Steven Hawkins <shawkins@redhat.com> made a comment on jira TEIID-2189

Our logic isn't really amenable to handling || differently than concat.  This would probably have to be implemented as having both || and concat map to concat2.  That by itself is not a large change especially if it's enabled via a system property and affects the system globally (user queries transformations etc.)  If that's not desirable then we need more information.

Steven Hawkins <shawkins@redhat.com> updated the status of jira TEIID-2189 to Resolved

Steven Hawkins <shawkins@redhat.com> made a comment on jira TEIID-2189

Reasons against this:
- it's non-ansi behavior
- Oracle doesn't even guarantee this behavior in later version
- addressing isn't straight-forward.  We can just change the behavior of the function method.  It would require us to internally differentiate between the || and the concat function and to ensure that all of the internal views (mostly in pg) use the concat function.

So marking as won't fix unless there is an additional driver to add this behavior.

Steven Hawkins <shawkins@redhat.com> updated the status of jira TEIID-2189 to Closed