Any method for user level CSV import to map issue key?

[JC]

We're on Jira 6.4x and need individual users (no admin access) to import CSV files that map/update existing issues. This is because there is no integration between another outdated ticketing system (but still in active use) and JIRA. We have the issue importer enabled - but the "Issue Key" field isn't exposed. This isn't something we can complete without the use of an importer or automation, the ticket counts are too great.

I know at the admin level its easy to map and issue key, but i'm not seeing the possibility at the user level. I was able to map ParentIDs and i'm assuming doing so I can update subtasks - but how would I update the initial tickets or parent tickets? Ultimately I suppose we can just create 1 ticket as a container to hold our imported 'subtask' tickets but that means we wouldn't be able to attach a subtask to a subtask... which we will occasionally need. 

I have attempted just importing the data, letting JIRA assign a key and prevent duplicates based off of the summary but encountered a bug https://ecosystem.atlassian.net/browse/JIM-1735 that prevents the users from importing anything if a duplicate exists.  Even if this were resolved, it wouldn't allow end users to have tickets updated on import. 

JIRA has to be mature enough that this sort of thing is trivial and i'm missing it - right? 
Any help or words of wisdom is appreciated,
Jacob 

[JC]

Thanks Nic, 
Understanding subtasks with subtaks is impossible means I have to use a container ticket with multiple subtasks.

I have attempted to perform the import a few dozen different ways and im not having any success in updating a subtask that's already been created. It simply creates a duplicate subtask

If my fields are (Just to cover the basics)
Parent ID, Issue ID, Summary

I define a parent ticket, its called CM-1
I don't need to update CM-1, but I do need to attach and/or update its children

Parent ID,Issue ID, Summary

CM-1,,CM2
CM-1, CM2, CM-2
CM-1. 1,CM-2
CM-1,2,CM-2

All of these will create a new subtask of CM1. None of these will update. If the actual key issued to CM-2 is CM-2 it won't update the ticket. Ive inspected the "xml export" option on CM-2 and I see something like 

<item>
 
<title>[CM2-2] CM2</title>
 
 <link>
https://website.net/browse/CM-2
</link>
 
<project id="1199" key="CM">Change Management</project>
 
<description>CM2</description>
 
<environment/>
 
<key id="191150">CM-2</key>
 
<summary>CM2 Subtask</summary>

I have tried to use the "key id" value for the "Issue ID" field when importing and it also fails to update the ticket. 

Any chance you can point out to me what im doing wrong? I have read through the documentation multiple times, for both the admin level and user level bulk create and im not seeing my mistake.
It's helpful to note we need end users (or team leads) to do this, there won't be administrative access on the production env.  

[Nic Brough -Adaptavist-]

>we wouldn't be able to attach a subtask to a subtask

As a minor point, you can't do that at all.  So you can forget that idea.