and more in a single search tool across platforms. Read the announcement here. |
04/21/2023 11:55 AM
Use case is to be able to capture a user's firstname and lastname in the user registration form and to normalize and propagate to different attributes within the identity.
The original firstname would end up in a customproperty while the normalized value would end up in firstname, stripped of all accents and other inappropriate characters. A similar logic would apply to lastname.
What is the optimal way to handle this from EIC?
version 23.3
Solved! Go to Solution.
04/23/2023 04:42 PM
You can map customproperty in users column in user register form dynamic attribute also you can create hidden attribute and map in actual attribute ( first name) . You can normalize using sql query
04/23/2023 05:23 PM - edited 04/23/2023 05:24 PM
Hi @rushikeshvartak ,
Thanks for the quick reply. I managed to get the simple part set up where the form's fields go to my customproperty.
I am stuck with the normalization. Would you have a simple sql statement that would allow to normalize the values? Earlier experimentation yielded something absolutely horrendous involving nearly 40 nested replace() statements. I (and the client) would greatly appreciate something a tad more elegant.
Thanks again!
François
04/23/2023 08:13 PM
Give me some sample data
04/23/2023 08:20 PM
Let's say: firstname: Étienne, lastname: L'espérance
Expected end result would be "Etienne" and "L'esperance"
Thanks again!
François
04/23/2023 08:31 PM
Check if below helps
04/24/2023 05:36 AM
Hi,
I don't think it would work, as I would like to replace all accented characters by their non-accented value, whereas regexp replace would replace all characters matching the pattern with a single value
Also, not sure if I am doing anything wrong but calls to this routine are denied on my test tenant: "Error : execute command denied to user '[user]' for routine 'saviyntazure.REGEXP_REPLACE'"
Thanks!
04/26/2023 02:29 AM
Looked left and right and had to resort to using nested replace statements. I still don't like it but it seems to work...