An XML parsing/building library built using nodejs. Also runs in the browser thanks to browserify.



Parse a little document at once:

var xml = ltx.parse('<document/>')

Push parser:

    p = new ltx.Parser()
    p.on('tree', function(tree) {
        proceed(null, tree);
    p.on('error', function(error) {


ltx implements multiple SAX backends:

  • node-expat: libexpat binding
  • ltx: fast native-JavaScript parser without error handling
  • saxjs: native-JavaScript parser

If present, they are available through ltx.availableSaxParsers. Mostly, you'll want to do:

    parser = new ltx.bestSaxParser();

Refer to lib/parse.js for the interface.

Element traversal

  • is(name, xmlns?): check
  • getName(): name without ns prefix
  • getNS(): element's xmlns, respects prefixes and searches upwards
  • findNS(prefix?): search for xmlns of a prefix upwards
  • getChild(name, xmlns?): find first child
  • getChildren(name, xmlns?): find all children
  • getChildByAttr(attr, value, xmlns?, recursive?): find first child by a specific attribute
  • getChildrenByAttr(attr, value, xmlns?, recursive?): find all children by a specific attribute
  • getChildrenByFilter(filter, recursive?): find all children by a custom filter function
  • getText(): appends all text nodes recursively
  • getChildText(name): a child's text contents
  • root(): uppermost parent in the tree
  • up(): parent or self

Element attributes

  • attrs is an object of the Element's attributes
  • name contains optional prefix, colon, name
  • parent points to its parent, this should always be consistent with children
  • children is an Array of Strings and Elements

Modifying XML Elements

  • new Element(name, attrs?): constructor
  • remove(child): remove child by reference
  • remove(name, xmlns): remove child by tag name and xmlns
  • attr(attrName, value?): modify or get an attribute's value
  • text(value?): modify or get the inner text
  • clone(): clones an element that is detached from the document

Building XML Elements

    var el = new ltx.Element('root').
    el.c('child', { age: 5 }).t('Hello').up()
      .c('child', { age: 7 }).t('Hello').up()
      .c('child', { age: 99 }).t('Hello').up()
    console.log('Serialized document:', el.root().toString());

This resembles Strophejs a bit.

strophejs' XML Builder is very convenient for producing XMPP stanzas. node-xmpp includes it in a much more primitive way: the c(), cnode() and t() methods can be called on any Element object, returning the child element.

This can be confusing: in the end, you will hold the last-added child until you use up(), a getter for the parent. Connection.send() first invokes tree() to retrieve the uppermost parent, the XMPP stanza, before sending it out the wire.

Destructive manipulation

Please always make sure parent and children are consistent. Don't append children of other parents to your own element. We're not adoption-safe!