Change how the width of the gutter is calculated considering propotional fonts.

[matj1]

Check for existing issues

• [X] Completed

Describe the bug / provide steps to reproduce it

Pull request #4135 sets that there are always 4 ems of space reserved for the line number to prevent changing the width much. The idea is reasonable, but em is defined as the width of ‘m’. In proportional fonts, ‘m’ is usually much wider than numerals, so the gutter is too wide. That can be seen in this comment.

The same em is used to calculate padding, so the gutter also has too much padding.

Reproduction

1. Set the buffer font to something proportional (like Noto Sans): in settings "buffer_font_family": "Noto Sans"
2. See that the gutter is too wide.

Solution

I suggest that the width of the numeral ‘0’ would be used as the width of a digit in the gutter. The width of digits usually does not vary much, and, if so, usually just because ‘1’ is narrower. So there is practically no risk of not enough space in the gutter for 4 digits.

It seems that it would be solved by redefining the variables em_width and em_advance in the appropriate functions. When #4135 was merged, the appropriate functions were compute_layout and compute_auto_height_layout in crates/editor/src/element.rs. But the file has changed, so I don't know which are the appropriate functions now.

The variables would be redefined so:

let em_width = cx
    .text_system()
    .typographic_bounds(font_id, font_size, '0') // not 'm'
    .unwrap()
    .size
    .width;
let em_advance = cx
    .text_system()
    .advance(font_id, font_size, '0') // not 'm'
    .unwrap()
    .width;

The names no longer reflect their function, so they should be changed to something more suitable.

Environment

Zed: v0.164.2 (Zed) OS: Linux X11 arch unknown Memory: 15.6 GiB Architecture: x86_64 GPU: AMD Radeon RX 580 2048SP (RADV POLARIS10) || radv || Mesa 24.3.1-arch1.1

If applicable, add mockups / screenshots to help explain present your vision of the feature

[iamnbutler]

I agree, that seems like a resoable solution.

I added the good first issue tag, as the fix should be reasonably straightforward and would be a great way for new contributors to jump in and contribute.

Otherwise someone can feel free to jump in and fix this.

[TimmyMcPickles]

I'd like to try and check this issue out, if that would be alright

[flvrone]

@iamnbutler, hey, in that case - would you consider looking at the tiny changes I have made in my local PR? https://github.com/flvrone/zed/pull/1

[iamJ3]

is the issue open?