diff --git a/del_bfz.py b/del_bfz.py
index 35c18fd..27e4170 100644
--- a/del_bfz.py
+++ b/del_bfz.py
@@ -4,7 +4,9 @@
 
 import stardict
 
+# sqlite3 和 csv文件路径，根据实际情况修改
 MYSQLITE = 'ecdictSqlite.db'
+CSVFILE = 'ecdict.csv'
 
 
 def new_inflection(exchange):
@@ -57,7 +59,7 @@ def new_inflection(exchange):
 
 def init_ecdict_sqlite():
 
-    stardict.convert_dict(MYSQLITE, 'ecdict.csv')
+    stardict.convert_dict(MYSQLITE, CSVFILE)
     con = sqlite3.connect(MYSQLITE)
     cur = con.cursor()
 
@@ -80,7 +82,8 @@ def init_ecdict_sqlite():
     con.close()
 
 
-init_ecdict_sqlite()  # 只需运行一次，生成sqlite3 db文件
+def test_del_bfz():
+    init_ecdict_sqlite()  # 只需运行一次，将CSV文件转换为sqlite3 db文件
 
-# 转换SQLITE为csv文件
-stardict.convert_dict('ecdict.csv', MYSQLITE)
+    # 再将sqlite3 db文件转换回csv文件，覆盖原文件
+    stardict.convert_dict(CSVFILE, MYSQLITE)
diff --git a/stardict.py b/stardict.py
index eb5cb92..b0ff332 100644
--- a/stardict.py
+++ b/stardict.py
@@ -66,8 +66,8 @@ def __open (self):
             "definition" TEXT,
             "translation" TEXT,
             "pos" VARCHAR(16),
-            "collins" INTEGER DEFAULT(0),
-            "oxford" INTEGER DEFAULT(0),
+            "collins" INTEGER DEFAULT(NULL),
+            "oxford" INTEGER DEFAULT(NULL),
             "tag" VARCHAR(64),
             "bnc" INTEGER DEFAULT(NULL),
             "frq" INTEGER DEFAULT(NULL),
@@ -398,8 +398,8 @@ def init (self):
             `definition` TEXT,
             `translation` TEXT,
             `pos` VARCHAR(16),
-            `collins` SMALLINT DEFAULT 0,
-            `oxford` SMALLINT DEFAULT 0,
+            `collins` SMALLINT DEFAULT NULL,
+            `oxford` SMALLINT DEFAULT NULL,
             `tag` VARCHAR(64),
             `bnc` INT DEFAULT NULL,
             `frq` INT DEFAULT NULL,
@@ -1774,20 +1774,10 @@ def convert_dict(dstname, srcname):
     for word in src.dumps():
         pc.next()
         data = src[word]
-        x = data['oxford']
-        if isinstance(x, int) or isinstance(x, long):
-            if x <= 0:
-                data['oxford'] = None
-        elif isinstance(x, str) or isinstance(x, unicode):
-            if x == '' or x == '0':
-                data['oxford'] = None
-        x = data['collins']
-        if isinstance(x, int) or isinstance(x, long):
-            if x <= 0:
-                data['collins'] = None
-        elif isinstance(x, str) or isinstance(x, unicode):
-            if x in ('', '0'):
-                data['collins'] = None
+        for i in ['oxford', 'collins', 'frq', 'bnc']:
+            x = data.get(i)
+            if not isinstance(x, (int, long)) or x <=0:
+                data[i] = None
         dst.register(word, data, False)
     dst.commit()
     pc.done()