Summary

Several array/list functions from upstream DataFusion are not yet exposed in datafusion-python. This includes new functions and missing list_* aliases for existing array_* functions.

Missing Functions (new)

• array_any_value / list_any_value — returns any non-null element from the array
• array_contains / list_contains — alias for array_has
• array_distance / list_distance — computes distance between two arrays
• array_max / list_max — returns the maximum element
• array_min / list_min — returns the minimum element
• array_reverse / list_reverse — reverses elements in the array
• arrays_overlap — checks if two arrays share any elements
• arrays_zip / list_zip — zips multiple arrays into an array of structs
• generate_series — generates a series of values
• string_to_array / string_to_list — splits a string into an array by delimiter

Missing list_* Aliases

The following list_* aliases exist upstream but are not in __all__:

• list_empty
• list_pop_back
• list_pop_front
• list_has
• list_has_all
• list_has_any

Upstream Reference

• https://datafusion.apache.org/user-guide/sql/scalar_functions.html (Array Functions section)

Implementation

• Rust bindings: crates/core/src/functions.rs
• Python wrappers: python/datafusion/functions.py

Note: This gap analysis was performed using an AI agent comparing upstream DataFusion v53 documentation against the current datafusion-python codebase.