[jira] [Comment Edited] (HADOOP-12121) smarter branch detection

["Allen Wittenauer (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/HADOOP-12121?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14601230#comment-14601230 ] 

Allen Wittenauer edited comment on HADOOP-12121 at 6/25/15 1:56 PM:
--------------------------------------------------------------------

I've had a few thoughts rolling around in my head on how to make this work better/more predictable.  The key I think is that when we get the patch name, we should perform a few operations on it before passing through a parser:

a) replace .patch, .diff, .txt with a .
b) de-dupe any .
c) strip end-of-filename .

After that, we can make some reasonable assumptions about what's left over.  Additionally, it makes it much easier to build a string that contains multiple periods which can then be passed through verify_valid_branch.


was (Author: aw):
I've had a few thoughts rolling around in my head on how to make this work better/more predictable.  The key I think is that when we get the patch name, we should a few operations on it before passing through a parser:

a) replace .patch, .diff, .txt with a .
b) de-dupe any .
c) strip end-of-filename .

After that, we can make some reasonable assumptions about what's left over.  Additionally, it makes it much easier to build a string that contains multiple periods which can then be passed through verify_valid_branch.

> smarter branch detection
> ------------------------
>
>                 Key: HADOOP-12121
>                 URL: https://issues.apache.org/jira/browse/HADOOP-12121
>             Project: Hadoop Common
>          Issue Type: Sub-task
>          Components: yetus
>    Affects Versions: HADOOP-12111
>            Reporter: Allen Wittenauer
>
> We should make branch detection smarter so that it works on micro versions.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)