") if t.strip()])
+        max_reps = primary["srs_reps"] or 0
+        max_interval = primary["srs_interval"] or 0
+        max_ease = primary["srs_ease"] or 2.5
+        latest_review = primary["srs_last_review"]
+        srs_state = primary["srs_state"]
+
+        for dup in cards[1:]:
+            dup_id = dup["id"]
+            if dup["context_sentence"] and dup["context_sentence"] not in merged_contexts:
+                merged_contexts.append(dup["context_sentence"])
+            for t in (dup["tags"] or "").split(","):
+                if t.strip():
+                    merged_tags.add(t.strip())
+            if (dup["srs_reps"] or 0) > max_reps:
+                max_reps = dup["srs_reps"]
+                max_interval = dup["srs_interval"]
+                max_ease = dup["srs_ease"]
+                latest_review = dup["srs_last_review"]
+                srs_state = dup["srs_state"]
+            # Reassign SRS logs to primary card
+            conn.execute("UPDATE srs_logs SET card_id = ? WHERE card_id = ?", (primary_id, dup_id))
+            # Delete duplicate row
+            conn.execute("DELETE FROM vocabulary_cards WHERE id = ?", (dup_id,))
+
+        # Update primary card with merged data
+        conn.execute(
+            """
+            UPDATE vocabulary_cards
+            SET context_sentence = ?, tags = ?, srs_reps = ?, srs_interval = ?, srs_ease = ?,
+                srs_last_review = ?, srs_state = ?
+            WHERE id = ?
+            """,
+            (
+                "\n\n".join(merged_contexts),
+                ", ".join(sorted(merged_tags)),
+                max_reps,
+                max_interval,
+                max_ease,
+                latest_review,
+                srs_state,
+                primary_id,
+            ),
+        )
+
+    conn.execute(
+        "CREATE UNIQUE INDEX IF NOT EXISTS idx_vocab_user_term_nocase ON vocabulary_cards(user_id, term COLLATE NOCASE);"
+    )
+
 
 def init_db():
     config.DB_PATH.parent.mkdir(parents=True, exist_ok=True)