htmlparse

Htmlparse is a go tool for parsing a html document.

It converts a html document into a tree. Each node in the tree is either a tag or a text. Given a tag, a programmer

can easily get its original infos, including its metadata, its children, its siblings and the text wrapped in it.

One can also modify a tree, by writing something into or delete a tag.

It can be used in web crawlers, analysis, batch formating and etc.

---

• Install
• Api

---

Install

go get -u github.com/tancehao/htmlparse

api

Parser

    type Parser struct {
        //unexported fields
    }

• NewParser(data []byte) *Parser

  Initiate a parser using a bytes.

• (p *Parser)Parse() (*Tag, error)

  The only one method needed to convert the original bytes to a tree. The root tag was returned.

  Example:

   	import (
   		"github.com/tancehao/htmlparse"		
   	)
       
   	//...
   	
       content, _ := ioutil.ReadFile("index.html")
   	parser := htmlparse.NewParser(content)
   	root, err := parser.Parse()

Tag

    type Tag struct {
        TagName    string
        Attributes map[string]string
        Class      []string
        NoEnd      bool    //whether it's a single tag
        
        //unexported fields
    }

• (t *Tag)HasClass(class string) bool

  Whether a given tag has a class.

  Example

      div, _ := htmlparse.NewParser(`<div class="product box">...</div>`).Parse()
      fmt.Println(div.HasClass("box"))  // true

• (t *Tag)Find(map[string]string) *TagSets

  Find the tags from a tag's children by a set of conditions.

  Example

      formBytes := []byte(`<form method="post" action"test">
        <input type="hidden" name="foo1" value="bar1" />
        <input type="hidden" name="foo2" value="bar2" />
        <input type="text" name="foo3" value="bar3" />
        <textarea name="foo4" value="bar4" />
        </form>
      `)
      formTag, _ := htmlparse.NewParser(formBytes).Parse()
      hiddenInputes := formTag.Find(map[string]string{
          "tagName": "input",
          "type": "hidden",
      })
      fmt.Println(hiddenInputes)
      //<input type="hidden" name="foo1" value="bar1" />
      //<input type="hidden" name="foo2" value="bar2" />

• (t *Tag)FindByName(name string) *TagSets

  Find a set of tags by their names.

• (t *Tag)FindByClass(class string) *TagSets

  Find a set of tags by their classes.

• (t *Tag)FindByFunc(f func(*Tag) bool) *TagSets

  Find a set of tags by a function which tells whether a tag should by returned.

  Example:

      root.FindByFunc(func(t *htmlparse.Tag) bool {
          cond1 := t.TagName == "img"
          cond2 := t.HasClass("product")
          cond3 := strings.Contains(t.Attributes["Src"], ".png")
          return cond1 && cond2 && cond3
      })

• (t *Tag)FindByCssSelector(path string) *TagSets

  Find a set of tags by a css selector.

  Example:

      productBytes := []byte(`<div class="product">
          <div class="pictures"></div>
          <div class="infos">
              <div class="brand">BRAND</brand>
              <div class="name">NAME</div>
              <div class="prices">
                  $1000.00 <span><s>$1200.00</s></span>
              </div>
          </div>
      </div>`)
      productTag, _ := htmlparse.Next(productBytes).Parse()
      originalPrices := productTag.FindByCssSelector(" .infos .prices span s")
      fmt.Println(originalPrices)
      //$1200.00

• (t *Tag)GetContent() []byte

  Return the original bytes of a tag in the document, the tag's metadata is included.
  By design, each tag or text has a pair of pointers pointing to its head and tail in the original document. This function searchs for the contents using the pair of indexes, which guarantees the speed when it works.

• (t *Tag)String() string

  Implements the fmt.Stringer.

• (t *Tag)Extract() []byte

  Filter the text from the original data of a tag. Tags wont't be included.

