You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@flink.apache.org by GitBox <gi...@apache.org> on 2020/09/24 08:18:01 UTC

[GitHub] [flink] dianfu commented on a change in pull request #13327: [FLINK-19134][python] Add BasicArrayTypeInfo and coder for PrimitiveArrayTypeInfo for Python DataStream API.

dianfu commented on a change in pull request #13327:
URL: https://github.com/apache/flink/pull/13327#discussion_r494116254



##########
File path: flink-python/pyflink/common/typeinfo.py
##########
@@ -384,6 +444,35 @@ def PRIMITIVE_ARRAY(element_type: TypeInformation):
         else:
             raise TypeError("Invalid element type for a primitive array.")
 
+    @staticmethod
+    def BASIC_ARRAY(element_type: TypeInformation):
+        """
+        Returns type information for arrays of boxed primitive type (such as Integer[]).
+
+        :param element_type element type of the array (e.g. Types.BOOLEAN(), Types.INT(),
+        Types.DOUBLE())
+        """
+        if element_type == Types.BOOLEAN():
+            return BasicArrayTypeInfo.BOOLEAN_ARRAY_TYPE_INFO()
+        elif element_type == Types.BYTE():
+            return BasicArrayTypeInfo.BYTE_ARRAY_TYPE_INFO()
+        elif element_type == Types.SHORT():
+            return BasicArrayTypeInfo.SHORT_ARRAY_TYPE_INFO()
+        elif element_type == Types.INT():
+            return BasicArrayTypeInfo.INT_ARRAY_TYPE_INFO()
+        elif element_type == Types.LONG():
+            return BasicArrayTypeInfo.LONG_ARRAY_TYPE_INFO()
+        elif element_type == Types.FLOAT():
+            return BasicArrayTypeInfo.FLOAT_ARRAY_TYPE_INFO()
+        elif element_type == Types.DOUBLE():
+            return BasicArrayTypeInfo.DOUBLE_ARRAY_TYPE_INFO()
+        elif element_type == Types.CHAR():
+            return BasicArrayTypeInfo.CHAR_ARRAY_TYPE_INFO()
+        elif element_type == Types.STRING():
+            return BasicArrayTypeInfo.STRING_ARRAY_TYPE_INFO()
+        else:
+            raise TypeError("Invalid element type for a boxed primitive array.")

Review comment:
       ```suggestion
               raise TypeError("Invalid element type for a boxed primitive array: %s" % str(element_type))
   ```

##########
File path: flink-python/pyflink/fn_execution/coders.py
##########
@@ -32,7 +32,7 @@
 __all__ = ['RowCoder', 'BigIntCoder', 'TinyIntCoder', 'BooleanCoder',
            'SmallIntCoder', 'IntCoder', 'FloatCoder', 'DoubleCoder',
            'BinaryCoder', 'CharCoder', 'DateCoder', 'TimeCoder',
-           'TimestampCoder', 'ArrayCoder', 'MapCoder', 'DecimalCoder']
+           'TimestampCoder', 'BasicArrayCoder', 'MapCoder', 'DecimalCoder']

Review comment:
       Also expose PrimitiveArrayCoder?

##########
File path: flink-python/pyflink/fn_execution/beam/beam_coders.py
##########
@@ -194,7 +194,7 @@ def _to_data_type(field_type):
                                                field_type.nullable)
             elif field_type.type_name == flink_fn_execution_pb2.Schema.TIMESTAMP:
                 return TimestampType(field_type.timestamp_info.precision, field_type.nullable)
-            elif field_type.type_name == flink_fn_execution_pb2.Schema.ARRAY:

Review comment:
       should we also add PRIMITIVE_ARRAY?

##########
File path: flink-python/pyflink/fn_execution/coders.py
##########
@@ -467,8 +478,8 @@ def from_proto(field_type):
     if field_type_name == type_name.LOCAL_ZONED_TIMESTAMP:
         timezone = pytz.timezone(os.environ['table.exec.timezone'])
         return LocalZonedTimestampCoder(field_type.local_zoned_timestamp_info.precision, timezone)
-    elif field_type_name == type_name.ARRAY:
-        return ArrayCoder(from_proto(field_type.collection_element_type))
+    elif field_type_name == type_name.BASIC_ARRAY:

Review comment:
       Should we also support PRIMITIVE_ARRAY here?




----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

For queries about this service, please contact Infrastructure at:
users@infra.apache.org