Refactor keyword parsing/rendering & add support for ATTR_HTML

This commit is contained in:
Niklas Fasching 2018-12-11 17:30:02 +01:00
parent a859264420
commit 81f74f4ad9
9 changed files with 169 additions and 76 deletions

View file

@ -83,7 +83,11 @@ table, .highlight > pre, pre.example {
font-size: 0.85rem; font-size: 0.85rem;
border: 1px solid rgba(250, 100, 50, 0.5); } border: 1px solid rgba(250, 100, 50, 0.5); }
.caption { figure {
margin: 1em 0;
}
figcaption {
font-family: monospace; font-family: monospace;
font-size: 0.75em; font-size: 0.75em;
text-align: center; text-align: center;

View file

@ -4,6 +4,9 @@ import (
"fmt" "fmt"
"html" "html"
"strings" "strings"
h "golang.org/x/net/html"
"golang.org/x/net/html/atom"
) )
type HTMLWriter struct { type HTMLWriter struct {
@ -45,10 +48,14 @@ func (w *HTMLWriter) emptyClone() *HTMLWriter {
return &wcopy return &wcopy
} }
func (w *HTMLWriter) before(d *Document) { func (w *HTMLWriter) nodesAsString(nodes ...Node) string {
w.document = d tmp := w.emptyClone()
tmp.writeNodes(nodes...)
return tmp.String()
} }
func (w *HTMLWriter) before(d *Document) {}
func (w *HTMLWriter) after(d *Document) { func (w *HTMLWriter) after(d *Document) {
w.writeFootnotes(d) w.writeFootnotes(d)
} }
@ -199,9 +206,7 @@ func (w *HTMLWriter) writeRegularLink(l RegularLink) {
} }
description := url description := url
if l.Description != nil { if l.Description != nil {
descriptionWriter := w.emptyClone() description = w.nodesAsString(l.Description...)
descriptionWriter.writeNodes(l.Description...)
description = descriptionWriter.String()
} }
switch l.Kind() { switch l.Kind() {
case "image": case "image":
@ -245,17 +250,27 @@ func (w *HTMLWriter) writeHorizontalRule(h HorizontalRule) {
w.WriteString("<hr>\n") w.WriteString("<hr>\n")
} }
func (w *HTMLWriter) writeNodeWithMeta(m NodeWithMeta) { func (w *HTMLWriter) writeNodeWithMeta(n NodeWithMeta) {
nodeW := w.emptyClone() out := w.nodesAsString(n.Node)
nodeW.writeNodes(m.Node) if p, ok := n.Node.(Paragraph); ok {
nodeString := nodeW.String() if len(p.Children) == 1 && isImageOrVideoLink(p.Children[0]) {
if rawCaption, ok := m.Meta["CAPTION"]; ok { out = w.nodesAsString(p.Children[0])
nodes, captionW := w.document.parseInline(rawCaption), w.emptyClone() }
captionW.writeNodes(nodes...)
caption := `<p class="caption">` + "\n" + captionW.String() + "\n</p>\n"
nodeString = `<div class="captioned">` + "\n" + nodeString + caption + `</div>` + "\n"
} }
w.WriteString(nodeString) for _, attributes := range n.Meta.HTMLAttributes {
out = withHTMLAttributes(out, attributes...) + "\n"
}
if len(n.Meta.Caption) != 0 {
caption := ""
for i, ns := range n.Meta.Caption {
if i != 0 {
caption += " "
}
caption += w.nodesAsString(ns...)
}
out = fmt.Sprintf("<figure>\n%s<figcaption>\n%s\n</figcaption>\n</figure>\n", out, caption)
}
w.WriteString(out)
} }
func (w *HTMLWriter) writeTable(t Table) { func (w *HTMLWriter) writeTable(t Table) {
@ -289,3 +304,24 @@ func (w *HTMLWriter) writeTableHeader(t TableHeader) {
func (w *HTMLWriter) writeTableSeparator(t TableSeparator) { func (w *HTMLWriter) writeTableSeparator(t TableSeparator) {
w.WriteString("<tr></tr>\n") w.WriteString("<tr></tr>\n")
} }
func withHTMLAttributes(input string, kvs ...string) string {
if len(kvs)%2 != 0 {
panic(fmt.Sprintf("len of kvs must be even: %#v", kvs))
}
context := &h.Node{Type: h.ElementNode, Data: "body", DataAtom: atom.Body}
nodes, err := h.ParseFragment(strings.NewReader(strings.TrimSpace(input)), context)
if err != nil || len(nodes) != 1 {
panic(fmt.Sprintf("could not extend html attributes of %s: %v (%s)", input, len(nodes), err))
}
out, node := strings.Builder{}, nodes[0]
for i := 0; i < len(kvs)-1; i += 2 {
k, v := strings.TrimPrefix(kvs[i], ":"), kvs[i+1]
node.Attr = append(node.Attr, h.Attribute{Namespace: "", Key: k, Val: v})
}
err = h.Render(&out, nodes[0])
if err != nil {
panic(fmt.Sprintf("could not extend html attributes of %s: %#v (%s)", input, nodes, err))
}
return out.String()
}

