nested CDTs and Writing Data to Database

Question

Hi, I have multiple CDTs that have various relationships set up. What I am not clear on is how the ids that would enable the records to relate appropriately are passed from the parent CDT to the child CDT? For example, if an app allows you to create a customer and that customer has many phone numbers, I would have a parent CDT with a one to many relationship with a phone number CDT. When a user enters information into the form, how should the customer key flow into the phone number CDT to allow the relationship to work? Is it expected that this occurs automatically, or would you just program for this where a script task runs in the process model to add the applicable customer id into the applicable phone number rows in the child CDT? (i.e. script task using: repeat(count(pv!childCDT),pv!parentCDT.id) 
 
 THANKS!

ManuelHTG · Answer

It should work if you set it up properly. You will need just to save your main CDT and the process should do the rest. I however, never do that. I find messy to handle the nested CDT's and Appian at times does not manage it so well. 
What I do is to save the key of the parent CDT on the child CDT's.

Stefan Helzle · Answer

Did you have a chance to read this: docs.appian.com/.../cdt_design_guidance.html

gangadharr0001 · Answer

ankitt 
 first data store entity just remove save into from input tab and in the output tab stored vales just do data pass back to PV 
 after that you can use script task in output tab just map first cdt primarky key to second cdt of foreign key .then you can continue with store data to second data store entity

meredithd0001 · Answer

lcrostarosa I may not have explained my initial question well but essentially the key linking the child CDT to my parent CDT was not populating as I expected it to. I went away from this method after researching it more extensively. I found what ManuelHTG found, which was that it was not as easy/clean of a methodology as I initially thought it would be. I did end up just using the process model and script task functionality to grab the parent key and put it in the child CDT. This resulted in a much cleaner process and so far has enabled me to do everything I need. So for anyone looking to do something similar, it seems the best method is to just us the process model to grab the parent keys and put them into the child CDTs.

lcrostarosa · Answer

Lets say you are persisting to the &#96;parent&#96; entity, which is a parent of &#96;nest&#96;with the following data: 
 
&#96;&#96;&#96; 
parent = { 
 "foo" : "bar", 
 "nest": { 
 "hello": "world" 
 } 
} 
&#96;&#96;&#96; 
 
If you just put this data into a write to data store entity, and you have CascadeType.ALL, you should have primary keys for both, as well as a matching foreign key in the foo data 
 
so it will look like 
 
parent: 
id: 1 
nest_id: 1 
foo: bar 
 
nest 
id 1 
hello: world