Minidom.Node doesn't call __getattr__

[ArseAssassin]

I'm using minidom to parse XML and to simplify accessing child nodes, I'm trying to implement __getattr__ for the Node class. Yes, I know what most of you will think of that; I'll just have to be careful, won't I. ;)

import xml
from xml.dom.minidom import parseString

def getChild(self, name):
    print "Getting child"
    return self.getElementsByTagName(name)[0]

xml.dom.minidom.Node.__getattr__ = getChild

xmlDoc = parseString('''<?xml version="1.0" encoding="ISO-8859-1"?>
<CATALOG>
  <cd>
    <title>Empire Burlesque</title>
    <ARTIST>Bob Dylan</ARTIST>
    <COUNTRY>USA</COUNTRY>
    <COMPANY>Columbia</COMPANY>
    <PRICE>10.90</PRICE>
    <YEAR>1985</YEAR>
  </cd>
  <cd>
    <title>Hide your heart</title>
    <ARTIST>Bonnie Tylor</ARTIST>
    <COUNTRY>UK</COUNTRY>
    <COMPANY>CBS Records</COMPANY>
    <PRICE>9.90</PRICE>
    <YEAR>1988</YEAR>
  </cd>
</CATALOG>''')

xmlRoot = xmlDoc.childNodes[0]
print xmlRoot.__getattr__
print xmlRoot.cd.title.childNodes[0].nodeValue

outputs:

<bound method Element.getChild of <DOM Element: CATALOG at 0x1c50850>>
  File "C:\Users\Tuomas\Projektit\wrestlevania\ezXML.py", line 31, in <module>
    print xmlRoot.cd.title.childNodes[0].nodeValue
AttributeError: Element instance has no attribute 'cd'

As far as I can tell, everything works fine, except __getattr__ never gets called. Can anyone help me understand why that is or how to work around it?

[jlm699]

If you're trying to call __getattr__ you need ().

[ArseAssassin]

If you're trying to call __getattr__ you need ().

Not trying to call it - I'm overriding it.

[jlm699]

Not trying to call it - I'm overriding it.

No I know that... but you still need the (name) arguments that getChild needs

[ArseAssassin]

No I know that... but you still need the (name) arguments that getChild needs

I'm not sure what you're getting at. getChild is just an implementation of the special method __getattr__, which should be called when I'm trying to access attributes that aren't defined.

[jlm699]

I'm not sure what you're getting at. getChild is just an implementation of the special method __getattr__, which should be called when I'm trying to access attributes that aren't defined.

But aren't you over-riding __getattr__ with getChild ? The way that your code is setup, instead of __getattr__ being called, it calls for getChild... I don't know, maybe I'm misunderstandin g what you're trying to accomplish in this code... Where does the .cd. method come from since it's not an attribute of the class that you are trying to call it from?

[bvdet]

I'm using minidom to parse XML and to simplify accessing child nodes, I'm trying to implement __getattr__ for the Node class. Yes, I know what most of you will think of that; I'll just have to be careful, won't I. ;)

import xml
from xml.dom.minidom import parseString

def getChild(self, name):
    print "Getting child"
    return self.getElementsByTagName(name)[0]

xml.dom.minidom.Node.__getattr__ = getChild

xmlDoc = parseString('''<?xml version="1.0" encoding="ISO-8859-1"?>
<CATALOG>
  <cd>
    <title>Empire Burlesque</title>
    <ARTIST>Bob Dylan</ARTIST>
    <COUNTRY>USA</COUNTRY>
    <COMPANY>Columbia</COMPANY>
    <PRICE>10.90</PRICE>
    <YEAR>1985</YEAR>
  </cd>
  <cd>
    <title>Hide your heart</title>
    <ARTIST>Bonnie Tylor</ARTIST>
    <COUNTRY>UK</COUNTRY>
    <COMPANY>CBS Records</COMPANY>
    <PRICE>9.90</PRICE>
    <YEAR>1988</YEAR>
  </cd>
</CATALOG>''')

xmlRoot = xmlDoc.childNodes[0]
print xmlRoot.__getattr__
print xmlRoot.cd.title.childNodes[0].nodeValue

outputs:

<bound method Element.getChild of <DOM Element: CATALOG at 0x1c50850>>
  File "C:\Users\Tuomas\Projektit\wrestlevania\ezXML.py", line 31, in <module>
    print xmlRoot.cd.title.childNodes[0].nodeValue
AttributeError: Element instance has no attribute 'cd'

As far as I can tell, everything works fine, except __getattr__ never gets called. Can anyone help me understand why that is or how to work around it?

Modify getChild() slightly:[code=Python]def getChild(self, name):
print "Getting child"
return self.getElement sByTagName(name )[/code]Modify the calls:[code=Python]xmlRoot = xmlDoc.childNod es[0]
print xmlRoot.__getat tr__('cd')
print xmlRoot.__getat tr__('cd')[0].__getattr__('t itle')[0].firstChild.nod eValue
print xmlRoot.__getat tr__('ARTIST')[0].firstChild.nod eValue[/code]Output:[code=Python]>>> Getting child
[<DOM Element: cd at 0x100d4b8>, <DOM Element: cd at 0x100d788>]
Getting child
Getting child
Empire Burlesque
Getting child
Bob Dylan
>>> [/code]

[jlm699]

Modify the calls:[code=Python]
print xmlRoot.__getat tr__('cd')
[/code]

That's what I was trying to say by using the (name) argument

[bvdet]

That's what I was trying to say by using the (name) argument

And you were correct!

