markup/asciidoc: Add support for .TableOfContents

Fill the .TableOfContents template variable when writing Asciidoc content.
This is done by letting Asciidoc render its TOC as HTML, then extract this
HTML rendered TOC, parse it into a tableofcontents.Root and finally remove
it from the HTML content.
This aims to stay in the logic that the Asciidoc parsing is entirely done
by the external helper.

See #1687

docs/content/en/content-management/toc.md

@@ -92,6 +92,33 @@ The following is a [partial template][partials] that adds slightly more logic fo
 With the preceding example, even pages with > 400 words *and* `toc` not set to `false` will not render a table of contents if there are no headings in the page for the `{{.TableOfContents}}` variable to pull from.
 {{% /note %}}
 
+## Usage with asciidoc
+
+Hugo supports table of contents with Asciidoc content format.
+
+In the header of your content file, specify the Asciidoc TOC directives, by using the macro style:
+
+```asciidoc
+// <!-- Your front matter up here -->
+:toc: macro
+// Set toclevels to be at least your hugo [markup.tableOfContents.endLevel] config key
+:toclevels: 4
+toc::[]
+
+== Introduction
+
+One morning, when Gregor Samsa woke from troubled dreams, he found himself transformed in his bed into a horrible vermin.
+
+== My Heading
+
+He lay on his armour-like back, and if he lifted his head a little he could see his brown belly, slightly domed and divided by arches into stiff sections. The bedding was hardly able to cover it and seemed ready to slide off any moment.
+
+=== My Subheading
+
+A collection of textile samples lay spread out on the table - Samsa was a travelling salesman - and above it there hung a picture that he had recently cut out of an illustrated magazine and housed in a nice, gilded frame. It showed a lady fitted out with a fur hat and fur boa who sat upright, raising a heavy fur muff that covered the whole of her lower arm towards the viewer. Gregor then turned to look out the window at the dull weather. Drops
+```
+Hugo will take this Asciddoc and create a table of contents store it in the page variable `.TableOfContents`, in the same as described for Markdown.
+
 [conditionals]: /templates/introduction/#conditionals
 [front matter]: /content-management/front-matter/
 [pagevars]: /variables/page/

markup/asciidocext/convert.go

@@ -17,6 +17,7 @@
 package asciidocext
 
 import (
+	"bytes"
 	"os/exec"
 	"path/filepath"
 
@@ -24,6 +25,8 @@ import (
 	"github.com/gohugoio/hugo/markup/asciidocext/asciidocext_config"
 	"github.com/gohugoio/hugo/markup/converter"
 	"github.com/gohugoio/hugo/markup/internal"
+	"github.com/gohugoio/hugo/markup/tableofcontents"
+	"golang.org/x/net/html"
 )
 
 /* ToDo: RelPermalink patch for svg posts not working*/
@@ -45,16 +48,32 @@ func (p provider) New(cfg converter.ProviderConfig) (converter.Provider, error)
 	}), nil
 }
 
+type asciidocResult struct {
+	converter.Result
+	toc tableofcontents.Root
+}
+
+func (r asciidocResult) TableOfContents() tableofcontents.Root {
+	return r.toc
+}
+
 type asciidocConverter struct {
 	ctx converter.DocumentContext
 	cfg converter.ProviderConfig
 }
 
 func (a *asciidocConverter) Convert(ctx converter.RenderContext) (converter.Result, error) {
-	return converter.Bytes(a.getAsciidocContent(ctx.Src, a.ctx)), nil
+	content, toc, err := extractTOC(a.getAsciidocContent(ctx.Src, a.ctx))
+	if err != nil {
+		return nil, err
+	}
+	return asciidocResult{
+		Result: converter.Bytes(content),
+		toc:    toc,
+	}, nil
 }
 
-func (c *asciidocConverter) Supports(feature identity.Identity) bool {
+func (a *asciidocConverter) Supports(_ identity.Identity) bool {
 	return false
 }
 
