Last active
May 25, 2023 14:22
-
-
Save HalCanary/fd4ec75ae950196454f09051a5d8779a to your computer and use it in GitHub Desktop.
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
package main | |
import ( | |
"regexp" | |
"sort" | |
"strings" | |
"golang.org/x/net/html" | |
) | |
func AppendNodes(node *html.Node, children ...*html.Node) *html.Node { | |
for _, c := range children { | |
if c != nil { | |
node.AppendChild(c) | |
} | |
} | |
return node | |
} | |
func DocumentNode(doctype string, children ...*html.Node) *html.Node { | |
node := &html.Node{Type: html.DocumentNode} | |
node.AppendChild(&html.Node{Type: html.DoctypeNode, Data: doctype}) | |
return AppendNodes(node, children...) | |
} | |
func TextNode(data string) *html.Node { | |
return &html.Node{Type: html.TextNode, Data: data} | |
} | |
func CommentNode(data string) *html.Node { | |
return &html.Node{Type: html.CommentNode, Data: data} | |
} | |
// Return an element with given attributes and children. | |
func Element(tag string, attributes map[string]string, children ...*html.Node) *html.Node { | |
node := &html.Node{Type: html.ElementNode, Data: tag} | |
if len(attributes) > 0 { | |
keys := make([]string, 0, len(attributes)) | |
for k := range attributes { | |
keys = append(keys, k) | |
} | |
sort.Strings(keys) | |
for _, k := range keys { | |
node.Attr = append(node.Attr, html.Attribute{Key: k, Val: attributes[k]}) | |
} | |
} | |
return AppendNodes(node, children...) | |
} | |
//////////////////////////////////////////////////////////////////////////////// | |
func FindNodeByTagAndAttrib(root *html.Node, tag, key, value string) *html.Node { | |
// I unrolled a recursive function to use no recursion, or heap allocation, | |
// only loops. This is allowed by the existance of Parent pointer. | |
if root == nil { | |
return nil | |
} | |
node := root | |
for { | |
if node.Type == html.ElementNode && (tag == "" || node.Data == tag) { | |
if key == "" { | |
return node | |
} | |
for _, attr := range node.Attr { | |
if attr.Key == key && attr.Val == value { | |
return node | |
} | |
} | |
} | |
if node.FirstChild != nil { | |
node = node.FirstChild | |
continue | |
} | |
for { | |
if node == root || node == nil { | |
return nil | |
} | |
if node.NextSibling != nil { | |
node = node.NextSibling | |
break | |
} | |
node = node.Parent | |
} | |
} | |
} | |
func FindNodeByTag(node *html.Node, tag string) *html.Node { | |
return FindNodeByTagAndAttrib(node, tag, "", "") | |
} | |
func FindNodeById(node *html.Node, id string) *html.Node { | |
return FindNodeByTagAndAttrib(node, "", "id", id) | |
} | |
func FindNodeByAttribute(node *html.Node, key, value string) *html.Node { | |
return FindNodeByTagAndAttrib(node, "", key, value) | |
} | |
var whitespaceRegexp = regexp.MustCompile("\\s+") | |
// Extract and combine all Text Nodes under given node. | |
func ExtractText(root *html.Node) string { | |
if root == nil { | |
return "" | |
} | |
var accumulator strings.Builder | |
node := root | |
for { | |
if node.Type == html.TextNode { | |
accumulator.WriteString(whitespaceRegexp.ReplaceAllString(node.Data, " ")) | |
} | |
if node.Type == html.ElementNode { | |
switch node.Data { | |
case "br": | |
accumulator.WriteString("\n") | |
case "hr": | |
accumulator.WriteString("\n* * *\n") | |
case "p": | |
accumulator.WriteString("\n\n") | |
} | |
} | |
if node.FirstChild != nil { | |
node = node.FirstChild | |
continue | |
} | |
for { | |
if node == root || node == nil { | |
return accumulator.String() | |
} | |
if node.NextSibling != nil { | |
node = node.NextSibling | |
break | |
} | |
node = node.Parent | |
} | |
} | |
} | |
//////////////////////////////////////////////////////////////////////////////// | |
func example(w io.Writer) error { | |
return html.Render(w, | |
DocumentNode("html", | |
TextNode("\n"), | |
Element("html", map[string]string{"lang": "en"}, | |
TextNode("\n"), | |
Element("head", nil, | |
TextNode("\n"), | |
Element("meta", map[string]string{"charset": "utf-8"}), | |
TextNode("\n"), | |
CommentNode("\nthis is an example\n"), | |
TextNode("\n"), | |
Element("meta", map[string]string{ | |
"name": "viewport", "content": "width=device-width, initial-scale=1.0"}), | |
TextNode("\n"), | |
Element("title", nil, TextNode("Hello World!")), | |
TextNode("\n"), | |
Element("style", nil, TextNode("\nbody{font-family: sans-serif;}\n")), | |
TextNode("\n"), | |
), | |
TextNode("\n"), | |
Element("body", nil, | |
Element("h1", nil, TextNode("Hello World!")), | |
), | |
TextNode("\n"), | |
), | |
TextNode("\n"), | |
), | |
) | |
} |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment