Home | About | Sematext search-lucene.com search-hadoop.com
 Search Hadoop and all its subprojects:

Switch to Plain View
Pig, mail # user - Best Practice: How to encapsulate frequently used native pig code to "methods"


+
Markus Resch 2012-03-14, 12:41
Copy link to this message
-
Re: Best Practice: How to encapsulate frequently used native pig code to "methods"
Norbert Burger 2012-03-14, 13:02
Pig has support for macros; take a look at DEFINE (macro) and IMPORT here:

http://pig.apache.org/docs/r0.9.1/cont.html#define-macros

Norbert

On Wed, Mar 14, 2012 at 8:41 AM, Markus Resch <[EMAIL PROTECTED]>wrote:

> Hi everyone,
>
> In future we are going to have a bunch of Pig scripts in parallel which
> are doing different stuff. In front of that special code we are going to
> have a very common "pre-filtering part" which I don't like to copy
> manually all over. I also don't like to use UDFs (as long as I can avoid
> them) because of my concerns about the optimization. I think I can doo
> all that filtering in native Pig code.
> What is the most common way to encapsulate such stuff in some kind of
> "methods" wich are returning relations just like the native calls do?
>
>
> Thanks
>
> Markus
>
>
>