Add support for Table of Contents

README.org

@@ -4,12 +4,14 @@ Take a look at [[https://niklasfasching.github.io/go-org/][github pages]] for so
 * next
 - more keywords: https://orgmode.org/manual/In_002dbuffer-settings.html
 - table of contents
+  - see hugo ExtractTOC
+  - loop the headlines and print an hX for each headline, ul for children
 - rethink frontmatter
-  - keyword customizable mapping. e.g. #+ARRAY_KWS: list kws that should become []string
+  - use toml/yaml/json frontmatter from hugo (see [[https://github.com/gohugoio/hugo/issues/5436][hugo #5436]]):
-  - json/yaml values.
+    - complex values are a requirement of hugo, not Org mode
-  - +lisp syntax from [[https://github.com/kaushalmodi/ox-hugo/blob/master/ox-hugo.el#L2791][ox-hugo]]+
+    - by giving up on the ability to mix front matter in org keyword (=#+=) and other formats we save a lot of complexity (that has little benefit)
-    - too much complexity and not native to Org mode, i.e. does not even increase compatibility
+    - Maybe allow for []string via KEYWORD[]: Tag Foo Bar - can be done in hugo
-  - use toml/yaml/json frontmatter from hugo (see [[https://github.com/gohugoio/hugo/issues/5436][hugo #5436]])
+    -> the other problem is that org mode normally allows markup in the title - but we would have to render - and to what format
 ** headlines
 - auto-generate unique ids: see [[https://github.com/kaushalmodi/ox-hugo/blob/8472cf2d8667754c9da3728255634e8001a1da6d/ox-hugo.el#L1785-L1850][ox-hugo]]
    - what about name conflicts?
@@ -41,3 +43,4 @@ Nonetheless, the html output can be compared by taking a look in the developer c
   - https://code.orgmode.org/bzg/org-mode/src/master/lisp/org-element.el
   - mostly those & ox-html.el, but yeah, all of [[https://code.orgmode.org/bzg/org-mode/src/master/lisp/]]
 - existing Org mode implementations: [[https://github.com/emacsmirror/org][org]], [[https://github.com/bdewey/org-ruby/blob/master/spec/html_examples][org-ruby]], [[https://github.com/chaseadamsio/goorgeous/][goorgeous]], [[https://github.com/jgm/pandoc/][pandoc]]
+

org/document.go

@@ -15,6 +15,7 @@ type Document struct {
 	tokens              []token
 	Nodes               []Node
 	Footnotes           Footnotes
+	Outline             Outline
 	StatusKeywords      []string
 	MaxEmphasisNewLines int
 	AutoLink            bool
@@ -71,6 +72,7 @@ func FrontMatterHandler(fm FrontMatter, k, v string) error {
 }
 
 func NewDocument() *Document {
+	outlineSection := &Section{}
 	return &Document{
 		Footnotes: Footnotes{
 			Title:       "Footnotes",
@@ -78,6 +80,7 @@ func NewDocument() *Document {
 		},
 		AutoLink:            true,
 		MaxEmphasisNewLines: 1,
+		Outline:             Outline{outlineSection, outlineSection, 0},
 		BufferSettings:      map[string]string{},
 		DefaultSettings: map[string]string{
 			"TODO":         "TODO | DONE",
@@ -232,6 +235,12 @@ func (d *Document) addFootnote(name string, definition *FootnoteDefinition) {
 	d.Footnotes.addOrder = append(d.Footnotes.addOrder, name)
 }
 
+func (d *Document) addHeadline(headline *Headline) int {
+	d.Outline.last.add(&Section{Headline: headline})
+	d.Outline.count++
+	return d.Outline.count
+}
+
 func tokenize(line string) token {
 	for _, lexFn := range lexFns {
 		if token, ok := lexFn(line); ok {

org/headline.go

@@ -1,12 +1,26 @@
 package org
 
 import (
+	"fmt"
 	"regexp"
 	"strings"
 	"unicode"
 )
 
