Django get a DOM object from an HttpResponse

Question

I want to do a unit test on a html page which is returned as a byte string in an HttpResponse object... e.g. "find_elements_by_tag_name". Is the solution simply to xml.dom.minidom.parseString the bytes of response.content?

I couldn't find any examples of people doing this online or in Django manuals or tutorials, which makes me wonder if there's a reason for not doing it this way? If it's bad practice and there's a better way to do this please can you say why and what?

Answer 1

for interest, this is what I came up with in the end:

def test_that_all_inputs_have_name_attr( self ):
    response = home_page( HttpRequest() )
    from xml.dom.minidom import parseString
    page_as_doc = parseString( response.content )
    inputs = page_as_doc.getElementsByTagName( 'input' )
    for input_el in inputs:
        attr_map = input_el.attributes
        fail_msg = ''
        for i in range( attr_map.length ):
            attr = attr_map.item( i )
            fail_msg += "\nname '%s' value '%s'" % ( attr.name, attr.value )
        if not fail_msg: 
            fail_msg = "none!"
        self.assertTrue( input_el.hasAttribute( "name" ), "\nINPUT attributes: %s" % fail_msg )

... although my very newbie-ish understanding that this is not a brilliant test in that assertTrue, if ever it failed on one INPUT tag, would then cause the method to end... whereas you'd want info about all INPUTs which failed.

Answer 2

Yes that's a way to parse HTML into a DOM tree. If other people don't do that, they might have other requirements.

In general your idea is not bad, it might require more CPU time then other testing method (for example regular expressions. But if it fits your needs for testing, just do it. Performance it rarely a problem at testing time.