fix: Ensure List element truncation ellipses respect ASCII* table formats

[alexander-beedie]

Closes #19821.

Follows-up #19404, which missed the List dtype ellipsis truncation.

Example

import polars as pl

df = pl.DataFrame({
    "a": [1, 2],
    "b": [4, 5],
    "c": [[list(range(1, 26))], [list(range(1, 76))]],
})
with pl.Config(ascii_tables=True):
    print(df)   

Before:
(utf8 ellipsis char used in ascii table)

# shape: (2, 3)
# +-----+-----+-----------------+
# | a   | b   | c               |
# | --- | --- | ---             |
# | i64 | i64 | list[list[i64]] |
# +=============================+
# | 1   | 4   | [[1, 2, … 25]]  |
# | 2   | 5   | [[1, 2, … 75]]  |
# +-----+-----+-----------------+

After:
(ascii ellipsis chars)

# shape: (2, 3)
# +-----+-----+------------------+
# | a   | b   | c                |
# | --- | --- | ---              |
# | i64 | i64 | list[list[i64]]  |
# +==============================+
# | 1   | 4   | [[1, 2, ... 25]] |
# | 2   | 5   | [[1, 2, ... 75]] |
# +-----+-----+------------------+

[codecov]

Codecov Report

Attention: Patch coverage is 80.00000% with 3 lines in your changes missing coverage. Please review.

Project coverage is 79.33%. Comparing base (6ccb187) to head (05b232e).
Report is 2 commits behind head on main.

Files with missing lines	Patch %	Lines
crates/polars-core/src/fmt.rs	80.00%	3 Missing ⚠️

Additional details and impacted files

@@           Coverage Diff           @@
##             main   #19835   +/-   ##
=======================================
  Coverage   79.33%   79.33%           
=======================================
  Files        1548     1548           
  Lines      214245   214237    -8     
  Branches     2460     2460           
=======================================
+ Hits       169968   169972    +4     
+ Misses      43719    43707   -12     
  Partials      558      558           

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

[kevinli1993]

Thanks @alexander-beedie - this is slightly out of scope for this PR, but I wonder if we should adopt the Oxford comma?

e.g. from

[[1, 2, ... 25]]

to

[[1, 2, ..., 25]]

[alexander-beedie]

I wonder if we should adopt the Oxford comma?

We should not - it gives up valuable horizontal space for no obvious gain ;)

The only reference I found (after an admittedly very casual search!) talking about this situation seems to agree:

The ellipsis itself is typically used to indicate omitted content. Since it replaces part of the list, treating it as part of the structure (without an additional comma) is common.

[kevinli1993]

I see, that makes sense. Though I will say that mathematical writing follows the opposite convention of including the Oxford comma in this context.

[alexander-beedie]

I see, that makes sense. Though I will say that mathematical writing follows the opposite convention of including the Oxford comma in this context.

Yup. Different domain; here we prioritise visualisation and effective use of horizontal space. The priority in a mathematical domain will understandably err towards formal precision. Each is appropriate in their respective context 👍