go/adbc/driver/snowflake: GetObjects panics while retrieving table columns of vector type

[frbvianna]

What happened?

When a table contains a column of VECTOR data type, the driver panics:
panic: runtime error: invalid memory address or nil pointer dereference

To my understanding, this is not a supported type by this driver or Go clients in general.

However, the driver could return a clearer error message.

Stack Trace

panic: runtime error: invalid memory address or nil pointer dereference
[signal SIGSEGV: segmentation violation code=0x2 addr=0x20 pc=0x103f614c0]

goroutine 74 [running]:
github.com/apache/arrow-adbc/go/adbc/driver/internal/driverbase.ToXdbcDataType({0x0, 0x0})
        /Users/<redacted>/go/pkg/mod/github.com/apache/arrow-adbc/go/adbc@v1.4.0/driver/internal/driverbase/connection.go:703 +0x20
github.com/apache/arrow-adbc/go/adbc/driver/snowflake.(*connectionImpl).GetObjects.func4()
        /Users/<redacted>/go/pkg/mod/github.com/apache/arrow-adbc/go/adbc@v1.4.0/driver/snowflake/connection.go:368 +0x280
created by github.com/apache/arrow-adbc/go/adbc/driver/snowflake.(*connectionImpl).GetObjects in goroutine 1
        /Users/<redacted>/go/pkg/mod/github.com/apache/arrow-adbc/go/adbc@v1.4.0/driver/snowflake/connection.go:354 +0x164c
exit status 2

How can we reproduce the bug?

• Create the table described in this doc: https://docs.snowflake.com/en/sql-reference/data-types-vector#loading-and-unloading-vector-data
• Send a GetObjects at ObjectDepthColumns targetting that table

Environment/Setup

No response

[zeroshade]

Absolutely correct, it should have a better error message. I'll take a look next week at updating this to provide a better error.

[frbvianna]

@zeroshade thank you for the quick feedback. Have a nice weekend!

[zeroshade]

Okay so it's an easy fix to avoid the failure. But it's really annoying because the data is actually fully supported by the ADBC driver, coming back as a FixedSizeList, but I can't find any way from the metadata to determine the vector size of type of the vector from what is available in INFORMATION_SCHEMA. COLUMNS :(