[jlm699]

[code=Python]xmlRoot = xmlDoc.childNod es[0]
print xmlRoot.__getat tr__('cd')
print xmlRoot.__getat tr__('cd')[0].__getattr__('t itle')[0].firstChild.nod eValue
print xmlRoot.__getat tr__('ARTIST')[0].firstChild.nod eValue[/code]Output:[code=Python]>>> Getting child
[<DOM Element: cd at 0x100d4b8>, <DOM Element: cd at 0x100d788>]
Getting child
Getting child
Empire Burlesque
Getting child
Bob Dylan
>>> [/code]

In my calls I needed to modify it a little bit...
[CODE=python]
>>> xmlRoot.__getat tr__('cd').__ge tattr__('title' ).firstChild.no deValue
Getting child
Getting child
u'Empire Burlesque'
>>> xmlRoot.__getat tr__('ARTIST'). firstChild.node Value
Getting child
u'Bob Dylan'
>>>
[/CODE]

[bvdet]

In my calls I needed to modify it a little bit...
[CODE=python]
>>> xmlRoot.__getat tr__('cd').__ge tattr__('title' ).firstChild.no deValue
Getting child
Getting child
u'Empire Burlesque'
>>> xmlRoot.__getat tr__('ARTIST'). firstChild.node Value
Getting child
u'Bob Dylan'
>>>
[/CODE]

I had modified function getChild() to return the NodeList object. Otherwise you could not access the other elements in the object.
[code=Python]
>>> for elem in xmlRoot.__getat tr__('cd'):
... print elem
...
Getting child
<DOM Element: cd at 0xf854b8>
<DOM Element: cd at 0xf8b490>
>>> [/code]

[ArseAssassin]

In my calls I needed to modify it a little bit...
[CODE=python]
>>> xmlRoot.__getat tr__('cd').__ge tattr__('title' ).firstChild.no deValue
Getting child
Getting child
u'Empire Burlesque'
>>> xmlRoot.__getat tr__('ARTIST'). firstChild.node Value
Getting child
u'Bob Dylan'
>>>
[/CODE]

Well, obviously that'll work. Here's what I'm going for:

http://pyref.infogami.com/__getattr__

Called when an attribute lookup has not found the attribute in the usual places (i.e. it is not an instance attribute nor is it found in the class tree for self).

With any custom class I define, it will work:

>>> class example:
... 	def __init__(self):
... 		print 'Initiated'
... 		
>>> a = example()
Initiated
>>> def getAttribute(self, name):
... 	print name, 'was not found'
... 	
>>> example.__getattr__ = getAttribute
>>> print a.unknownAttribute
unknownAttribute was not found

I'm having trouble understanding why it won't work for minidom nodes.

[ArseAssassin]

I was finally able to figure it out: since the nodes I'm accessing are actually instances of the Element class and __getattr__ isn't part of its parent's (Node class) original class definition, the new instances aren't, I suppose, aware of the method's existence. It's definitely inherited since I can access it; the instances just won't know they're supposed to call it. That's what I figure anyway. Here's the working code:

import xml
from xml.dom.minidom import parseString

def getChild(self, name):
    print "Getting child"
    return self.getElementsByTagName(name)[0]

xml.dom.minidom.Element.__getattr__ = getChild

xmlDoc = parseString('''<?xml version="1.0" encoding="ISO-8859-1"?>
<CATALOG>
  <cd>
    <title>Empire Burlesque</title>
    <ARTIST>Bob Dylan</ARTIST>
    <COUNTRY>USA</COUNTRY>
    <COMPANY>Columbia</COMPANY>
    <PRICE>10.90</PRICE>
    <YEAR>1985</YEAR>
  </cd>
  <cd>
    <title>Hide your heart</title>
    <ARTIST>Bonnie Tylor</ARTIST>
    <COUNTRY>UK</COUNTRY>
    <COMPANY>CBS Records</COMPANY>
    <PRICE>9.90</PRICE>
    <YEAR>1988</YEAR>
  </cd>
</CATALOG>''')

xmlRoot = xmlDoc.childNodes[0]
print xmlRoot.__getattr__
print xmlRoot.cd.title.childNodes[0].nodeValue

Thanks for the help, guys. :)

[bvdet]

I was finally able to figure it out: since the nodes I'm accessing are actually instances of the Element class and __getattr__ isn't part of its parent's (Node class) original class definition, the new instances aren't, I suppose, aware of the method's existence. It's definitely inherited since I can access it; the instances just won't know they're supposed to call it. That's what I figure anyway. Here's the working code:

import xml
from xml.dom.minidom import parseString

def getChild(self, name):
    print "Getting child"
    return self.getElementsByTagName(name)[0]

xml.dom.minidom.Element.__getattr__ = getChild

xmlDoc = parseString('''<?xml version="1.0" encoding="ISO-8859-1"?>
<CATALOG>
  <cd>
    <title>Empire Burlesque</title>
    <ARTIST>Bob Dylan</ARTIST>
    <COUNTRY>USA</COUNTRY>
    <COMPANY>Columbia</COMPANY>
    <PRICE>10.90</PRICE>
    <YEAR>1985</YEAR>
  </cd>
  <cd>
    <title>Hide your heart</title>
    <ARTIST>Bonnie Tylor</ARTIST>
    <COUNTRY>UK</COUNTRY>
    <COMPANY>CBS Records</COMPANY>
    <PRICE>9.90</PRICE>
    <YEAR>1988</YEAR>
  </cd>
</CATALOG>''')

xmlRoot = xmlDoc.childNodes[0]
print xmlRoot.__getattr__
print xmlRoot.cd.title.childNodes[0].nodeValue

Thanks for the help, guys. :)

You are welcome, and thank you for sharing what you have found out.