markup/goldmark: Add an optional Blackfriday auto ID strategy

Fixes #6707

2 files changed, 65 insertions, 1 deletions

diff --git a/markup/blackfriday/convert.go b/markup/blackfriday/convert.go
index bbbc2b377..d844c5554 100644
--- a/markup/blackfriday/convert.go
+++ b/markup/blackfriday/convert.go
@@ -15,6 +15,8 @@
 package blackfriday
 
 import (
+	"unicode"
+
 	"github.com/gohugoio/hugo/identity"
 	"github.com/gohugoio/hugo/markup/blackfriday/blackfriday_config"
 	"github.com/gohugoio/hugo/markup/converter"
@@ -61,7 +63,27 @@ type blackfridayConverter struct {
 }
 
 func (c *blackfridayConverter) SanitizeAnchorName(s string) string {
-	return blackfriday.SanitizedAnchorName(s)
+	return SanitizedAnchorName(s)
+}
+
+// SanitizedAnchorName is how Blackfriday sanitizes anchor names.
+// Implementation borrowed from https://github.com/russross/blackfriday/blob/a477dd1646916742841ed20379f941cfa6c5bb6f/block.go#L1464
+func SanitizedAnchorName(text string) string {
+	var anchorName []rune
+	futureDash := false
+	for _, r := range text {
+		switch {
+		case unicode.IsLetter(r) || unicode.IsNumber(r):
+			if futureDash && len(anchorName) > 0 {
+				anchorName = append(anchorName, '-')
+			}
+			futureDash = false
+			anchorName = append(anchorName, unicode.ToLower(r))
+		default:
+			futureDash = true
+		}
+	}
+	return string(anchorName)
 }
 
 func (c *blackfridayConverter) AnchorSuffix() string {
diff --git a/markup/blackfriday/convert_test.go b/markup/blackfriday/convert_test.go
index b4d66dec6..d2d8d927e 100644
--- a/markup/blackfriday/convert_test.go
+++ b/markup/blackfriday/convert_test.go
@@ -179,3 +179,45 @@ This is a footnote.[^1] And then some.
 	c.Assert(s, qt.Contains, "This is a footnote.<sup class=\"footnote-ref\" id=\"fnref:testid:1\"><a href=\"#fn:testid:1\">1</a></sup>")
 	c.Assert(s, qt.Contains, "<a class=\"footnote-return\" href=\"#fnref:testid:1\"><sup>[return]</sup></a>")
 }
+
+// Tests borrowed from https://github.com/russross/blackfriday/blob/a925a152c144ea7de0f451eaf2f7db9e52fa005a/block_test.go#L1817
+func TestSanitizedAnchorName(t *testing.T) {
+	tests := []struct {
+		text string
+		want string
+	}{
+		{
+			text: "This is a header",
+			want: "this-is-a-header",
+		},
+		{
+			text: "This is also          a header",
+			want: "this-is-also-a-header",
+		},
+		{
+			text: "main.go",
+			want: "main-go",
+		},
+		{
+			text: "Article 123",
+			want: "article-123",
+		},
+		{
+			text: "<- Let's try this, shall we?",
+			want: "let-s-try-this-shall-we",
+		},
+		{
+			text: "        ",
+			want: "",
+		},
+		{
+			text: "Hello, 世界",
+			want: "hello-世界",
+		},
+	}
+	for _, test := range tests {
+		if got := SanitizedAnchorName(test.text); got != test.want {
+			t.Errorf("SanitizedAnchorName(%q):\ngot %q\nwant %q", test.text, got, test.want)
+		}
+	}
+}