@@ -182,6 +201,112 @@ func getAsciidoctorExecPath() string {
 	return path
 }
 
+// extractTOC extracts the toc from the given src html.
+// It returns the html without the TOC, and the TOC data
+func extractTOC(src []byte) ([]byte, tableofcontents.Root, error) {
+	var buf bytes.Buffer
+	buf.Write(src)
+	node, err := html.Parse(&buf)
+	if err != nil {
+		return nil, tableofcontents.Root{}, err
+	}
+	var (
+		f       func(*html.Node) bool
+		toc     tableofcontents.Root
+		toVisit []*html.Node
+	)
+	f = func(n *html.Node) bool {
+		if n.Type == html.ElementNode && n.Data == "div" {
+			for _, a := range n.Attr {
+				if a.Key == "id" && a.Val == "toc" {
+					toc, err = parseTOC(n)
+					if err != nil {
+						return false
+					}
+					n.Parent.RemoveChild(n)
+					return true
+				}
+			}
+		}
+		if n.FirstChild != nil {
+			toVisit = append(toVisit, n.FirstChild)
+		}
+		if n.NextSibling != nil {
+			if ok := f(n.NextSibling); ok {
+				return true
+			}
+		}
+		for len(toVisit) > 0 {
+			nv := toVisit[0]
+			toVisit = toVisit[1:]
+			if ok := f(nv); ok {
+				return true
+			}
+		}
+		return false
+	}
+	f(node)
+	if err != nil {
+		return nil, tableofcontents.Root{}, err
+	}
+	buf.Reset()
+	err = html.Render(&buf, node)
+	if err != nil {
+		return nil, tableofcontents.Root{}, err
+	}
+	// ltrim <html><head></head><body> and rtrim </body></html> which are added by html.Render
+	res := buf.Bytes()[25:]
+	res = res[:len(res)-14]
+	return res, toc, nil
+}
+
+// parseTOC returns a TOC root from the given toc Node
+func parseTOC(doc *html.Node) (tableofcontents.Root, error) {
+	var (
+		toc tableofcontents.Root
+		f   func(*html.Node, int, int)
+	)
+	f = func(n *html.Node, parent, level int) {
+		if n.Type == html.ElementNode {
+			switch n.Data {
+			case "ul":
+				if level == 0 {
+					parent += 1
+				}
+				level += 1
+				f(n.FirstChild, parent, level)
+			case "li":
+				for c := n.FirstChild; c != nil; c = c.NextSibling {
+					if c.Type != html.ElementNode || c.Data != "a" {
+						continue
+					}
+					var href string
+					for _, a := range c.Attr {
+						if a.Key == "href" {
+							href = a.Val[1:]
+							break
+						}
+					}
+					for d := c.FirstChild; d != nil; d = d.NextSibling {
+						if d.Type == html.TextNode {
+							toc.AddAt(tableofcontents.Header{
+								Text: d.Data,
+								ID:   href,
+							}, parent, level)
+						}
+					}
+				}
+				f(n.FirstChild, parent, level)
+			}
+		}
+		if n.NextSibling != nil {
+			f(n.NextSibling, parent, level)
+		}
+	}
+	f(doc.FirstChild, 0, 0)
+	return toc, nil
+}
+
 // Supports returns whether Asciidoctor is installed on this computer.
 func Supports() bool {
 	return getAsciidoctorExecPath() != ""

markup/asciidocext/convert_test.go

@@ -270,3 +270,38 @@ func TestConvert(t *testing.T) {
 	c.Assert(err, qt.IsNil)
 	c.Assert(string(b.Bytes()), qt.Equals, "<div class=\"paragraph\">\n<p>testContent</p>\n</div>\n")
 }
+
+func TestTableOfContents(t *testing.T) {
+	if !Supports() {
+		t.Skip("asciidoc/asciidoctor not installed")
+	}
+	c := qt.New(t)
+	p, err := Provider.New(converter.ProviderConfig{Logger: loggers.NewErrorLogger()})
+	c.Assert(err, qt.IsNil)
+	conv, err := p.New(converter.DocumentContext{})
+	c.Assert(err, qt.IsNil)
+	b, err := conv.Convert(converter.RenderContext{Src: []byte(`:toc: macro
+:toclevels: 4
+toc::[]
+
+=== Introduction
+
+== Section 1
+
+=== Section 1.1
+
+==== Section 1.1.1
+
+=== Section 1.2
+
+testContent
+
+== Section 2
+`)})
+	c.Assert(err, qt.IsNil)
+	toc, ok := b.(converter.TableOfContentsProvider)
+	c.Assert(ok, qt.Equals, true)
+	root := toc.TableOfContents()
+	c.Assert(root.ToHTML(2, 4, false), qt.Equals, "<nav id=\"TableOfContents\">\n  <ul>\n    <li><a href=\"#_introduction\">Introduction</a></li>\n    <li><a href=\"#_section_1\">Section 1</a>\n      <ul>\n        <li><a href=\"#_section_1_1\">Section 1.1</a>\n          <ul>\n            <li><a href=\"#_section_1_1_1\">Section 1.1.1</a></li>\n          </ul>\n        </li>\n        <li><a href=\"#_section_1_2\">Section 1.2</a></li>\n      </ul>\n    </li>\n    <li><a href=\"#_section_2\">Section 2</a></li>\n  </ul>\n</nav>")
+	c.Assert(root.ToHTML(2, 3, false), qt.Equals, "<nav id=\"TableOfContents\">\n  <ul>\n    <li><a href=\"#_introduction\">Introduction</a></li>\n    <li><a href=\"#_section_1\">Section 1</a>\n      <ul>\n        <li><a href=\"#_section_1_1\">Section 1.1</a></li>\n        <li><a href=\"#_section_1_2\">Section 1.2</a></li>\n      </ul>\n    </li>\n    <li><a href=\"#_section_2\">Section 2</a></li>\n  </ul>\n</nav>")
+}