[
https://issues.apache.org/jira/browse/HADOOP-12266?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14647145#comment-14647145
]
Kengo Seki commented on HADOOP-12266:
-------------------------------------
bq. I'm confused by something. Why is the output from calcdiffs always an
append and not an overwrite?
It's my mistake. Originally, {noformat}>> "${diffout}"{noformat} was in a
for-loop, so the output was needed to be appended. It is no longer required to
be so. Also, the followings should be overwrite.
{code}
${AWK} -F: '{print $NF}' "${orig}" >> "${tmp}.branch"
${AWK} -F: '{print $NF}' "${new}" >> "${tmp}.patch"
{code}
> make test-patch plugins more consistent
> ---------------------------------------
>
> Key: HADOOP-12266
> URL: https://issues.apache.org/jira/browse/HADOOP-12266
> Project: Hadoop Common
> Issue Type: Sub-task
> Components: yetus
> Reporter: Kengo Seki
> Assignee: Kengo Seki
> Attachments: HADOOP-12266.HADOOP-12111.00.patch
>
>
> In response to:
> https://issues.apache.org/jira/browse/HADOOP-12207?focusedCommentId=14637515&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-14637515
> * make *_calcdiffs to be a common function
> * move add_footer_table after all examinations finished
> * replace ${PATCH_BRANCH} in output filenames with "branch-"
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
