accumulo-notifications mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Hudson (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (ACCUMULO-871) ColumnVisibilities should do more to create normalized representations of expressions
Date Thu, 22 Nov 2012 09:08:57 GMT

    [ https://issues.apache.org/jira/browse/ACCUMULO-871?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13502664#comment-13502664
] 

Hudson commented on ACCUMULO-871:
---------------------------------

Integrated in Accumulo-Trunk #563 (See [https://builds.apache.org/job/Accumulo-Trunk/563/])
    ACCUMULO-871

Updated the ColumnVisibility class to be more aggressive and consistent when flattening/normalizing
expressions. Labels are dedupe'd and sorted and equivalent subtrees are pruned. (Revision
1412448)

     Result = SUCCESS
ujustgotbilld : 
Files : 
* /accumulo/trunk/core/src/main/java/org/apache/accumulo/core/security/ColumnVisibility.java
* /accumulo/trunk/core/src/test/java/org/apache/accumulo/core/security/ColumnVisibilityTest.java

                
> ColumnVisibilities should do more to create normalized representations of expressions
> -------------------------------------------------------------------------------------
>
>                 Key: ACCUMULO-871
>                 URL: https://issues.apache.org/jira/browse/ACCUMULO-871
>             Project: Accumulo
>          Issue Type: Improvement
>    Affects Versions: 1.5.0
>         Environment: OSX
>            Reporter: William Slacum
>            Assignee: William Slacum
>            Priority: Minor
>         Attachments: col.vis.patch, CVTest.java
>
>
> ColumnVisibilities offer a `flatten()` functionality, which attempts to normalize a given
expression. So, if I had an expression, `b&a&c`, and I `flatten`'d it, I'd get back
`a&b&c`. 
> Through some testing, I found that this is applied and not applied depending on how the
expression is written. For instance, if I had something like `(b&a)&c`, I would get
back something like `c&a&b`.
> It's not much more code to provide a correct normalized form of expressions as well do
some more work to detect and eliminate expressions that boil down to `expr&expr` or `expr|expr`.
> I've attached a sample program that shows some output of what the current capability
is and what I think it should output.

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira

Mime
View raw message