topic Re: tMap instead of tReplicate + tFilterRow? in Talend Studio

tMap instead of tReplicate + tFilterRow?

Anonymous — Wed, 31 May 2017 08:37:28 GMT

I have some table rows which I'm sorting into a series of tables based on the contents of a column (location).

My brute force method (which works) is to use tReplicate, then a separate duplicate Main threads to tFilterRow. After seeing that tMap has the ability to create multiple output flows, I've been digging through documentation to see if that could be used to output multiple filtered output flows. Is this possible?

EDIT: Sample Table - Goal is to output the location A lines to one output stream, location B to one, location C to another.

|=-------+--------+-------=|
|Location|Name    |Quantity|
|=-------+--------+-------=|
|A       |Johnson |5       |
|A       |Jackson |12      |
|B       |Johnson |23      |
|B       |Jackson |2       |
|C       |Williams|33      |
|C       |Johnson |23      |
|C       |Jackson |8       |
'--------+--------+--------'

Re: tMap instead of tReplicate + tFilterRow?

vapukov — Wed, 31 May 2017 09:20:06 GMT

Hi,

yes, of course - it possible

I prefer tMap as well because it allow not only split/filter flows, but at same time make transformations

what better - always depends from current task

Re: tMap instead of tReplicate + tFilterRow?

Anonymous — Wed, 31 May 2017 10:21:40 GMT

Great! How do you set up multiple filters?

Re: tMap instead of tReplicate + tFilterRow?

vapukov — Wed, 31 May 2017 10:30:01 GMT

each input and output flow have filter tools

You can combine conditions by && or || as many as You need

Re: tMap instead of tReplicate + tFilterRow?

Anonymous — Wed, 31 May 2017 10:46:48 GMT

Found what I was looking for in the retired forum thread: https://community.talend.com/t5/Design-and-Development/If-Statement-tMap-Expression-Builder/td-p/2066

In the example table (which I added to the original post), the tMap would look like this:

Re: tMap instead of tReplicate + tFilterRow?

Anonymous — Wed, 31 May 2017 10:47:47 GMT

You must have posted while I was typing my findings. Thanks for your help!