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

The script below shows how polars (1.18.0) has expected behaviour for datetime datetime objects and fails to parse the pendulum counterpart.

The error given by the script is:

polars.exceptions.ComputeError: unable to parse time zone: 'FixedTimezone(0, name="+00:00")'. Please check the Time Zone Database for a list of available time zones

import datetime as dt
import pendulum as pm
import polars as pl

# Set arbitrary string timestamp
dt_string = "2025-01-03 10:06:00.232323+00:00"

# Parse string as pendulum.DateTime
pm_dt = pm.from_format(dt_string, fmt="YYYY-MM-DD HH:mm:ss.SSSSSSZ")

# Parse string as datetime.datetime
dt_dt = dt.datetime.strptime(dt_string, "%Y-%m-%d %H:%M:%S.%f%z")

# Empty DataFrame #
df = pl.DataFrame()

# Add new column
# datetime_col works
# pendulum_col breaks
df = df.with_columns([
        pl.lit(pm_dt).alias("pendulum_col"),
        # pl.lit(dt_dt).alias("datetime_col"),
    ])

if __name__ == "__main__":
    print(df)

Log output

No response

Issue description

Polars does not parse pendulum.DateTime objects when adding to DataFrame with the with_columns method.

Issue appears to be with how timezone is encoded, although polars works fine with a timezone-aware datetime.datetime object.

Expected behavior

Polars to parse the pendulum.DateTime column and add as column to DataFrame with type polars.Datetime.

Installed versions

--------Version info---------
Polars:              1.18.0
Index type:          UInt32
Platform:            macOS-15.2-arm64-arm-64bit
Python:              3.11.9 (main, Aug 19 2024, 09:48:03) [Clang 15.0.0 (clang-1500.3.9.4)]
LTS CPU:             False