View file

@ -1,6 +1,7 @@
package org package org
import ( import (
"encoding/csv"
"regexp" "regexp"
"strings" "strings"
) )
@ -12,14 +13,18 @@ type Keyword struct {
type NodeWithMeta struct { type NodeWithMeta struct {
Node Node Node Node
Meta map[string]string Meta Metadata
}
type Metadata struct {
Caption [][]Node
HTMLAttributes [][]string
} }
type Comment struct{ Content string } type Comment struct{ Content string }
var keywordRegexp = regexp.MustCompile(`^(\s*)#\+([^:]+):(\s+(.*)|(\s*)$)`) var keywordRegexp = regexp.MustCompile(`^(\s*)#\+([^:]+):(\s+(.*)|(\s*)$)`)
var commentRegexp = regexp.MustCompile(`^(\s*)#(.*)`) var commentRegexp = regexp.MustCompile(`^(\s*)#(.*)`)
var affiliatedKeywordRegexp = regexp.MustCompile(`^(CAPTION)$`)
func lexKeywordOrComment(line string) (token, bool) { func lexKeywordOrComment(line string) (token, bool) {
if m := keywordRegexp.FindStringSubmatch(line); m != nil { if m := keywordRegexp.FindStringSubmatch(line); m != nil {
@ -30,47 +35,44 @@ func lexKeywordOrComment(line string) (token, bool) {
return nilToken, false return nilToken, false
} }
func (d *Document) parseKeyword(i int, stop stopFn) (int, Node) {
k := parseKeyword(d.tokens[i])
if affiliatedKeywordRegexp.MatchString(k.Key) {
consumed, node := d.parseAffiliated(i, stop)
if consumed != 0 {
return consumed, node
}
} else {
d.BufferSettings[k.Key] = strings.Join([]string{d.BufferSettings[k.Key], k.Value}, "\n")
}
return 1, k
}
func (d *Document) parseComment(i int, stop stopFn) (int, Node) { func (d *Document) parseComment(i int, stop stopFn) (int, Node) {
return 1, Comment{d.tokens[i].content} return 1, Comment{d.tokens[i].content}
} }
func (d *Document) parseAffiliated(i int, stop stopFn) (int, Node) { func (d *Document) parseKeyword(i int, stop stopFn) (int, Node) {
start, meta := i, map[string]string{} k := parseKeyword(d.tokens[i])
for ; !stop(d, i) && d.tokens[i].kind == "keyword"; i++ { if k.Key == "CAPTION" || k.Key == "ATTR_HTML" {
k := parseKeyword(d.tokens[i]) consumed, node := d.parseAffiliated(i, stop)
if !affiliatedKeywordRegexp.MatchString(k.Key) { if consumed != 0 {
return 0, nil return consumed, node
} }
if value, ok := meta[k.Key]; ok { }
meta[k.Key] = value + " " + k.Value d.BufferSettings[k.Key] = strings.Join([]string{d.BufferSettings[k.Key], k.Value}, "\n")
} else { return 1, k
meta[k.Key] = k.Value }
func (d *Document) parseAffiliated(i int, stop stopFn) (int, Node) {
start, meta := i, Metadata{}
for ; !stop(d, i) && d.tokens[i].kind == "keyword"; i++ {
switch k := parseKeyword(d.tokens[i]); k.Key {
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
}
meta.HTMLAttributes = append(meta.HTMLAttributes, attributes)
default:
return 0, nil
} }
} }
if stop(d, i) { if stop(d, i) {
return 0, nil return 0, nil
} }
consumed, node := 0, (Node)(nil) consumed, node := d.parseOne(i, stop)
if t := d.tokens[i]; t.kind == "text" {
if nodes := d.parseInline(t.content); len(nodes) == 1 && isImageOrVideoLink(nodes[0]) {
consumed, node = 1, Paragraph{nodes[:1]}
}
} else {
consumed, node = d.parseOne(i, stop)
}
if consumed == 0 || node == nil { if consumed == 0 || node == nil {
return 0, nil return 0, nil
} }

