Improve ATTR_HTML keyword parsing

Org mode separates kvs not as initially assumed by whitespace (~ csv) but
rather at keywords (~ :\w+).

This is still not replicating Org mode behaviour though as I decided against
attributes ignoring multi-definitions. Instead we stack their
values (and those existing on the element) for certain attributes (class, style
for now).

e.g.
[[foo]]

would become <foo class="a"> in Org mode but becomes <foo class="a b"> with
go-org.

README.org

@@ -3,9 +3,6 @@ A basic Org mode parser in go.
 Take a look at [[https://niklasfasching.github.io/go-org/][github pages]] for some examples and an online org -> html demo (requires wasm support).
 * next
 - test against [[https://raw.githubusercontent.com/kaushalmodi/ox-hugo/master/test/site/content-org/all-posts.org][ox-hugo all-posts.org]]
-  - fix attr_html splitting - not whitespace, :x keywords, e.g.
-    #+attr_html: :caption A unicorn! :style width: 100%; height: 100%; :caption foo
-    [[/images/org-mode-unicorn-logo.png]]
 - more keywords: https://orgmode.org/manual/In_002dbuffer-settings.html
 - headlines
   - unique ids

org/keyword.go

@@ -1,7 +1,6 @@
 package org
 
 import (
-	"encoding/csv"
 	"fmt"
 	"io/ioutil"
 	"path/filepath"
@@ -35,6 +34,7 @@ var keywordRegexp = regexp.MustCompile(`^(\s*)#\+([^:]+):(\s+(.*)|(\s*)$)`)
 var commentRegexp = regexp.MustCompile(`^(\s*)#(.*)`)
 
 var includeFileRegexp = regexp.MustCompile(`(?i)^"([^"]+)" (src|example|export) (\w+)$`)
+var attributeRegexp = regexp.MustCompile(`(?:^|\s+)(:[-\w]+)\s+(.*)$`)
 
 func lexKeywordOrComment(line string) (token, bool) {
 	if m := keywordRegexp.FindStringSubmatch(line); m != nil {
@@ -77,11 +77,21 @@ func (d *Document) parseAffiliated(i int, stop stopFn) (int, Node) {
 		case "CAPTION":
 			meta.Caption = append(meta.Caption, d.parseInline(k.Value))
 		case "ATTR_HTML":
-			r := csv.NewReader(strings.NewReader(k.Value))
-			r.Comma = ' '
-			attributes, err := r.Read()
-			if err != nil {
-				return 0, nil
+			attributes, rest := []string{}, k.Value
+			for {
+				if k, m := "", attributeRegexp.FindStringSubmatch(rest); m != nil {
+					k, rest = m[1], m[2]
+					attributes = append(attributes, k)
+					if v, m := "", attributeRegexp.FindStringSubmatchIndex(rest); m != nil {
+						v, rest = rest[:m[0]], rest[m[0]:]
+						attributes = append(attributes, v)
+					} else {
+						attributes = append(attributes, strings.TrimSpace(rest))
+						break
+					}
+				} else {
+					break
+				}
 			}
 			meta.HTMLAttributes = append(meta.HTMLAttributes, attributes)
 		default:

org/org.go

@@ -184,11 +184,6 @@ func (w *OrgWriter) writeNodeWithMeta(n NodeWithMeta) {
 	}
 	for _, attributes := range n.Meta.HTMLAttributes {
 		w.WriteString("#+ATTR_HTML: ")
-		for i := 0; i < len(attributes)-1; i += 2 {
-			if strings.ContainsAny(attributes[i+1], "\t ") {
-				attributes[i+1] = fmt.Sprintf(`"%s"`, attributes[i+1])
-			}
-		}
 		w.WriteString(strings.Join(attributes, " ") + "\n")
 	}
 	w.writeNodes(n.Node)

org/testdata/keywords.html

@@ -1,5 +1,5 @@
 <figure>
-<div class="highlight a b c" id="it">
+<div class="highlight a b c d" id="it">
 <pre>echo &#34;a bash source block with custom html attributes&#34;
 </pre>
 </div>
@@ -11,7 +11,7 @@ and <span style="text-decoration: underline;">multiple</span> lines of <strong>c
 and an image with custom html attributes and a caption
 </p>
 <figure>
-<img src="https://placekitten.com/200/200#.png" alt="https://placekitten.com/200/200#.png" title="https://placekitten.com/200/200#.png" style="border: 10px solid black"/>
+<img src="https://placekitten.com/200/200#.png" alt="https://placekitten.com/200/200#.png" title="https://placekitten.com/200/200#.png" style="height: 100%; border: 10px solid black;" id="kittens"/>
 <figcaption>
 kittens!
 </figcaption>

org/testdata/keywords.org

@@ -1,13 +1,14 @@
 
 #+CAPTION: and _multiple_
 #+CAPTION: lines of *captions*!
-#+ATTR_HTML: :class "a b c"
-#+ATTR_HTML: :id it
+#+ATTR_HTML: :class a b
+#+ATTR_HTML: :id it :class c d
 #+BEGIN_SRC sh
 echo "a bash source block with custom html attributes"
 #+END_SRC
 
 and an image with custom html attributes and a caption
 #+CAPTION: kittens!
-#+ATTR_HTML: :style "border: 10px solid black"
+#+ATTR_HTML: :style height: 100%; :id overwritten
+#+ATTR_HTML: :style border: 10px solid black; :id kittens
 [[https://placekitten.com/200/200#.png]]

org/testdata/misc.html

@@ -207,7 +207,7 @@ foo
 <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" width="300" style="border:2px solid black;"/>
+<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>
 <span class="todo">DONE</span>
 <a href="https://github.com/chaseadamsio/goorgeous/issues/75">#75</a>: Not parsing nested lists correctly

org/testdata/misc.org

@@ -60,8 +60,8 @@ or ~that~ foo.
 [[https://www.example.com][/this/ *is* _markup_!]]
 #+END_QUOTE
 *** DONE [[https://github.com/chaseadamsio/goorgeous/issues/72][#72]]: Support for #+ATTR_HTML
-#+ATTR_HTML: :alt "Go is fine though."
-#+ATTR_HTML: :width 300 :style "border:2px solid black;"
+#+ATTR_HTML: :alt Go is fine though. :id gopher-image
+#+ATTR_HTML: :width 300 :style border:2px solid black;
 [[https://golang.org/doc/gopher/pkg.png]]
 *** DONE [[https://github.com/chaseadamsio/goorgeous/issues/75][#75]]: Not parsing nested lists correctly
 - bullet 1