[
https://issues.apache.org/jira/browse/HADOOP-11553?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Allen Wittenauer updated HADOOP-11553:
--------------------------------------
Attachment: HADOOP-11553-00.patch
-00:
* Initial patch, "code-level" only
This patch doesn't change any functions but does:
* puts a stability level, param, etc, comment in front of every function, in a
kind of a fake javadoc format
* re-arranges hadoop-functions to be grouped by those stability levels
Some things I'm looking for here:
* Are these the right settings for these functions? Too liberal? Too
conservative?
* Is re-arranging the right thing to do?
I recognize that the last point isn't something we normally do, breaks patches,
etc, etc. However, in the case of the shell code, since there is one big file
rather than multiple, I think it would be useful for others reading this file.
It forces the most interesting functions closer to the top of a very big file.
To-do:
* Write all the docs (function hint guide, shell profile docs, .hadooprc, etc,
etc)
> Formalize the shell API
> -----------------------
>
> Key: HADOOP-11553
> URL: https://issues.apache.org/jira/browse/HADOOP-11553
> Project: Hadoop Common
> Issue Type: Improvement
> Components: scripts
> Affects Versions: 3.0.0
> Reporter: Allen Wittenauer
> Assignee: Allen Wittenauer
> Priority: Blocker
> Attachments: HADOOP-11553-00.patch
>
>
> After HADOOP-11485, we need to formally document functions and environment
> variables that 3rd parties can expect to be able to exist/use.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
