Fix dataframe auto datatype for empty and 1D inputs

[fengfeng-zi]

Summary

• guard set_auto_datatype() against empty lists and 1D numpy arrays
• fall back to "str" instead of indexing into unsupported shapes
• add regression tests covering both empty list and 1D ndarray inputs

Closes #13294

Testing

• python -m pytest test/components/test_dataframe.py -k "auto_datatype_handles_empty_list_and_1d_ndarray or is_empty"

---

Note

Low Risk
Low risk: adds simple guards in set_auto_datatype() to avoid indexing errors on empty/1D inputs, plus focused regression tests.

Overview
Fixes gr.Dataframe(datatype="auto") inference for unsupported/empty shapes by treating empty lists and 1D/empty numpy arrays as a single "str" datatype instead of indexing into missing columns.

Adds regression tests ensuring datatype falls back to "str" for these inputs.

^{Reviewed by Cursor Bugbot for commit 3c565e8. Bugbot is set up for automated code reviews on this repo. Configure here.}

[cursor]

Cursor Bugbot has reviewed your changes and found 1 potential issue.

^{❌ Bugbot Autofix is OFF. To automatically fix reported issues with cloud agents, enable autofix in the Cursor dashboard.}

^{Reviewed by Cursor Bugbot for commit 3c565e8. Configure here.}

[cursor]

1D non-empty list still crashes in set_auto_datatype

High Severity

The list branch in set_auto_datatype only guards against empty lists but not 1D non-empty lists. The numpy branch correctly guards against both (value.ndim < 2 or value.size == 0), but a 1D list like [1, 2, 3] passes the len(value) == 0 check and reaches for val in value[0], which tries to iterate over an integer, raising a TypeError. The guard needs to also check whether value[0] is itself a list/iterable, mirroring the 1D protection added for numpy arrays.

Additional Locations (1)

• gradio/components/dataframe.py#L588-L589

 

^{Reviewed by Cursor Bugbot for commit 3c565e8. Configure here.}

[freddyaboulton]

Please fix @fengfeng-zi