Empty struct's value change since version 1.3.0 when calling to_dicts

[HistoireDeBabar]

Checks

• I have checked that this issue has not already been reported.
• I have confirmed this bug exists on the latest version of Polars.

Reproducible example

This is the current behaviour (>= 1.5.0):

import polars as pl
pl.__version__
# '1.9.0' (happens on from 1.5.0 onwards. 1.3.0 works (see below)

df = pl.DataFrame([{'empty': {}}])
df.to_dicts()
# [{'empty': None}]

df.schema
#  Schema([('empty', Struct({'': Null}))])

However in previous versions the following happened:

import polars as pl
pl.__version__
# '1.3.0' 

df = pl.DataFrame([{'empty': {}}])
df.to_dicts()
# [{'empty': {'': None}}]

df.schema
#  Schema([('empty', Struct({'': Null}))])

Log output

No response

Issue description

Hello,

I've came across an issue when converting a data frame with nested data, to a dictionary. This behaviour has changed since 1.3.0. I can't find any reference to the updates in the release notes (so I think it's unintentional).

I'm curious if the change in behaviour is deliberate, and the context around the change.

Expected behavior

Ideally, I think calling to_dicts should return an empty dict in this case.

import polars as pl

df = pl.DataFrame([{'empty': {}}])
df.to_dicts()
# [{'empty': {}}]

Installed versions

--------Version info---------
Polars:              1.9.0
Index type:          UInt32
Platform:            macOS-14.7-arm64-arm-64bit
Python:              3.9.12 (main, May 20 2022, 16:50:25) 
[Clang 13.1.6 (clang-1316.0.21.2.3)]

----Optional dependencies----
adbc_driver_manager  <not installed>
altair               <not installed>
cloudpickle          <not installed>
connectorx           <not installed>
deltalake            <not installed>
fastexcel            <not installed>
fsspec               <not installed>
gevent               <not installed>
great_tables         <not installed>
matplotlib           <not installed>
nest_asyncio         1.6.0
numpy                1.26.4
openpyxl             <not installed>
pandas               <not installed>
pyarrow              17.0.0
pydantic             <not installed>
pyiceberg            <not installed>
sqlalchemy           <not installed>
torch                <not installed>
xlsx2csv             <not installed>
xlsxwriter           <not installed>```

</details>

[ritchie46]

Both were wrong. This is now fixed in #19123