+type Outline struct {
+	*Section
+	last  *Section
+	count int
+}
+
+type Section struct {
+	Headline *Headline
+	Parent   *Section
+	Children []*Section
+}
+
 type Headline struct {
+	Index      int
 	Lvl        int
 	Status     string
 	Priority   string
@@ -29,6 +43,9 @@ func lexHeadline(line string) (token, bool) {
 func (d *Document) parseHeadline(i int, parentStop stopFn) (int, Node) {
 	t, headline := d.tokens[i], Headline{}
 	headline.Lvl = len(t.matches[1])
+
+	headline.Index = d.addHeadline(&headline)
+
 	text := t.content
 	todoKeywords := strings.FieldsFunc(d.Get("TODO"), func(r rune) bool { return unicode.IsSpace(r) || r == '|' })
 	for _, k := range todoKeywords {
@@ -64,3 +81,19 @@ func (d *Document) parseHeadline(i int, parentStop stopFn) (int, Node) {
 	headline.Children = nodes
 	return consumed + 1, headline
 }
+
+func (h Headline) ID() string {
+	if customID, ok := h.Properties.Get("CUSTOM_ID"); ok {
+		return customID
+	}
+	return fmt.Sprintf("headline-%d", h.Index)
+}
+
+func (parent *Section) add(current *Section) {
+	if parent.Headline == nil || parent.Headline.Lvl < current.Headline.Lvl {
+		parent.Children = append(parent.Children, current)
+		current.Parent = parent
+	} else {
+		parent.Parent.add(current)
+	}
+}

org/html.go

@@ -4,6 +4,7 @@ import (
 	"fmt"
 	"html"
 	"log"
+	"regexp"
 	"strings"
 	"unicode"
 
@@ -42,6 +43,8 @@ var listItemStatuses = map[string]string{
 	"X": "checked",
 }
 
+var cleanHeadlineTitleForHTMLAnchorRegexp = regexp.MustCompile(`</?a[^>]*>`) // nested a tags are not valid HTML
+
 func NewHTMLWriter() *HTMLWriter {
 	return &HTMLWriter{
 		htmlEscape: true,
@@ -66,6 +69,7 @@ func (w *HTMLWriter) nodesAsString(nodes ...Node) string {
 func (w *HTMLWriter) before(d *Document) {
 	w.excludeTags = strings.Fields(d.Get("EXCLUDE_TAGS"))
 	w.log = d.Log
+	w.writeOutline(d)
 }
 
 func (w *HTMLWriter) after(d *Document) {
@@ -201,6 +205,31 @@ func (w *HTMLWriter) writeFootnotes(d *Document) {
 	w.WriteString("</div>\n</div>\n")
 }
 
+func (w *HTMLWriter) writeOutline(d *Document) {
+	if len(d.Outline.Children) != 0 {
+		w.WriteString("<nav>\n<ul>\n")
+		for _, section := range d.Outline.Children {
+			w.writeSection(section)
+		}
+		w.WriteString("</ul>\n</nav>\n")
+	}
+}
+
+func (w *HTMLWriter) writeSection(section *Section) {
+	w.WriteString("<li>\n")
+	h := section.Headline
+	title := cleanHeadlineTitleForHTMLAnchorRegexp.ReplaceAllString(w.nodesAsString(h.Title...), "")
+	w.WriteString(fmt.Sprintf("<a href=\"#%s\">%s</a>\n", h.ID(), title))
+	if len(section.Children) != 0 {
+		w.WriteString("<ul>\n")
+		for _, section := range section.Children {
+			w.writeSection(section)
+		}
+		w.WriteString("</ul>\n")
+	}
+	w.WriteString("</li>\n")
+}
+
 func (w *HTMLWriter) writeHeadline(h Headline) {
 	for _, excludeTag := range w.excludeTags {
 		for _, tag := range h.Tags {
@@ -210,12 +239,7 @@ func (w *HTMLWriter) writeHeadline(h Headline) {
 		}
 	}
 
