pr update

rqalpha/data/bundle.py

@@ -678,10 +678,6 @@ def _auto_update_task(self, instrument):
         start_date = START_DATE
         try:
             h5 = h5py.File(self._file, "a")
-        except OSError as e:
-            raise OSError(_("File {} update failed, if it is using, please update later, "
-                          "or you can delete then update again".format(self._file))) from e   
-        try:
             if order_book_id in h5:
                 if len(h5[order_book_id][:]) != 0:
                     last_date = datetime.datetime.strptime(str(h5[order_book_id][-1]['trading_dt']), "%Y%m%d").date()
@@ -692,8 +688,8 @@ def _auto_update_task(self, instrument):
                         return
             arr = self._get_array(instrument, start_date)
             if arr is None:
-                arr = np.array([])
                 if order_book_id not in h5:
+                    arr = np.array([])
                     h5.create_dataset(order_book_id, data=arr)
             else:
                 if order_book_id in h5:
@@ -704,6 +700,9 @@ def _auto_update_task(self, instrument):
                     h5.create_dataset(order_book_id, data=data)
                 else:
                     h5.create_dataset(order_book_id, data=arr)
+        except OSError as e:
+            raise OSError(_("File {} update failed, if it is using, please update later, "
+                          "or you can delete then update again".format(self._file))) from e 
         finally:
             h5.close()
 
@@ -716,14 +715,20 @@ def _get_array(self, instrument, start_date):
             dtype = [('trading_dt', 'int')]
             for field in self._fields:
                 dtype.append((field, record.dtype[field]))
-            update_list = []
-            for index, row in df.iterrows():
-                values = [row[field] for field in self._fields]
-                update_data = tuple([convert_date_to_date_int(
-                    self._env.data_proxy._data_source.get_future_trading_date(index)
-                )] + values)
-                update_list.append(update_data)
-            arr = np.array(update_list, dtype=dtype)
+
+            dt_arr = np.array(df.index.tolist()).reshape((-1, 1))
+            dt_arr = np.apply_along_axis(self._get_trading_date, 1, dt_arr)
+            dt_arr = (dt_arr.astype('datetime64[Y]').astype(int) + 1970) * 10000 + (dt_arr.astype('datetime64[M]').astype(int) % 12 + 1) * 100 + (dt_arr.astype('datetime64[D]') - dt_arr.astype('datetime64[M]') + 1)
+            dt_arr.astype(int)
+            arr = np.ones((dt_arr.shape[0], ), dtype=dtype)
+            arr['trading_dt'] = dt_arr
+            for field in self._fields:
+                arr[field] = df[field].values
             return arr
         return None
-
+
+    def _get_trading_date(self, dt):
+        # type: (numpy.ndarray) -> Timestamp
+        dt = dt[0]
+        dt = self._env.data_proxy._data_source.get_future_trading_date(dt)
+        return dt