You can use multi storage to write out separate files based on a grouping
column ..
you would need to first make a unified data set with one of the columns
(may be the 1st one ..as below) having the grouping/file name needed ..
e.g.
C1
20140801,{..some content},{..some content}...,
20140802,{..some content},{..some content}....
:
:
then use
STORE slias INTO '$path' USING
org.apache.pig.piggybank.storage.MultiStorage('path','0', 'none', '|');
http://svn.apache.org/repos/asf/pig/trunk/contrib/piggybank/java/src/main/java/org/apache/pig/piggybank/storage/MultiStorage.java
Cheers !!!
Arvind
On 15-Dec-2014 11:12 pm, "Margus Roo" <mar...@roo.ee> wrote:
> Hi
>
> I have files contain timestamp. I'd like to parse row by row and put them
> into file by timestamp.
> in example
>
> original file:
> 20140801,...,...,...,...,...
> 20140802,...,...,...,...,...
> 20140801,...,...,...,...,...
> ...
>
> So I'd like to parse this rows to separate files 20140801 and 20140802 so
> that file
> 20140801.csv contains:
> 20140801,...,...,...,...,...
> 20140801,...,...,...,...,...
>
> and 20140802.csv contains
> 20140802,...,...,...,...,...
>
> I tried to write my own custom StorageFunc but as much I understand I can
> not do it there.
> I read about MultyStorage maybe this is the right tool to try? Or Pig
> totally wrong tool for that problem?
>
> --
> Margus (margusja) Roo
> http://margus.roo.ee
> skype: margusja
> +372 51 480
>
>
