LAE

Note: This was originally posted by an inactive account. Content was preserved by moving under an admin account.

Originally posted by: gabygyy

hi,

i want to create a decision node.
The story is like this:
we have a graph generic that is reading different files(same extension) with diferent no of columns and names.
For eg:
source file 1 aaa.txt has the columns:
aa,bb,cc,kk
Source file 2 bbb.txt has the columns:
aa,cc,dd,kk
We need to map almost always field bb to a new column ww.
As you can see there are some cases when column bb is not present as a column in the source file.
in this case we need to build something to work like this:
if column is available(we can use a metacheck node) then map bb to ww
if column is not available we need to extract value of bb from column kk and map this value to ww.
The logic nedeed must work individualy with any of the mentioned source files.
Note! any reference in the nodes to an non existing column will cause a failure to that node.
Please advise if someone already did this!

Note: This was originally posted by an inactive account. Content was preserved by moving under an admin account.

Originally posted by: stonysmith

One question before I offer a solution:
if you HAVE the column BB in the file, can that column ever have the NULL value?

If not, there's a fairly simple way to deal with this.

Note: This was originally posted by an inactive account. Content was preserved by moving under an admin account.

Originally posted by: gabygyy

with the info that i have until now about the files / the field bb if it is part of the file structure it should not be null or empty.

Note: This was originally posted by an inactive account. Content was preserved by moving under an admin account.

Originally posted by: stonysmith

This should do the trick for you then.
Notice that on the CAT node, the "union" option has been selected.
Attachments:
Merging Columns.brg

Note: This was originally posted by an inactive account. Content was preserved by moving under an admin account.

Originally posted by: gabygyy

hi, thx for your fast answer but this not cover the full request
"The logic nedeed must work individualy with any of the mentioned source files.
"
first, the files are not loaded together! /same graph must run once to load source 1 and run once again to load source 2 without crash!

Note: This was originally posted by an inactive account. Content was preserved by moving under an admin account.

Originally posted by: ryeh

That's clever, Stony! Here's a "brute force" method. I thought I could use the "try" function, but didn't seem to work for me.
Attachments:
twoFields.brg

Note: This was originally posted by an inactive account. Content was preserved by moving under an admin account.

Originally posted by: ryeh

Gabygyy, regarding your last comment, what you can do is have the second data source be just one field - bb - with no records. That way, if 'bb' does not exist in the first source, a 'bb' column of nulls gets appended. Your input file will go into the first pin of the Cat node.

Note: This was originally posted by an inactive account. Content was preserved by moving under an admin account.

Originally posted by: gabygyy

yes i now the trick with cat but please note that the files are not related these must be treated individualy by the same graph. i will try the graph received.

Note: This was originally posted by an inactive account. Content was preserved by moving under an admin account.

Originally posted by: gabygyy

i think this was the best and fastest answer/ thx guys!
the solution proposed by ryeh, who help me also in the past,passed the first tests!