telegraph/content.go

70 lines
1.5 KiB
Go
Raw Normal View History

package telegraph
import (
"bytes"
"errors"
"strings"
2017-01-10 10:28:23 +00:00
"golang.org/x/net/html"
)
2016-12-25 03:10:55 +00:00
// ContentFormat transforms data to a DOM-based format to represent the
// content of the page.
func ContentFormat(data interface{}) ([]Node, error) {
var doc html.Node
switch dst := data.(type) {
case string:
dom, err := html.Parse(strings.NewReader(dst))
if err != nil {
return nil, err
}
doc = *dom
case []byte:
dom, err := html.Parse(bytes.NewReader(dst))
if err != nil {
return nil, err
}
doc = *dom
default:
return nil, errors.New("invalid data type, use []byte or string")
}
var content []Node
content = append(content, domToNode(doc.FirstChild))
return content, nil
}
func domToNode(domNode *html.Node) interface{} {
if domNode.Type == html.TextNode {
return domNode.Data
}
if domNode.Type != html.ElementNode {
return nil
}
var nodeElement NodeElement
switch strings.ToLower(domNode.Data) {
case "a", "aside", "b", "blockquote", "br", "code", "em", "figcaption", "figure", "h3", "h4", "hr", "i", "iframe", "img", "li", "ol", "p", "pre", "s", "strong", "u", "ul", "video":
2017-08-06 18:26:38 +00:00
nodeElement.Tag = domNode.Data
2017-09-04 21:09:59 +00:00
for i := range domNode.Attr {
switch strings.ToLower(domNode.Attr[i].Key) {
case "href", "src":
nodeElement.Attrs = map[string]string{
domNode.Attr[i].Key: domNode.Attr[i].Val,
}
default:
continue
}
}
}
for child := domNode.FirstChild; child != nil; child = child.NextSibling {
nodeElement.Children = append(nodeElement.Children, domToNode(child))
}
return nodeElement
}