0
Fixed

Unicode support in the CSV connector

Adrian Corston 1 year ago in UNIFYBroker Service updated by Beau Harrison (Senior Product Software Engineer) 1 year ago 4

I have Unicode data in a locker field, sourced from a PowerShell connector, which looks file in the UNIFYBroker UI:

Image 6404

However when I export it to a CSV connector, the non-latin1 character doesn't appear to be correct:

Image 6405

(viewed using Notepad++)

Does the CSV connector support Unicode characters?

Sorry, might have to ask you to bump this up to a Bug instead, and get it looked at sooner rather than later, because this is causing a bigger problem.  After an import subsequent to the export, I end up with this in my connector with an extra record created.

Image 6406

Now the underlying CSV file has two records, but they both seem to have the corrupt form:

Image 6407

So when the next import after this runs, the duplicated lines are now causing a duplicate key error:

Image 6408

Under review

Hi Adrien,

Yes, this is a bug. I've fixed this and provided a patch to the devop team. Once deployed, unicode characters should no longer be mangled in writes.

Hi Beau,

I've tested this in two of my customer environments and the patch is working great.

Thanks!