markdown: Pass emoji codes to yuin/goldmark-emoji

[jmooring]

Fixes #7332
Fixes #11587

[jmooring]

So, this works great, but right now you need to do two things in site config to use it:

enableEmoji = false
[markup.goldmark.extensions]
emoji = true # default is false

It would be better if enableEmoji were the primary on/off switch, with goldmark/emoji enabled by default. The problem I'm having is in the page parser and shortcode parser:

(

hugo/hugolib/page.go

Lines 796 to 801 in de4e466

	case it.Type == pageparser.TypeEmoji:
	if emoji := helpers.Emoji(it.ValStr(result.Input())); emoji != nil {
	rn.AddReplacement(emoji, it)
	} else {
	rn.AddBytes(it)
	}

)

What I want to do is:

if this_content_will_be_rendered_by_goldmark {
  rn.AddBytes(it) 
} else {
  if emoji := helpers.Emoji(it.ValStr(result.Input())); emoji != nil { 
    rn.AddReplacement(emoji, it) 
  } else { 
    rn.AddBytes(it) 
  } 
}

But I can't figure out how to determine this_content_will_be_rendered_by_goldmark.

[bep]

But I can't figure out how to determine this_content_will_be_rendered_by_goldmark.

So, the Emoji feature was added a long time ago. I'm tempted to make that into a Goldmark only thing. We need to reduce some of the complexity in Hugo, and, as the issues referenced above, what we have isn't really working so we might as well just cut it.

We should translate enableEmoji to the new Goldmark option.

[jmooring]

Goldmark only thing

I'm fine with that, as long as we understand that existing emoji in org, pandoc, rst, asciidoc, and html will no longer render (breaking change). I just tested the first four to verify that the renders don't already have this baked in.

I'll update the PR.

[jmooring]

enableEmoji in site config now enables/disables the yuin/goldmark-emoji extension. No other config setting required. For markdown users, there is one expected change that is covered in the test cases:

{{< sc >}}:smiley:{{< /sc >}}

The shortcode must now do this:

{{ .Inner | .Page.RenderString }}

Instead of this:

{{ .Inner }}

Shortcodes called with the {{% sc %}} notation continue to work as they did before.

---

And it is a bit faster. With a 500 page site with 1877 emoji on each page, yuin/goldmark-emoji reduces the build time by 20%. But that's pretty much noise, because site builds in about a second.

[bep]

A tip: I found that I repeated the above construct so often, so I added 2 convenience funcs named Test and TestRunning which you could rewrite the above to:

b := hugolib.Test(t, files)

[github-actions]

This pull request has been automatically locked since there has not been any recent activity after it was closed. Please open a new issue for related bugs.