feat(python): Add strict parameter to DataFrame constructor to allow non-strict construction

[stinodego]

Ref #14427

The DataFrame constructor had no strict parameter, while the Series constructor does. Non-strict construction is useful when the input data is not guaranteed to be clean.

import polars as pl

data = {"a": [-1, 0, 1]}

# Data does not match desired data type, cannot construct dataframe
df = pl.DataFrame(data, schema={"a": pl.UInt8})
# OverflowError

# Pass `strict=False` to handle this cleanly
df = pl.DataFrame(data, schema={"a": pl.UInt8}, strict=False)
print(df)

shape: (3, 1)
┌──────┐
│ a    │
│ ---  │
│ u8   │
╞══════╡
│ null │
│ 0    │
│ 1    │
└──────┘

In this example, the result is similar to a non-strict cast of the data to the desired data type.
This will also allow handling mixed ints/floats cleanly in the future.

The default behavior is unchanged, as previously DataFrame construction was always strict, corresponding to strict=True.

[codecov]

Codecov Report

Attention: Patch coverage is 91.30435% with 4 lines in your changes are missing coverage. Please review.

Project coverage is 81.00%. Comparing base (1d7b49e) to head (e3437d4).
Report is 4 commits behind head on main.

❗ Current head e3437d4 differs from pull request most recent head 770f9ff. Consider uploading reports for the commit 770f9ff to get more accurate results

Files	Patch %	Lines
py-polars/polars/_utils/construction/dataframe.py	91.30%	2 Missing and 2 partials ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main   #15034      +/-   ##
==========================================
- Coverage   81.00%   81.00%   -0.01%     
==========================================
  Files        1337     1338       +1     
  Lines      173326   173603     +277     
  Branches     2460     2461       +1     
==========================================
+ Hits       140403   140624     +221     
- Misses      32454    32509      +55     
- Partials      469      470       +1     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.