View file

@ -173,11 +173,25 @@ func (w *OrgWriter) writeKeyword(k Keyword) {
w.WriteString(w.indent + fmt.Sprintf("#+%s: %s\n", k.Key, k.Value)) w.WriteString(w.indent + fmt.Sprintf("#+%s: %s\n", k.Key, k.Value))
} }
func (w *OrgWriter) writeNodeWithMeta(m NodeWithMeta) { func (w *OrgWriter) writeNodeWithMeta(n NodeWithMeta) {
for k, v := range m.Meta { for _, ns := range n.Meta.Caption {
w.writeNodes(Keyword{k, v}) w.WriteString("#+CAPTION: ")
w.writeNodes(ns...)
w.WriteString("\n")
} }
w.writeNodes(m.Node) for _, attributes := range n.Meta.HTMLAttributes {
w.WriteString("#+ATTR_HTML: ")
for i := 0; i < len(attributes)-1; i += 2 {
w.WriteString(attributes[i] + " ")
if strings.ContainsAny(attributes[i+1], "\t ") {
w.WriteString(`"` + attributes[i+1] + `"`)
} else {
w.WriteString(attributes[i+1])
}
}
w.WriteString("\n")
}
w.writeNodes(n.Node)
} }
func (w *OrgWriter) writeComment(c Comment) { func (w *OrgWriter) writeComment(c Comment) {

View file

@ -1,4 +1,4 @@
<div class="captioned"> <figure>
<div class="highlight"> <div class="highlight">
<pre> <pre>
echo &#34;a bash source block&#34; echo &#34;a bash source block&#34;
@ -10,10 +10,10 @@ function hello {
hello hello
</pre> </pre>
</div> </div>
<p class="caption"> <figcaption>
block caption block caption
</p> </figcaption>
</div> </figure>
<div class="highlight"> <div class="highlight">
<pre> <pre>
a source block without a language a source block without a language

View file

@ -1,27 +1,30 @@
<p> <p>
Anything can be captioned. Also captions are not real, correct captions but just a paragraph below the element (bothe wrapped into a div) Anything can be captioned. Also captions are not real, correct captions but just a paragraph below the element (bothe wrapped into a div)
</p> </p>
<div class="captioned"> <figure>
<div class="highlight"> <div class="highlight">
<pre> <pre>
echo &#34;i have a caption!&#34; echo &#34;i have a caption!&#34;
</pre> </pre>
</div> </div>
<p class="caption"> <figcaption>
captioned soure block captioned soure block
</p> </figcaption>
</div> </figure>
<div class="captioned"> <figure>
<img src="http://placekitten.com/200/200#.png" alt="http://placekitten.com/200/200#.png" title="http://placekitten.com/200/200#.png" /><figcaption>
captioned link (image in this case)
</figcaption>
</figure>
<p> <p>
<video src="my-video.mp4" title="my-video.mp4">my-video.mp4</video> note that the whole paragraph is captioned, so a linebreak is needed for images to caption correctly
</p> </p>
<p class="caption"> <figure>
captioned link (video in this case)
</p>
</div>
<p> <p>
note that only that one line is captioned, not the whole paragraph <img src="http://placekitten.com/200/200#.png" alt="http://placekitten.com/200/200#.png" title="http://placekitten.com/200/200#.png" />
</p> see?
<p>
also, normal text lines can&#39;t be captioned
</p> </p>
<figcaption>
captioned link (image in this case)
</figcaption>
</figure>

View file

@ -5,9 +5,12 @@ Anything can be captioned. Also captions are not real, correct captions but just
echo "i have a caption!" echo "i have a caption!"
#+END_SRC #+END_SRC
#+CAPTION: captioned link (video in this case) #+CAPTION: captioned link (image in this case)
[[my-video.mp4]] [[http://placekitten.com/200/200#.png]]
note that only that one line is captioned, not the whole paragraph
note that the whole paragraph is captioned, so a linebreak is needed for images to caption correctly
#+CAPTION: captioned link (image in this case)
[[http://placekitten.com/200/200#.png]]
see?
#+CAPTION: not happening!
also, normal text lines can't be captioned

18
org/testdata/keywords.html vendored Normal file
View file

@ -0,0 +1,18 @@
<figure>
<div class="highlight" class="a b c" id="it">
<pre>echo &#34;a bash source block with custom html attributes&#34;
</pre>
</div>
<figcaption>
and <span style="text-decoration: underline;">multiple</span> lines of <strong>captions</strong>!
</figcaption>
</figure>
<p>
and an image with custom html attributes and a caption
</p>
<figure>
<img src="http://placekitten.com/200/200#.png" alt="http://placekitten.com/200/200#.png" title="http://placekitten.com/200/200#.png" style="border: 10px solid black"/>
<figcaption>
kittens!
</figcaption>
</figure>

13
org/testdata/keywords.org vendored Normal file
View file

@ -0,0 +1,13 @@
#+CAPTION: and _multiple_
#+CAPTION: lines of *captions*!
#+ATTR_HTML: :class "a b c"
#+ATTR_HTML: :id it
#+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"
[[http://placekitten.com/200/200#.png]]