glimpse() fails for lengthy column names

[CodeCox]

Polars version checks

• I have checked that this issue has not already been reported.

• I have confirmed this bug exists on the latest version of Polars.

Issue description

glimpse() fails for lengthy column names

The error msg ValueError: Sign not allowed in string format specifier suggests that a number in the justifiy format string may be negative?

This line of code in frame.py might be the cause of the negative number?

 # limit the amount of data printed such that total width is fixed
        max_col_values = 100 - max_col_name - max_col_dtype

Reproducible example

colc = 'a' * 96
df = pl.DataFrame({colc: [11,22,33]})
df.glimpse()

---------------------------------------------------------------------------
ValueError                                Traceback (most recent call last)
Cell In[97], line 3
      1 colc = 'a' * 96
      2 df = pl.DataFrame({colc: [11,22,33]})
----> 3 df.glimpse()

File ~\mambaforge\envs\bears\Lib\site-packages\polars\dataframe\frame.py:3577, in DataFrame.glimpse(self, return_as_string)
   3574 # print individual columns: one row per column
   3575 for col_name, dtype_str, val_str in data:
   3576     output.write(
-> 3577         f"$ {col_name:<{max_col_name}}"
   3578         f" {dtype_str:>{max_col_dtype}}"
   3579         f" {val_str:<{min(len(val_str), max_col_values)}}\n"
   3580     )
   3582 s = output.getvalue()
   3584 if not return_as_string:

ValueError: Sign not allowed in string format specifier

Expected behavior

same as this but with len() > 95, and of course use of the "..." in the column name as a 'truncation' to cater for width limit would be preferred to a failure!

colc = 'a' * 95
df = pl.DataFrame({colc: [11,22,33]})
df.glimpse()

Rows: 3
Columns: 1
$ aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa <i64> 11, 22, 33

Installed versions

Polars: 0.17.3

[zundertj]

Thank you for reporting, I have fixed this in #8281.