-	if id, ok := h.Properties.Get("CUSTOM_ID"); ok {
+	w.WriteString(fmt.Sprintf(`<h%d id="%s">`, h.Lvl, h.ID()) + "\n")
-		w.WriteString(fmt.Sprintf(`<h%d id="%s">`, h.Lvl, id) + "\n")
-	} else {
-		w.WriteString(fmt.Sprintf("<h%d>\n", h.Lvl))
-	}
-
 	if h.Status != "" {
 		w.WriteString(fmt.Sprintf(`<span class="todo">%s</span>`, h.Status) + "\n")
 	}

org/testdata/footnotes.html

@@ -1,4 +1,14 @@
-<h1>
+<nav>
+<ul>
+<li>
+<a href="#headline-1">Using some footnotes</a>
+</li>
+<li>
+<a href="#headline-2">Footnotes</a>
+</li>
+</ul>
+</nav>
+<h1 id="headline-1">
 Using some footnotes
 </h1>
 <ul>
@@ -24,7 +34,7 @@ Rather, they are gathered and exported at the end of the document in the footnot
 </p>
 </li>
 </ul>
-<h1>
+<h1 id="headline-2">
 Footnotes
 </h1>
 <p>

org/testdata/headlines.html

@@ -1,4 +1,26 @@
-<h1>
+<nav>
+<ul>
+<li>
+<a href="#headline-1">Simple Headline <code class="statistic">[1/2]</code></a>
+</li>
+<li>
+<a href="#headline-2">Headline with todo status &amp; priority</a>
+</li>
+<li>
+<a href="#this-will-be-the-id-of-the-headline">Headline with TODO status</a>
+</li>
+<li>
+<a href="#headline-4">Headline with tags &amp; priority</a>
+</li>
+<li>
+<a href="#headline-5">headline with custom status</a>
+</li>
+<li>
+<a href="#headline-6">excluded headline</a>
+</li>
+</ul>
+</nav>
+<h1 id="headline-1">
 Simple Headline <code class="statistic">[1/2]</code>
 </h1>
 <ul>
@@ -20,7 +42,7 @@ not just where they are actually meant to be - even here &gt; <code class="stati
 </p>
 </li>
 </ul>
-<h1>
+<h1 id="headline-2">
 <span class="todo">TODO</span>
 <span class="priority">[B]</span>
 Headline with todo status &amp; priority
@@ -32,7 +54,7 @@ Headline with TODO status
 <p>
 we can link to headlines that define a custom_id: <a href="#this-will-be-the-id-of-the-headline">#this-will-be-the-id-of-the-headline</a>
 </p>
-<h1>
+<h1 id="headline-4">
 <span class="priority">[A]</span>
 Headline with tags &amp; priority&#xa0;&#xa0;&#xa0;<span class="tags"><span>foo</span>&#xa0;<span>bar</span></span>
 </h1>
@@ -45,7 +67,7 @@ This is inside the drawer
 <p>
 Still outside the drawer
 </p>
-<h1>
+<h1 id="headline-5">
 <span class="todo">CUSTOM</span>
 headline with custom status
 </h1>

