Hi Christopher,

I may consider your second suggestion, but really prefer to "minimize" the 
parts that I am working with (if you know what I mean :)).

But, about the comment in your first paragraph, I don't see where a reference 
to the IteratorEnvironment is being passed into the VisibilityEvaluator? 

How can I access that from within the VE?

Thanks again,
Jim

--------------------------------------------
On Mon, 8/14/17, Christopher <[email protected]> wrote:

 Subject: Re: Another VisibilityEvaluator question
 To: [email protected], "o haya" <[email protected]>
 Date: Monday, August 14, 2017, 7:43 PM
 
 Not in the
 current implementation. As I understand it, though, you are
 writing an alternate VisibilityEvaluator. The
 IteratorEnvironment that is passed in contains a reference
 to the current table's configuration. It doesn't
 have the table name or id, but it does have its
 configuration, so if you were to insert a configuration
 property into that particular table, you could read it in
 the VisibilityFilter, and modify that to pass it to the
 VisibilityEvaluator.
 
 Alternatively, you could update the
 IteratorEnvironment interface to include a table ID getter
 (or name, but ID is more reliable). This would avoid
 requiring you to put anything in the table's
 configuration, but may require you to modify a more code.
 (Might be a good idea to add this upstream; I created an
 issue: https://issues.apache.org/jira/browse/ACCUMULO-4695)
 
 Either strategy would involve
 getting some information from the environment from within
 the VisibilityFilter, and passing that along to the
 VisibilityEvaluator.
 
 
 On Mon, Aug
 14, 2017 at 12:40 AM o haya <[email protected]>
 wrote:
 Hi,
 
 
 
 I am wondering if there is a way for code inside the
 VisibilityEvaluator to get the name of the
 "current" table (the table that is being
 processed)?
 
 
 
 Thanks,
 
 Jim
 
 

Reply via email to