What happens?

There is some limitation or bug during pivoting table using struct type field.

I'm using the CLI. I would like to pivot a table, more than 6000 columns to produce.
Here is the result :

pivot donnees_csv on variable_id using first(donnee) group by id_niv,codzone;
85% ▕███████████████████████████████████████████████████
[1] 62343 segmentation fault duckdb tests.db

The donnee field is a struct type.

To Reproduce

The table donnees_csv was materialized into a database like this :

create or replace table donnees_csv as (
select build_variable_id(codind,datefin::string) as variable_id,zonage as id_niv,codzone,valeur,
statut,{v:valeur,s:statut} as donnee
from read_csv('variables_donnees.csv', header=true) order by id_niv,codzone);

The build_variable_id is a simple function : create or replace function build_variable_id(codind,datefin) as codind||'_'||replace(datefin,'-','_');

The segmentation fault is coming where pivoting like this :
pivot donnees_csv on variable_id using first(donnee) group by id_niv,codzone;

When i limit the size of the donnees_csv table, it works fine. With the full size of the table, PIVOT has to produce 6000 columns.

Maybe important thing : it works when pivoting using another field than donnee: donnee is a struct type field, where using valeur field (double type) all is fine.

OS:

MacOS

DuckDB Version:

10.0.3

DuckDB Client:

CLI

Full Name:

Frank Jardillier

Affiliation:

compas

What is the latest build you tested with? If possible, we recommend testing with the latest nightly build.

I have tested with a stable release

Did you include all relevant data sets for reproducing the issue?

No - I cannot easily share my data sets due to their large size

Did you include all code required to reproduce the issue?

• Yes, I have

Did you include all relevant configuration (e.g., CPU architecture, Python version, Linux distribution) to reproduce the issue?

• Yes, I have