libxml2

Interface to libxml2, with DOM interface.

Why?

I needed to write go-xmlsec. This means we need to build trees using libxml2, and then muck with it in xmlsec: Two separate packages in Go means we cannot (safely) pass around C.xmlFooPtr objects (also, you pay a penalty for pointer types). This package carefully avoid references to C.xmlFooPtr types and uses uintptr to pass data around, so other libraries that needs to interact with libxml2 can safely interact with it.

Status

• This library should be considered alpha grade. API may still change.
• Much of commonly used functionalities from libxml2 that I use are there already, and are known to be functional

Package Layout:

Name	Description
libxml2	Globally available utility functions, such as ParseString
types	Common data types, such as types.Node
parser	Parser routines
dom	DOM-like manipulation of XML document/nodes
xpath	XPath related tools
xsd	XML Schema related tools
clib	Wrapper around C libxml2 library - DO NOT TOUCH IF UNSURE

Features

Create XML documents using DOM-like interface:

  d := dom.CreateDocument()
  e, err := d.CreateElement("foo")
  if err != nil {
    println(err)
    return
  }
  d.SetDocumentElement(e)
  ...

Parse documents:

  d, err := libxml2.ParseString(xmlstring)
  if err != nil {
    println(err)
    return
  }

Use XPath to extract node values:

  text := xpath.String(node.Find("//xpath/expression"))

Examples

Basic XML Example

import (
  "log"
  "net/http"

  "github.com/lestrrat-go/libxml2"
  "github.com/lestrrat-go/libxml2/parser"
  "github.com/lestrrat-go/libxml2/types"
  "github.com/lestrrat-go/libxml2/xpath"
)

func ExampleXML() {
  res, err := http.Get("http://blog.golang.org/feed.atom")
  if err != nil {
    panic("failed to get blog.golang.org: " + err.Error())
  }

  p := parser.New()
  doc, err := p.ParseReader(res.Body)
  defer res.Body.Close()

  if err != nil {
    panic("failed to parse XML: " + err.Error())
  }
  defer doc.Free()

  doc.Walk(func(n types.Node) error {
    log.Printf(n.NodeName())
    return nil
  })

  root, err := doc.DocumentElement()
  if err != nil {
    log.Printf("Failed to fetch document element: %s", err)
    return
  }

  ctx, err := xpath.NewContext(root)
  if err != nil {
    log.Printf("Failed to create xpath context: %s", err)
    return
  }
  defer ctx.Free()

  ctx.RegisterNS("atom", "http://www.w3.org/2005/Atom")
  title := xpath.String(ctx.Find("/atom:feed/atom:title/text()"))
  log.Printf("feed title = %s", title)
}

Basic HTML Example

func ExampleHTML() {
  res, err := http.Get("http://golang.org")
  if err != nil {
    panic("failed to get golang.org: " + err.Error())
  }

  doc, err := libxml2.ParseHTMLReader(res.Body)
  if err != nil {
    panic("failed to parse HTML: " + err.Error())
  }
  defer doc.Free()

  doc.Walk(func(n types.Node) error {
    log.Printf(n.NodeName())
    return nil
  })

  nodes := xpath.NodeList(doc.Find(`//div[@id="menu"]/a`))
  for i := 0; i < len(nodes); i++ {
    log.Printf("Found node: %s", nodes[i].NodeName())
  }
}

XSD Validation

import (
  "io/ioutil"
  "log"
  "os"
  "path/filepath"

  "github.com/lestrrat-go/libxml2"
  "github.com/lestrrat-go/libxml2/xsd"
)

func ExampleXSD() {
  xsdfile := filepath.Join("test", "xmldsig-core-schema.xsd")
  f, err := os.Open(xsdfile)
  if err != nil {
    log.Printf("failed to open file: %s", err)
    return
  }
  defer f.Close()

  buf, err := ioutil.ReadAll(f)
  if err != nil {
    log.Printf("failed to read file: %s", err)
    return
  }

  s, err := xsd.Parse(buf)
  if err != nil {
    log.Printf("failed to parse XSD: %s", err)
    return
  }
  defer s.Free()

  d, err := libxml2.ParseString(`<foo></foo>`)
  if err != nil {
    log.Printf("failed to parse XML: %s", err)
    return
  }

  if err := s.Validate(d); err != nil {
    for _, e := range err.(xsd.SchemaValidationError).Errors() {
      log.Printf("error: %s", e.Error())
    }
    return
  }

  log.Printf("validation successful!")
}

Caveats

Other libraries

There exists many similar libraries. I want speed, I want DOM, and I want XPath.When all of these are met, I'd be happy to switch to another library.

For now my closest contender was xmlpath, but as of this writing it suffers in the speed (for xpath) area a bit:

shoebill% go test -v -run=none -benchmem -benchtime=5s -bench .
PASS
BenchmarkXmlpathXmlpath-4     500000         11737 ns/op         721 B/op          6 allocs/op
BenchmarkLibxml2Xmlpath-4    1000000          7627 ns/op         368 B/op         15 allocs/op
BenchmarkEncodingXMLDOM-4    2000000          4079 ns/op        4560 B/op          9 allocs/op
BenchmarkLibxml2DOM-4        1000000         11454 ns/op         264 B/op          7 allocs/op
ok      github.com/lestrrat-go/libxml2  37.597s

See Also

• https://github.com/lestrrat-go/xmlsec

Credits

• Work on this library was generously sponsored by HDE Inc (https://www.hde.co.jp)