Contents

• Building a tree from SAX events
• Producing SAX events from an ElementTree or Element
• Interfacing with pulldom/minidom

Building a tree from SAX events

First of all, lxml has support for building a new tree given SAX events. Todo this, we use the special SAX content handler defined by lxml namedlxml.sax.ElementTreeContentHandler:

>>>importlxml.sax>>>handler=lxml.sax.ElementTreeContentHandler()

Now let's fire some SAX events at it:

>>>handler.startElementNS((None,'a'),'a',{})>>>handler.startElementNS((None,'b'),'b',{(None,'foo'):'bar'})>>>handler.characters('Hello world')>>>handler.endElementNS((None,'b'),'b')>>>handler.endElementNS((None,'a'),'a')

This constructs an equivalent tree. You can access it through theetreeproperty of the handler:

>>>tree=handler.etree>>>lxml.etree.tostring(tree.getroot())b'<a><b foo="bar">Hello world</b></a>'

By passing amakeelement function the constructor ofElementTreeContentHandler, e.g. the one of a parser you configured, youcan determine which element class lookup scheme should be used.

Producing SAX events from an ElementTree or Element

Let's make a tree we can generate SAX events for:

>>>f=StringIO('<a><b>Text</b></a>')>>>tree=lxml.etree.parse(f)

To see whether the correct SAX events are produced, we'll write a customcontent handler.:

>>>fromxml.sax.handlerimportContentHandler>>>classMyContentHandler(ContentHandler):...def__init__(self):...self.a_amount=0...self.b_amount=0...self.text=None......defstartElementNS(self,name,qname,attributes):...uri,localname=name...iflocalname=='a':...self.a_amount+=1...iflocalname=='b':...self.b_amount+=1......defcharacters(self,data):...self.text=data

Note that it only defines the startElementNS() method and not startElement().The SAX event generator in lxml.sax currently only supports namespace-awareprocessing.

To test the content handler, we can produce SAX events from the tree:

>>>handler=MyContentHandler()>>>lxml.sax.saxify(tree,handler)

This is what we expect:

>>>handler.a_amount1>>>handler.b_amount1>>>handler.text'Text'

Interfacing with pulldom/minidom

lxml.sax is a simple way to interface with the standard XML support in thePython library. Note, however, that this is a one-way solution, as Python'sDOM implementation cannot generate SAX events from a DOM tree.

You can use xml.dom.pulldom to build a minidom from lxml:

>>>fromxml.dom.pulldomimportSAX2DOM>>>handler=SAX2DOM()>>>lxml.sax.saxify(tree,handler)

PullDOM makes the result available through thedocument attribute:

>>>dom=handler.document>>>print(dom.firstChild.localName)a