Using unique aggregations with StringCache replaces values with null

[thomas-tran-de]

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

import polars as pl

df = pl.DataFrame(
    {"number": [1, 1, 2, 2, 3], "letter": ["a", "b", "b", "c", "c"]}
)

with pl.StringCache():
    result = (
        df.cast({"letter": pl.Categorical})
        .group_by("number")
        .agg(
            all=pl.col("letter"),
            unique=pl.col("letter").unique(),
            unique_with_order=pl.col("letter").unique(maintain_order=True),
        )
    )

result.sort("number")

In v 1.17.1 the output is

number	all	unique	unique_with_order
i64	list[cat]	list[cat]	list[cat]
1	["a", "b"]	["a", "b"]	["a", "b"]
2	["b", "c"]	["b", "c"]	["b", "c"]
3	["c"]	["c"]	["c"]

In v1.80.0 the output is

number	all	unique	unique_with_order
i64	list[cat]	list[cat]	list[cat]
1	["a", "b"]	["a", "b"]	["a", "b"]
2	["b", "c"]	["b", null]	["b", "c"]
3	["c"]	["c"]	["c"]

Log output

No response

Issue description

Using the unique operator in an aggregation while the StringCache is active, replaces some values with null.

Expected behavior

Behavior as it was in v1.17.1

Installed versions

--------Version info---------
Polars:              1.18.0
Index type:          UInt32
Platform:            Linux-5.15.167.4-microsoft-standard-WSL2-x86_64-with-glibc2.39
Python:              3.12.7 (main, Oct 18 2024, 08:38:59) [GCC 13.2.0]
LTS CPU:             False

----Optional dependencies----
adbc_driver_manager  <not installed>
altair               <not installed>
azure.identity       1.19.0
boto3                <not installed>
cloudpickle          <not installed>
connectorx           <not installed>
deltalake            0.17.4
fastexcel            <not installed>
fsspec               2024.12.0
gevent               <not installed>
google.auth          2.37.0
great_tables         <not installed>
matplotlib           <not installed>
nest_asyncio         1.6.0
numpy                2.2.1
openpyxl             <not installed>
pandas               2.2.3
pyarrow              18.1.0
pydantic             2.10.4
pyiceberg            <not installed>
sqlalchemy           2.0.36
torch                <not installed>
xlsx2csv             <not installed>
xlsxwriter           <not installed>