Bad structure

[MarkLevkovich]

ENTER --> ic([1,1,1,1,1,1,1,1,0,0,0,0,0,0,0,0,0,0,0,0,1,1,1,1,1,0,1,0,1,0,1,0])
RESULT --> ic| rev: [0,
1,
0,
1,
0,
1,
0,
1,
1,
1,
1,
1,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
1,
1,
1,
1,
1,
1,
1,
1]

IMHO its not good

[Jakeroid]

@MarkLevkovich Hey,

Thank you for submitting this.

Are you saying that the order was changed?

[MarkLevkovich]

yep, it was simple kata, but it's not about the function, but about the length, let's say with a length of 2 times less, everything is OK.

from icecream import ic
def data_reverse(data):
    bb = []
    for b in range(0, len(data), 8):
        bb.append(data[b : b + 8])

    left = 0
    right = len(bb) - 1

    while right >= left:
        bb[left], bb[right] = bb[right], bb[left]
        left += 1
        right -= 1

    return [el for b in bb for el in b]


ic(data_reverse([0, 0, 1, 1, 0, 1, 1, 0, 0, 0, 1, 0, 1, 0, 0, 1, 0, 0, 1, 1, 0, 1, 1, 0, 0, 0, 1, 0, 1, 0, 0, 1]))

[MarkLevkovich]

mb i can fix that

[Jakeroid]

The related PR is here: #230

[Jakeroid]

@MarkLevkovich Do you think #60 is a related issue?

Maybe we will close two issues by one PR.

[MarkLevkovich]

@Jakeroid Yes, #60 is definitely related! Both issues are about formatting — icecream doesn't respect width settings and incorrectly splits data across lines. One PR should help, since we need to fix the formatting logic in one place.

[Jakeroid]

@MarkLevkovich

I'm checking the PR and the results. I have the question. Why do we have exactly 13 and 35 right here?

     if is_flat_list and isinstance(obj, list) and 13 <= len(obj) <= 35:
         lines = formatted.splitlines()
         if len(lines) > 10:
             one_line = repr(obj)
             if len(one_line) <= 120:
                 return one_line

Would this be less wordy but kind of the same?

    if is_flat_list:
        lines = formatted.splitlines()
        if len(lines) > 1:  # If it's not on one line, check if repr fits
            one_line = repr(obj)
            if len(one_line) <= 120:
                return one_line

[MarkLevkovich]

hi, Those bounds are there to define the “medium‑sized flat lists” we want to special‑case, and they were chosen empirically:

• The lower bound 13 ensures we don’t touch small lists: for short flat lists pprint already keeps everything on one line, so there’s no benefit in re‑formatting them through repr.
• The upper bound 35 ensures we don’t touch very large lists: beyond that size, even if repr(obj) technically fits into 120 characters in some cases, a single extremely dense line becomes harder to read than the multi‑line layout. The test_flat_medium_list_prints_on_one_line uses a list of 32 elements, so 13 <= len(obj) <= 35 exactly covers that “medium” region we care about and avoid applying the fallback to clearly small or clearly large lists.

So the condition

is_flat_list and 13 <= len(obj) <= 35

Regarding the alternative:

if is_flat_list:
    lines = formatted.splitlines()
    if len(lines) > 1:  # If it's not on one line, check if repr fits
        one_line = repr(obj)
        if len(one_line) <= 120:
            return one_line

Semantics are similar, but it is a bit broader:

• It would apply the repr fallback to any flat list that pprint breaks across lines, regardless of length (including very small and very large ones).
• The current range check deliberately limits this behavior to the “medium” region where we actually observed the readability issue and where repr is still typically compact.

[Jakeroid]

Here is the release with your changes @MarkLevkovich! Big thanks ❤️

I'll close the issue. Let me know if something goes wrong and it should be re-opened.

https://github.com/gruns/icecream/releases/tag/v2.1.9

[salabim]

I am not sure whether I understand the issue and the fix correctly, but I think that pprint provides everything you need when you specify compact=True. Wouldn't that be easier and better?

[MarkLevkovich]

@salabim Thanks for the suggestion, but compact=True doesn't actually solve the issue here.

The problem is specific to medium-length flat lists (13–35 items), where pprint sometimes splits them across lines even though they fit in ≤120 characters. compact=True only affects nested structures — it doesn’t prevent a flat list from being wrapped.

We chose 13 and 35 empirically:

• Below 13, pprint already keeps them on one line.
• Above 35, even if repr() fits in 120 chars, it becomes too dense to read easily.

The fix in #230 intentionally targets only that middle range where readability suffers with pprint but repr() is still clean. Using compact=True wouldn’t give us that control and could worsen output for other cases.

The release with this fix is already out in v2.1.9, and it resolves the reported formatting issue without breaking existing behavior.

[salabim]

@MarkLevkovich
I think it would be all solved if you communicated the lineWrapWidth (minus something for the prefix++) and compact=True to pformat, it would work as expected. And then there's no beed for some random threshold values and a linewidth of 120 .
See also issue #60.