org/testdata/misc.html

@@ -1,12 +1,88 @@
-<h2>
+<nav>
+<ul>
+<li>
+<a href="#headline-1">issues from goorgeous (free test cases, yay!)</a>
+</li>
+<li>
+<a href="#headline-2">#19: Support #+HTML</a>
+</li>
+<li>
+<a href="#headline-3">#29: Support verse block</a>
+</li>
+<li>
+<a href="#headline-4">#31: Support #+INCLUDE</a>
+</li>
+<li>
+<a href="#headline-5">#33: Wrong output when mixing html with Org mode</a>
+</li>
+<li>
+<a href="#headline-6">#46: Support for symbols like ndash and mdash</a>
+</li>
+<li>
+<a href="#headline-7">#47: Consecutive <code>code</code> wrapped text gets joined</a>
+</li>
+<li>
+<a href="#headline-8">#50: LineBreaks in lists are preserved</a>
+</li>
+<li>
+<a href="#headline-9">#68: Quote block with inline markup</a>
+</li>
+<li>
+<a href="#headline-10">#72: Support for #+ATTR_HTML</a>
+</li>
+<li>
+<a href="#headline-11">#75: Not parsing nested lists correctly</a>
+</li>
+<li>
+<a href="#headline-12">#77: Recognize <code class="verbatim">code</code>— as code plus dash</a>
+</li>
+<li>
+<a href="#headline-13">#78: Emphasis at beginning of line</a>
+</li>
+<li>
+<a href="#headline-14">#82: Crash on empty headline</a>
+</li>
+<li>
+<a href="#headline-15"></a>
+</li>
+<li>
+<a href="#headline-16">#84: Paragraphs that are not followed by an empty line are not parsed correctly</a>
+</li>
+<li>
+<a href="#headline-17">Foo</a>
+</li>
+<li>
+<a href="#headline-18">Bar</a>
+</li>
+<li>
+<a href="#headline-19">#86: Multiple hyphens not converted to dashes</a>
+</li>
+<li>
+<a href="#headline-20">#87: Markup in footnotes is rendered literally</a>
+</li>
+<li>
+<a href="#headline-21">issues (wrongly) filed with hugo</a>
+</li>
+<li>
+<a href="#headline-22">#3874 exporting images in org mode</a>
+</li>
+<li>
+<a href="#headline-23">#4006 source code blocks in org not rendered correctly</a>
+</li>
+<li>
+<a href="#headline-24">Footnotes</a>
+</li>
+</ul>
+</nav>
+<h2 id="headline-1">
 issues from goorgeous (free test cases, yay!)
 </h2>
-<h3>
+<h3 id="headline-2">
 <span class="todo">DONE</span>
 <a href="https://github.com/chaseadamsio/goorgeous/issues/19">#19</a>: Support #+HTML
 </h3>
 <p style="border: 1px dotted grey">neato!</p>
-<h3>
+<h3 id="headline-3">
 <span class="todo">DONE</span>
 <a href="https://github.com/chaseadamsio/goorgeous/issues/29">#29:</a> Support verse block
 </h3>
@@ -23,7 +99,7 @@ or even a <strong>totally</strong> <em>custom</em> kind of block
 crazy ain&#39;t it?
 </p>
 </div>
-<h3>
+<h3 id="headline-4">
 <span class="todo">DONE</span>
 <a href="https://github.com/chaseadamsio/goorgeous/issues/31">#31</a>: Support #+INCLUDE
 </h3>
@@ -115,7 +191,7 @@ deploy:
 </pre>
 </li>
 </ul>
-<h3>
+<h3 id="headline-5">
 <span class="todo">DONE</span>
 <a href="https://github.com/chaseadamsio/goorgeous/issues/33">#33</a>: Wrong output when mixing html with Org mode
 </h3>
@@ -133,7 +209,7 @@ deploy:
 </tbody>
 </table>
 </div>
-<h3>
+<h3 id="headline-6">
 <span class="todo">DONE</span>
 <a href="https://github.com/chaseadamsio/goorgeous/issues/46">#46</a>: Support for symbols like ndash and mdash
 </h3>
@@ -164,7 +240,7 @@ note that —— is replaced with 2 mdashes and …. becomes ellipsis+. and so o
 </p>
 </li>
 </ul>
-<h3>
+<h3 id="headline-7">
 <span class="todo">DONE</span>
 <a href="https://github.com/chaseadamsio/goorgeous/issues/47">#47:</a> Consecutive <code>code</code> wrapped text gets joined
 </h3>
@@ -173,7 +249,7 @@ either <code>this</code> or <code>that</code> foo.
 either <code>this</code>
 or <code>that</code> foo.
 </p>
-<h3>
+<h3 id="headline-8">
 <span class="todo">DONE</span>
 <a href="https://github.com/chaseadamsio/goorgeous/issues/50">#50</a>: LineBreaks in lists are preserved
 </h3>
@@ -208,7 +284,7 @@ foo
 </p>
 </li>
 </ol>
-<h3>
+<h3 id="headline-9">
 <span class="todo">DONE</span>
 <a href="https://github.com/chaseadamsio/goorgeous/issues/68">#68</a>: Quote block with inline markup
 </h3>
@@ -217,12 +293,12 @@ foo
 <a href="https://www.example.com"><em>this</em> <strong>is</strong> <span style="text-decoration: underline;">markup</span>!</a>
 </p>
 </blockquote>
-<h3>
+<h3 id="headline-10">
 <span class="todo">DONE</span>
 <a href="https://github.com/chaseadamsio/goorgeous/issues/72">#72</a>: Support for #+ATTR_HTML
 </h3>
 <img src="https://golang.org/doc/gopher/pkg.png" alt="Go is fine though." title="https://golang.org/doc/gopher/pkg.png" id="gopher-image" width="300" style="border:2px solid black;"/>
-<h3>
+<h3 id="headline-11">
 <span class="todo">DONE</span>
 <a href="https://github.com/chaseadamsio/goorgeous/issues/75">#75</a>: Not parsing nested lists correctly
 </h3>
@@ -240,11 +316,11 @@ sub bullet
 </ul>
 </li>
 </ul>
-<h3>
+<h3 id="headline-12">
 <span class="todo">DONE</span>
 <a href="https://github.com/chaseadamsio/goorgeous/issues/77">#77</a>: Recognize <code class="verbatim">code</code>— as code plus dash
 </h3>
-<h3>
+<h3 id="headline-13">
 <span class="todo">DONE</span>
 <a href="https://github.com/chaseadamsio/goorgeous/issues/78">#78</a>: Emphasis at beginning of line
 </h3>
@@ -255,33 +331,33 @@ sub bullet
 Text
 <em>italics</em>
 </p>
-<h3>
+<h3 id="headline-14">
 <span class="todo">DONE</span>
 <a href="https://github.com/chaseadamsio/goorgeous/issues/82">#82</a>: Crash on empty headline
 </h3>
-<h4>
+<h4 id="headline-15">
 