  Example

      navifatorBytes := []byte(`<div class="navs">
          <a href="/category/society">Society</a>
          <a href="/category/sports">Sports</a>
          <a href="/category/mil">Mil</a>
          <a href="/category/cars">Cars</a>
          <a href="/category/tech">Tech</a>
      </div>`)
      navigatorTag, _ := htmlparse.NewParser(navifatorBytes).Parse()
      fmt.Println(string(navigatorTag.Extract()))
      //Society
      //Sports
      //Mil
      //Cars
      //Tech

• (t *Tag)Index() int

  Get the index of a tag among its siblings.

• (t *Tag)Prev() *Tag

  Get the previous tag of an existing one under the same parent.

• (t *Tag)Next() *Tag

  Similar to Prev().

• (t *Tag)Modify() string

  Return the modified data of a tag. When reading something from a tag, one gets the original data from the document. However, when writing a tag, one is not writing to the original document but the tag struct, so this method gets the original data plus the differences one has made to a tag.

• (t *Tag)WriteText(position int, data []byte) (*Text, error)

  Write text into a tag at the given index in the tag's children.
  Example:

   	names := products.Find(map[string]string{"class": "productName"})
   	fmt.Println(names)
   	//prints:
   	//<div class="productName">Product1</div>
   	//<div class="productName">Product2</div>
   	//<div class="productName">Product3</div>
  
   	for _, name := range names.All() {
   		name.WriteText(0, []byte("[ONSALE] "))
   		fmt.Println(name.Modify())
   	}
  
   	//prints:
   	//<div class="productName">[ONSALE] Product1</div>
   	//<div class="productName">[ONSALE] Product2</div>
   	//<div class="productName">[ONSALE] Product3</div>

• (t *Tag)WriteTag(position int, tagname string) (*Tag, error)

  Write a tag into a tag at the given index in the tag's chidren. Example:

   	script, _ := body.WriteTag(1000, "script")
   	//if ths position is greater than the count of the tag's children, it'll be set to the last
   	script.Attributes["src"] = "http://www.foo.com"
   	fmt.Println(body.Modify())
  
       //a script tag was inserted into the html document.

• (t *Tag)Delete() error

  Delete a tag.
  Example:

   	garbages := html.FindByClass("ad").All()
       for _, g := range garbages {
           g.Delete()
       }
       fmt.Println(html.Modify())

TagSets

    type TagSets struct {
        //unexported fields
    }

• (ts *TagSets)Find(map[string]string) *TagSets

  Return a set of tags from a set of tags or their children using a filter.
  It can be used in a chain style. Example:

   	photos := tree.Find(map[string]string{
   		"tagName": "div", 
   		"class": "product",
   	}).Find(map[string]string{
   		"tagName": "img",
   		"class": "photo",
   	})

• (ts *TagSets)All() []*Tag

  Get all the tags in this set.

• (ts *TagSets)GetAttributes(attr ...string) []map[string]string

  Get some attributes from each tag in this set.
  Example:

       
       formBytes := []byte(`<form method="post" action"test">
         <input type="hidden" name="foo1" value="bar1" />
         <input type="hidden" name="foo2" value="bar2" />
         <input type="text" name="foo3" value="bar3" />
         <textarea name="foo4" value="bar4" />
         </form>
       `)
       formTag, _ := htmlparse.NewParser(formBytes).Parse()
   	inputs := formTag.Find(map[string]string{
   		"tagName": "input"
   	}).GetAttributes("type", "name", "value")
   	for _, input := range inputs {
   		fmt.Printf("%s,%s,%s\n", 
   			input["type"], input["name"], input["value"]
   		)
   	}
       //hidden,foo1,bar1
       //hidden,foo2,bar2
       //text,foo3,bar3
       

• (ts *TagSets)String() string

Text

    type Text struct {
        Text []byte
        
        //unexported fields
    }

• (ts *Text)String() string

  Smilar to tag.

• (ts *Text)Index() int

  Smilar to tag.

• (t *Text)Modify() string

  Smilar to tag.

• (t *Text)Delete()

  Smilar to tag.