This is more of a question to explore than an actual issue.
Working with the framework on a project, the ids which are integers at the source, were appended a prefix which transformed them to varchars. The only "issue" is when you're debugging and want to get to the source's id, it's quicker to copy the whole value instead of removing the prefix. Also, I think it might be better practice to stick to the source's natural field type instead of forcing it to varchars.
My preference for this project has been
- to not use prefixes
- name the field as
source_table_natural_key
- keep source's natural field type.
Example:
select
id as platform_transaction_natural_key
┆Issue is synchronized with this Jira Task by Unito
This is more of a question to explore than an actual issue.
Working with the framework on a project, the ids which are integers at the source, were appended a prefix which transformed them to varchars. The only "issue" is when you're debugging and want to get to the source's id, it's quicker to copy the whole value instead of removing the prefix. Also, I think it might be better practice to stick to the source's natural field type instead of forcing it to varchars.
My preference for this project has been
source_table_natural_keyExample:
┆Issue is synchronized with this Jira Task by Unito