Codecov Report

Merging #8834 (35d1e2d) into master (0780204) will decrease coverage by 0.03%.
The diff coverage is 100.00%.

❗ Current head 35d1e2d differs from pull request most recent head f8be693. Consider uploading reports for the commit f8be693 to get more accurate results

@@            Coverage Diff             @@
##           master    #8834      +/-   ##
==========================================
- Coverage   90.18%   90.15%   -0.04%     
==========================================
  Files         151      147       -4     
  Lines      170880   170175     -705     
==========================================
- Hits       154102   153413     -689     
+ Misses      16778    16762      -16     

Flags with carried forward coverage won't be shown. Click here to find out more.

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 0780204...f8be693. Read the comment docs.

Isn't that something, that you can already do using something like this:

call matchadd('Conceal', ' \@<= ', 10, -1, {'conceal':'—'})
call matchadd('Conceal', '\S\zs  \@=', 10, -1, {'conceal':'—'})

This requires setting conceallevel=2 and concealcursor=nvic (maybe?) to achieve the effect similar to listchars, which is not very convenient in some cases (e.g. when editing LaTeX files).

Well yes, and in your case it needs an additional :set list. I am just wondering if the maintenance burden for adding that feature is justified, considering, that we already have a feature that allows this as well.

Everything in listchars can be toggled on and off, so it is unlikely that this will conflict with other use cases, unlike conceallevel and concealcursor which affect everything that use the conceal feature.

This requires setting conceallevel=2 and concealcursor=nvic (maybe?) to achieve the effect similar to listchars, which is not very convenient in some cases (e.g. when editing LaTeX files).

Maybe "convenient" is not the right word here. I wanted to mean that it may conflict with other use cases of conceal, especially those that don't work well with concealcursor=nvic.

Instead of a single character, how about making this a string? E.g. setting it to "---+" would show a longer sequence of spaces as "---+---+---+". Would not be used for a single space.

Instead of a single character, how about making this a string? E.g. setting it to "---+" would show a longer sequence of spaces as "---+---+---+". Would not be used for a single space.

Should the string be char_u * or int * in structs.h?

I'm using int *.

I think this is done now.

Several continue; statements in set_chars_option always lead to return e_invarg; (as the value of pointer p is not changed). Maybe they can be changed to return e_invarg; to be more straightforward?

This is starting to look good. A couple of notes.

The name "contig" is not at all obvious. How about "multspace"? or "space2"?

The help should give an example, e.g. "---.---+".

It seems "multiple" is more often abbreviated to "multi" than as "mult" in Vim. "multispace" indeed seems a more straightforward name.