Issue: Handling large numbers in the Custom Connector Input in GCP Application Integration,

[muhilvarnan-v]

Hi,

Recently, in my team, we have started implementing application integration for one of the workflow. 

During that, we found a interesting problem with the platform. For particular task, we need to pass a integer value something like this 29211504443291. 

But when the task executes, the platform converts the value to '2.9211504443291E13'. Since the input is valid the task is failing. we did tried different ways datamapping, javascript task, but no luck.

i feel, this has to be fixed at the application integration platform. if anyone have workaround and suggestions. please do share

[fengwan]

Are you able to use String to represent the large number?  But this seems like an interesting issue. We will investigate further internally. Thanks. 

[fengwan]

I tried the large number in JS task which plus 1 to the large number. It looks normal. Can you let me know what logic in JS task that failed? Thanks,

[muhilvarnan-v]

HI @fengwan 

Thanks for the responding to my query.

To reproduce the issue, you can follow the following step.

1. Create a Custom Connector -> in my case, i created a custom connector using zendesk openapi spec - Download OpenAPI file 
2. In the zendesk openapi spec, the resources are identified using integer value, not as string. example: ticket_id, macro_id. sometimes these ids will be very large numbers in the zendesk, so this large number should be passed a request params to connector action.
3. in the integration, add a connector task and do a simple get macro by id. 
4. use data mapping/javascript to pass the large id to the task, you will see that large number is converted to different format.
5. the zendesk custom connector will not have the resouce for given id. 

My gut feel is that the problem is at connector task level, where the payload is prepared. i believe this issue will come any type of custom connector where we need to pass large integer value.

If you need, we can connect on zoom call to share my use case.