[
https://issues.apache.org/jira/browse/HADOOP-12693?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15086913#comment-15086913
]
Kai Zheng commented on HADOOP-12693:
------------------------------------
This can be corrected incrementally. For the long term, I would suggest we
introduce [AssertJ|http://joel-costigliola.github.io/assertj/] for the new
tests avoiding such mistakes.
> Many misusage of assertEquals(expected, actual)
> -----------------------------------------------
>
> Key: HADOOP-12693
> URL: https://issues.apache.org/jira/browse/HADOOP-12693
> Project: Hadoop Common
> Issue Type: Test
> Reporter: Akihiro Suda
> Priority: Trivial
> Attachments: just-rough-approx.txt
>
>
> The first arg of {{org.JUnit.Assert.assertEquals()}} should be an
> {{expected}} value, and the second one should be an {{actual}} value.
> {code}
> void assertEquals(T expected, T actual);
> {code}
> http://junit.org/apidocs/org/junit/Assert.html#assertEquals(java.lang.Object,
> java.lang.Object)
> However, there are so many violations in Hadoop, which can make a misleading
> message like this:
> {code}
> AssertionError: expected:<actual> but was:<expected>
> {code}.
> Please refer to {{just-rough-approx.txt}}.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