 </h4>
 <p>
 just a space as title…
 </p>
-<h3>
+<h3 id="headline-16">
 <span class="todo">DONE</span>
 <a href="https://github.com/chaseadamsio/goorgeous/issues/84">#84</a>: Paragraphs that are not followed by an empty line are not parsed correctly
 </h3>
-<h4>
+<h4 id="headline-17">
 Foo
 </h4>
 <p>
 Foo paragraph.
 </p>
-<h4>
+<h4 id="headline-18">
 Bar
 </h4>
 <p>
 Bar paragraph
 </p>
-<h3>
+<h3 id="headline-19">
 <span class="todo">DONE</span>
 <a href="https://github.com/chaseadamsio/goorgeous/issues/86">#86</a>: Multiple hyphens not converted to dashes
 </h3>
@@ -327,17 +403,17 @@ src/example/export blocks should not be converted!
 </pre>
 </li>
 </ul>
-<h3>
+<h3 id="headline-20">
 <span class="todo">DONE</span>
 <a href="https://github.com/chaseadamsio/goorgeous/issues/87">#87</a>: Markup in footnotes is rendered literally
 </h3>
 <p>
 footnotes can contain <strong>markup</strong> - and other elements and stuff <sup class="footnote-reference"><a id="footnote-reference-2" href="#footnote-2">2</a></sup>
 </p>
-<h2>
+<h2 id="headline-21">
 issues (wrongly) filed with hugo
 </h2>
-<h3>
+<h3 id="headline-22">
 <a href="https://github.com/gohugoio/hugo/issues/3874">#3874</a> exporting images in org mode
 </h3>
 <p>
@@ -346,7 +422,7 @@ Hello, I&#39;m writing hugo blogs using org-mode.
 <p>
 When inserting an image link like <img src="/home/amos/Pictures/Screenshots/img-2017-09-11-165647.png" alt="/home/amos/Pictures/Screenshots/img-2017-09-11-165647.png" title="/home/amos/Pictures/Screenshots/img-2017-09-11-165647.png" />, hugo doesn&#39;t export the image.
 </p>
-<h3>
+<h3 id="headline-23">
 <a href="https://github.com/gohugoio/hugo/issues/4006">#4006</a> source code blocks in org not rendered correctly
 </h3>
 <div class="highlight">
@@ -359,7 +435,7 @@ When inserting an image link like <img src="/home/amos/Pictures/Screenshots/img-
   (ansi-term))
 </pre>
 </div>
-<h1>
+<h1 id="headline-24">
 Footnotes
 </h1>
 <div class="footnotes">