beam-commits mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "ASF GitHub Bot (JIRA)" <j...@apache.org>
Subject [jira] [Work logged] (BEAM-5376) Row interface doesn't support nullability on all fields.
Date Thu, 13 Sep 2018 16:29:00 GMT

     [ https://issues.apache.org/jira/browse/BEAM-5376?focusedWorklogId=144002&page=com.atlassian.jira.plugin.system.issuetabpanels:worklog-tabpanel#worklog-144002
]

ASF GitHub Bot logged work on BEAM-5376:
----------------------------------------

                Author: ASF GitHub Bot
            Created on: 13/Sep/18 16:28
            Start Date: 13/Sep/18 16:28
    Worklog Time Spent: 10m 
      Work Description: akedin commented on a change in pull request #6383: [BEAM-5376] Support
nullability on all Row types
URL: https://github.com/apache/beam/pull/6383#discussion_r217450564
 
 

 ##########
 File path: sdks/java/core/src/test/java/org/apache/beam/sdk/schemas/JavaBeanSchemaTest.java
 ##########
 @@ -155,10 +155,10 @@ public void testRecursiveGetters() throws NoSuchSchemaException {
 
     Row nestedRow = row.getRow("nested");
     assertEquals("string", nestedRow.getString("str"));
-    assertEquals((byte) 1, nestedRow.getByte("aByte"));
-    assertEquals((short) 2, nestedRow.getInt16("aShort"));
-    assertEquals((int) 3, nestedRow.getInt32("anInt"));
-    assertEquals((long) 4, nestedRow.getInt64("aLong"));
+    assertEquals((byte) 1, (Object) nestedRow.getByte("aByte"));
 
 Review comment:
   It is unfortunate that there's `assertEquals(long, long)` overload that is applicable here.
   
   Correct way of doing `Byte.valueOf()` is `Byte.valueOf((byte) 1)`, which is ugly. Probably
even uglier way is `(Byte)(byte)1`. Another ugly approach is to create a local wrapper method
`void myAssertEquals(Object a, Object b) { assertEquals(a, b); }`.
   
   I mentioned `assertTrue()` as an example, not suggesting to use it.
   
   To Rui's point, though, if you look at how `Byte.equals()` works, it accepts an `Object`
and casts it to `Byte`, so in this case for semantics it shouldn't matter whether to cast
it at all, to cast it to `Byte` or `Object`. Casting here only allows to unambiguously choose
the `assertEquals(Object, Object)` overload over `assertEquals(long, long)`, and I don't have
an idea how to do that cleanly, so I don't mind either your approach or any of the proposed
in the discussion.

----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on 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


Issue Time Tracking
-------------------

    Worklog Id:     (was: 144002)
    Time Spent: 1.5h  (was: 1h 20m)

> Row interface doesn't support nullability on all fields.
> --------------------------------------------------------
>
>                 Key: BEAM-5376
>                 URL: https://issues.apache.org/jira/browse/BEAM-5376
>             Project: Beam
>          Issue Type: Improvement
>          Components: dsl-sql
>            Reporter: Andrew Pilloud
>            Assignee: Andrew Pilloud
>            Priority: Major
>          Time Spent: 1.5h
>  Remaining Estimate: 0h
>
> For example: 
> {code:java}
> public boolean getBoolean(int idx);{code}
> Should be:
> {code:java}
> public Boolean getBoolean(int idx);{code}



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)

Mime
View raw message