How to do full outer join

[Anonymous]

Hi Shong, 
I have another question, 
please answer me, 
when I'm triying to merge two CSV files, I would like to do a FULL OUTER JOIN, I would like to have all rows in both files, 
what can I do ? 
thanks 

[Anonymous]

Hi 
Using tMap to do left outer join two times, then unite the two output results and filter the duplicated rows, the job looks like:

[Anonymous]

Hi 
Using tMap to do left outer join two times, then unite the two output results and filter the duplicated rows, the job looks like:
674/mini_1.png_20160126-0324.png

While I am still baffled as to why tMap just doesn't give a Full Outer Join option, in its absence glad to see a workaround. SHong, I am not clear though on how to use the tFilterrow component to eliminate the duplicate rows... can you provide a sample of what the filter conditions would be?
Thanks!

[Anonymous]

Hi Shon,
I followed your post for applying the full outer join, but I am facing difficulty in joining tmap outputs using tunite.
one of the tmap output gets connected but the other one does not. Please view the attachment:

Can you please suggest, what should i do here.

[Anonymous]

Adding Snapshot of my job:

[Anonymous]

Hi sunny240884
I don't see your image, however, I think you are trying to create a cycle flow in the job, it is now allowed to do it. Please take a look at this article and follows the solution.
Regards
Shong

[Anonymous]

Thanks Shong,
I was doing the cycle flow. your article helped me.
Cheers,
Shong

[Anonymous]

Hi Shong,

What if file1 or file2 have some duplicate data in it.

I believe full outer join preserves it but with the solution you suggested, it will remove those duplicates.

How to handle that situation?

---

@shong wrote:
Hi 
Using tMap to do left outer join two times, then unite the two output results and filter the duplicated rows, the job looks like:

---