Task JSON file

[fj40wdh]

I'm trying to automate creation of the task JSON file, What is the configurations section used for?  Is it just a listing of tables that have changes/transformation made?

"configurations": [{

"name": "dbo.CODE_BODY_LOCATION"

}, {

"name": "dbo.DLR_ASSIGNED"

}, {

"name": "dbo.DLR_INVENTORY"

[SachinB]

Hello @fj40wdh ,

Thank you for reaching out to the Qlik community !

Could you please elaborate more about your requirements, are you looking for an option to extract the task JSON for the existing task? or you want to create new task by importing it? 

Accordingly, we can check and help you.

Regards,

Sachin B

[fj40wdh]

We have a replication from Oracle  to SQL Server, which requires transformation for date columns. There are over 200 hundred tables to transform so this would take a lot time to do through the web console. The plan is to create the task with one table, then add the other tables by scripting the JSON entries. I noticed the configuration section in the JSON file and need info on how to script it. 

[SachinB]

Hello @fj40wdh ,

You can set it to a different datatype using Global transformation for all the tables once.

You may also need to change a specific data type in the source to a different data type in the target for many or all of the tables in the task. Global transformation will allow you to accomplish this without having to define a transformation for each individual table.

Please refer the below user guide for the same

https://help.qlik.com/en-US/replicate/May2023/Content/Global_Common/Content/SharedEMReplicate/Custom...

https://community.qlik.com/t5/Video-Transcripts/STT-Customizing-Data-with-Qlik-Replicate/ta-p/194848...

Regards,

Sachin B

[fj40wdh]

It's not just a data type change. All the columns are Char so it has to be done by table, by column. The columns also need to be manipulated. 

Again what is the configurations section used for?

[Heinvandenheuvel · Accepted solution]

The 'configurations' section is a relatively recent addition to the task JSON (2023.5?)

Given the scale of what you intent to do I'd highly recommend to create a test task with two or three tables and just try with/without the whole section, and with or without a specific test table in the list.

I haven't personally explored it, but because of a desire for Replicate to be backwards compatible with JSON files I am 99% sure you can just leave the whole section. At the same time, since you'll be generating JSON, you might as well just add it, adding each source schema.table.

I have written many script to pick  up  table from CSV files as well as scripts to clone manipulations. Tedious but very doable and more reliable (imho) than going through the UI. I'll attach one such script  (in PERL)  to perhaps give you some ideas. LMK.

Hein.