ClaudioV
6 months agoNew Member
Status:
New
SQL: Add OUTER UNION CORR–like operator (union by column name with automatic null padding)
In SAS PROC SQL, the OUTER UNION CORR operator allows vertically concatenating tables with different schemas by aligning columns by name and automatically creating missing columns filled with NULLs.
In Microsoft Fabric (Warehouse / SQL analytics endpoint / Spark SQL), this functionality is missing. Users must manually normalize schemas by explicitly adding NULL AS column_name in each SELECT, which becomes error‑prone and hard to maintain when unioning many heterogeneous datasets.
This feature is particularly important for:
- migrations from SAS to Fabric
- data lake / lakehouse ingestion patterns
- append‑only fact tables from heterogeneous sources
Adding an operator or syntax variant such as:
SELECT * FROM AOUTER UNION BY NAMESELECT * FROM Bwould significantly reduce friction and improve competitiveness with other analytics platforms.
No CommentsBe the first to comment
Recent ideas
Restore "no-reply" option for Power BI Subscriptions
Please add a setting in the Power BI Subscription pane to choose the sender identity. With the recent update the platform forces emails to route through the subscription owner's personal Exchange On...LenaFid7 hours agoRegular VisitorNew300Views52likes11CommentsEnforcing OneLake Security on external semantic model API calls without adding security risks
Problem External applications cannot query Fabric semantic models on a user’s behalf - with OneLake RLS security rules enforced - unless the Fabric API is exposed to any and every client an authoriz...sean_cochran10 hours agoResolver INew21Views0likes0CommentsAllow ownership transfer (“Take over”) for Activator items
Activator rules run under the identity of the item’s creator. When that person leaves the organization and their account is disabled, all rules stop working silently. Unlike other Fabric items, there...ndelcastillo10 hours agoRegular VisitorNew29Views1like0Comments