Impossible to create a new column containing a struct with one column containing a datetime literal

[ismailhammounou]

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

I tested the code in 1.8.0 and above and it did not work but it seems to work < 1.8.0 (tested on 1.7.1)

import polars as pl
from datetime import datetime
import pytz
 
df = pl.DataFrame({"col1": ["a", "b"]})
metadata = pl.struct(
        pl.lit(datetime.now(pytz.UTC)).alias('creation_ts'), pl.lit("a").alias("col_2"),
        schema={"creation_ts": pl.Datetime(time_unit='us', time_zone="UTC"), "col_2": pl.Utf8()}
    )
df = df.with_columns(metadata.alias("__metadata__"))
print(df)

Log output

thread '<unnamed>' panicked at crates/polars-core/src/scalar/mod.rs:46:92:
called `Result::unwrap()` on an `Err` value: SchemaMismatch(ErrString("unexpected value while building Series of type Datetime(Microseconds, Some(\"UTC\")); found value of type Datetime(Microseconds, Some(\"UTC\")): 2024-10-03 12:18:33.247992 UTC"))
---------------------------------------------------------------------------
PanicException                            Traceback (most recent call last)
/tmp/ipykernel_11476/316418030.py in ?()
      7         pl.lit(datetime.now(pytz.UTC)).alias('creation_ts'), pl.lit("a").alias("col_2"),
      8         schema={"creation_ts": pl.Datetime(time_unit='us', time_zone="UTC"), "col_2": pl.Utf8()}
      9     )
     10 df = df.with_columns(metadata.alias("__metadata__"))
---> 11 print(df)

/opt/conda/lib/python3.11/site-packages/polars/dataframe/frame.py in ?(self)
   1170     def __str__(self) -> str:
-> 1171         return self._df.as_str()

PanicException: called `Result::unwrap()` on an `Err` value: SchemaMismatch(ErrString("unexpected value while building Series of type Datetime(Microseconds, Some(\"UTC\")); found value of type Datetime(Microseconds, Some(\"UTC\")): 2024-10-03 12:18:33.247992 UTC"))

Issue description

When I use datetime object in literal inside a struct, I am getting an error but when I use an existing column from the df inside the struct it works along with the datetime literal.

This code works:

import polars as pl
from datetime import datetime
import pytz
 
df = pl.DataFrame({"col1": ["a", "b"], "col_2": ["a", "a"]})
metadata = pl.struct(
        pl.lit(datetime.now(pytz.UTC)).alias('creation_ts'), pl.col("col_2"),
        schema={"creation_ts": pl.Datetime(time_unit='us', time_zone="UTC"), "col_2": pl.Utf8()}
    )
df = df.with_columns(metadata.alias("__metadata__"))
print(df)

but this code does not work as expected:

import polars as pl
from datetime import datetime
import pytz
 
df = pl.DataFrame({"col1": ["a", "b"]})
metadata = pl.struct(
        pl.lit(datetime.now(pytz.UTC)).alias('creation_ts'), pl.lit("a").alias("col_2"),
        schema={"creation_ts": pl.Datetime(time_unit='us', time_zone="UTC"), "col_2": pl.Utf8()}
    )
df = df.with_columns(metadata.alias("__metadata__"))
print(df)

Expected behavior

shape: (2, 3)
┌──────┬─────---─-------------------┐
│ col1 ┆__metadata__ │
│ --- ┆--- │
│ str ┆struct[2] │
╞══════╪══════════════════╡
│ a ┆{2024-10-03 12:23:55.949877 UT… │
│ b ┆{2024-10-03 12:23:55.949877 UT… │
└───┴────-------──────────────┘

Installed versions

--------Version info---------
Polars:              1.8.0
Index type:          UInt32
Platform:            Linux-5.15.0-1071-azure-x86_64-with-glibc2.35
Python:              3.11.10 | packaged by conda-forge | (main, Sep 22 2024, 14:10:38) [GCC 13.3.0]

----Optional dependencies----
adbc_driver_manager  <not installed>
altair               5.4.1
cloudpickle          3.0.0
connectorx           <not installed>
deltalake            0.20.1
fastexcel            <not installed>
fsspec               2022.11.0
gevent               <not installed>
great_tables         <not installed>
matplotlib           3.9.2
nest_asyncio         1.6.0
numpy                2.0.2
openpyxl             3.1.5
pandas               2.2.3
pyarrow              17.0.0
pydantic             2.9.2
pyiceberg            <not installed>
sqlalchemy           2.0.35
torch                <not installed>
xlsx2csv             <not installed>
xlsxwriter           <not installed>

[cmdlineluser]

Can reproduce on 1.9.0 also.

import pytz
import polars as pl
from datetime import datetime

pl.DataFrame({"a": [1]}).with_columns(pl.struct(pl.lit(datetime.now(pytz.UTC))))
# shape: (1, 2)
# ┌─────┬─────────────────────────────────┐
# │ a   ┆ literal                         │
# │ --- ┆ ---                             │
# │ i64 ┆ struct[1]                       │
# ╞═════╪═════════════════════════════════╡
# │ 1   ┆ {2024-10-03 12:45:25.810150 UT… │
# └─────┴─────────────────────────────────┘

pl.DataFrame({"a": [1, 2]}).with_columns(pl.struct(pl.lit(datetime.now(pytz.UTC))))

# thread '<unnamed>' panicked at crates/polars-core/src/scalar/mod.rs:46:92:
# PanicException: called `Result::unwrap()` on an `Err` value: 
# SchemaMismatch(ErrString("unexpected value while building Series of type 
# Datetime(Microseconds, Some(\"UTC\")); found value of type 
# Datetime(Microseconds, Some(\"UTC\")): 2024-10-03 12:45:27.592791 UTC"))

It works in 1.7.2