Skip to main content

I have the following list as an example:

Original list:

mylist{0}.attr1 = a1

mylist{0}.attr2 = a2

mylist{0}.attr3 = a3

mylist{1}.attr1 = b1

mylist{1}.attr2 = b2

mylist{1}.attr3 = b3

mylist{2}.attr1 = c1

mylist{2}.attr2 = c2

mylist{2}.attr3 = c3

mylist{3}.attr1 = d1

mylist{3}.attr2 = d2

mylist{3}.attr3 = d3

mylist{4}.attr1 = e1

mylist{4}.attr2 = e2

mylist{4}.attr3 = e3

 

I want to regroup this list elements by recreating multiple lists by grouping the elements into 2 pairs while taking into account the previous element index as follows:

 

New regrouping of elements:

lstGroup1{0}.attr1 = a1

lstGroup1{0}.attr2 = a2

lstGroup1{0}.attr3 = a3

lstGroup1{1}.attr1 = b1

lstGroup1{1}.attr2 = b2

lstGroup1{1}.attr3 = b3

 

lstGroup2{0}.attr1 = b1

lstGroup2{0}.attr2 = b2

lstGroup2{0}.attr3 = b3

lstGroup2{1}.attr1 = c1

lstGroup2{1}.attr2 = c2

lstGroup2{1}.attr3 = c3

 

lstGroup3{0}.attr1 = c1

lstGroup3{0}.attr2 = c2

lstGroup3{0}.attr3 = c3

lstGroup3{1}.attr1 = d1

lstGroup3{1}.attr2 = d2

lstGroup3{1}.attr3 = d3

 

lstGroup4{0}.attr1 = d1

lstGroup4{0}.attr2 = d2

lstGroup4{0}.attr3 = d3

lstGroup4{1}.attr1 = e1

lstGroup4{1}.attr2 = e2

lstGroup4{1}.attr3 = e3

I tried creating a custom looping transformer using a Sampler, but not succeeding, may be a Python code can be easier to do the task..please help.

Note the challenge is the original list length can be varied (dynamic) could be 2, 3, 4, ..50+ etc

Do you actually want this output?

lstGroup2{1}.attr1 = b1

lstGroup2{1}.attr2 = b2

lstGroup2{1}.attr3 = b3

lstGroup2{2}.attr1 = c1

lstGroup2{2}.attr2 = c2

lstGroup2{2}.attr3 = c3

 

or this?

lstGroup2{0}.attr1 = b1

lstGroup2{0}.attr2 = b2

lstGroup2{0}.attr3 = b3

lstGroup2{1}.attr1 = c1

lstGroup2{1}.attr2 = c2

lstGroup2{1}.attr3 = c3


oh thanks for the hint @egomm..you're right as the index for the new list will start at 0 , I'll edit the output example..


What is the intention afterwards? It's fairly straightforward to build the list you want in python but you would have to type in all the lstGroup names to expose them. 

import fme
import fmeobjects

def processFeature(feature):
        attr1 = feature.getAttribute('mylist{}.attr1')
        attr2 = feature.getAttribute('mylist{}.attr2')
        attr3 = feature.getAttribute('mylist{}.attr3')
        for i, val in enumerate(attr1):
            if i < len(attr1)-1:
                feature.setAttribute('lstGroup'+str(i+1)+'{0}.attr1',attr1(i])
                feature.setAttribute('lstGroup'+str(i+1)+'{1}.attr1',attr1ii+1])
                feature.setAttribute('lstGroup'+str(i+1)+'{0}.attr2',attr2li])
                feature.setAttribute('lstGroup'+str(i+1)+'{1}.attr2',attr2ui+1])
                feature.setAttribute('lstGroup'+str(i+1)+'{0}.attr3',attr3Âi])
                feature.setAttribute('lstGroup'+str(i+1)+'{1}.attr3',attr3Âi+1])

 


What is the intention afterwards? It's fairly straightforward to build the list you want in python but you would have to type in all the lstGroup names to expose them. 

import fme
import fmeobjects

def processFeature(feature):
        attr1 = feature.getAttribute('mylist{}.attr1')
        attr2 = feature.getAttribute('mylist{}.attr2')
        attr3 = feature.getAttribute('mylist{}.attr3')
        for i, val in enumerate(attr1):
            if i < len(attr1)-1:
                feature.setAttribute('lstGroup'+str(i+1)+'{0}.attr1',attr1(i])
                feature.setAttribute('lstGroup'+str(i+1)+'{1}.attr1',attr1ii+1])
                feature.setAttribute('lstGroup'+str(i+1)+'{0}.attr2',attr2li])
                feature.setAttribute('lstGroup'+str(i+1)+'{1}.attr2',attr2ui+1])
                feature.setAttribute('lstGroup'+str(i+1)+'{0}.attr3',attr3Âi])
                feature.setAttribute('lstGroup'+str(i+1)+'{1}.attr3',attr3Âi+1])

 

many thanks!..the big challenge is I found the original list can be varied in length.. so it could be 4 elements or 50+ in some cases (dynamic list length), so the grouping should accommodate the list length.


Reply