Normalize consecutive spaces?

[bpj]

Didn't there use to be an option/extension to normalize consecutive spaces and the like in output?

My use case is that I have a filter which removes certain spans, and I end up with double spaces in the output where there are spaces before and after those spans.

I'm just wondering as I seem to remember this. It is no big deal as the obvious workaround is to put the leading space inside the span, which I probably should do anyway if there is a terminal punctuation after the span.

[jgm]

When you remove the spans, are you replacing them with an empty span or with nothing {}?
In the latter case I would not expect you'd get doubled spaces in the output.

% pandoc -f native -t html
[Str "hi",Space,Space,Str "hi"]
hi hi

[jgm]

In any case, it should be possible to write a Lua filter that collapses adjacent Space elements.
use Inlines and iterate over the list.

[bpj]

I will probably do that in a second pass. I have anyway found that it is better to put spaces which go before the content of conditional spans inside the span, or I may end up with spaces before punctuation.

[jgm]

It is nowrap that is doing this. The implementation could be improved.

[bpj]

Do you want me to close this and/or open an issue?

[jgm]

It's a pretty low priority thing, but if you want to open an issue, you could open it on jgm/doclayout.
The issue would be that nowrap simply converts breaking spaces (which normally collapse) into literal spaces, so that two adjacent ones will produce two spaces.