[
https://issues.apache.org/jira/browse/FLINK-2275?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14602726#comment-14602726
]
ASF GitHub Bot commented on FLINK-2275:
---------------------------------------
Github user StephanEwen commented on the pull request:
https://github.com/apache/flink/pull/866#issuecomment-115652087
All in all this looks good.
The only think that strikes me is the of the test pattern(that way used
also before by the tests. I am not a big fan of validating results in an
`@After` method. I find the tests more readable if they simply call the
comparison themselves at the end:
```java
List<Tuple2<String, Integer>> result = data.collect()
String expected = "Hello,2\n" +
"Dude,5\n"
compareResultAsTuples(expexted, result)
```
This also solves the issue with the raw type lists, which is usually an
indicator that something is not modeled in the right way (unless you
dynamically load or instantiate types).
> Migrate test from package 'org.apache.flink.test.javaApiOperators'
> ------------------------------------------------------------------
>
> Key: FLINK-2275
> URL: https://issues.apache.org/jira/browse/FLINK-2275
> Project: Flink
> Issue Type: Sub-task
> Components: Tests
> Affects Versions: 0.10
> Reporter: Matthias J. Sax
> Assignee: Matthias J. Sax
> Priority: Minor
>
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
