Import CSV ; The "schema" field required or not --> 2 datasets for 1 table

[xavier-GitHub76]

Bug description

Description

The "schema" field is not required :

with version 4 value 'public' was set by default but with V5.0.0, the default value is not set

The error message should precise

The import is OK if user set a schema but 2 datasets are existing for 1 table

Screenshots/recordings

No response

Superset version

master / latest-dev

Python version

3.9

Node version

16

Browser

Chrome

Additional context

No response

Checklist

• I have searched Superset docs and Slack and didn't find a solution to my problem.
• I have searched the GitHub issue tracker and didn't find a similar bug report.
• I have checked Superset's logs for errors and if I found a relevant Python stacktrace, I included it here as text in the "additional context" section.

[xavier-GitHub76]

import_public_duplicated_datasets.mp4

[ompharate]

@xavier-GitHub76
I was able to reproduce a related (but slightly stricter) issue on current master during CSV upload.

If I provide a table name but leave the Schema field empty and click Upload, Superset ends up generating SQL like:

CREATE TABLE undefined.algoritham1 (...);

which fails on Postgres with:

psycopg2.errors.InvalidSchemaName: schema "undefined" does not exist

The Schema field doesn’t appear to be required in the UI and shows no validation error, so it’s easy to miss.
If I explicitly select public, everything works as expected.

This feels like a small frontend → backend contract mismatch when schema is undefined.

[rusackas]

So if I understand we just need to restore setting 'public' by default? CC @betodealmeida in case they're aware of any change here.

[tysoncung]

I'm interested in working on this issue. Could you provide more context about what you're looking for? Any additional details about requirements or constraints would be helpful.

[xavier-GitHub76]

I think it would be possible to solve the problem by making the schema mandatory everywhere.

In the database :

• mandatory, and if no value then the value 'public' can be set (for PostgreSQL databases)

In application for forms:

• mandatory, with the default value set to 'public'

These forms are:

• SQL Lab
• creating a physical dataset
• the source tab of the dataset editing screen
• importing CSV, Excel, and column data

With schema always defined, it would no longer be possible to create duplicate datasets pointing to the same target.

[YousufFFFF]

Thanks for the detailed context and screenshots.

Based on the discussion, it seems the preferred fix is to restore the previous behavior by defaulting the schema (e.g. public for Postgres) during CSV import when it’s left empty, rather than making schema mandatory everywhere.

I can work on a PR following this approach (frontend + backend contract fix) unless there are any objections or additional constraints I should consider.

[YousufFFFF]

Confirmed this issue still reproduces on Superset v6.

When uploading a CSV to Postgres with the schema left empty, Superset attempts to create the table under schema = "undefined" and fails with:

psycopg2.errors.InvalidSchemaName: schema "undefined" does not exist

I’m working on a fix to restore the default schema behavior during CSV import, as discussed above.

[vighneshtule]

@xavier-GitHub76 I can take this up. The issue seems to be on the frontend where "undefined" is being sent as the schema value when no schema is selected. I’ll fix this by adding schema to NonNullFields in UploadDataModal so it’s not appended to FormData when empty.

Since the backend already supports optional schema, this should resolve the error. If needed, I can also add a small defensive check in UploadPostSchema to treat empty or "undefined" values as None.

[YousufFFFF]

I’ve opened a PR to address this issue.
PR: #36792

Feedback and review would be appreciated. Thanks!

[aditikh-76]

Hi, I’m new to Apache Superset and would like to work on this issue.
Please let me know if it’s okay for me to take it.