An exception will be raised if the passed name does not represent an object in the scene. This has several advantages:
- you will never unwittingly attempt to use a node or attribute that does not exist, either due to a typo or unexpected context
- it brings PyMEL’s attribute handling more in line with pythonic rules, where attributes must exist before accessing them
- it prevents the awkward situation of having a python object for which only a handful of methods will actually work
PyMEL has three exceptions which can be used to test for existence errors when creating new PyNodes: MayaObjectError, MayaNodeError, and MayaAttributeError.
>>> for x in [ 'fooBar.spangle', 'superMonk' ] :
... try:
... PyNode( x )
... print "It Exists"
... except MayaNodeError:
... print "The Node Doesn't Exist:", x
... except MayaAttributeError:
... print "The Attribute Doesn't Exist:", x
...
The Attribute Doesn't Exist: fooBar.spangle
The Node Doesn't Exist: superMonk
Both exceptions can be caught by using the parent exception MayaObjectError. In addition MayaAttributeError can also be caught with the builtin exception AttributeError.
Note that you will get different exceptions depending on how you access the attribute. This is because the shorthand notation can also be used to access functions, in which case the MayaAttributeError does not make sense to raise. As mentioned above, you can always use AttributeError to catch both.
>>> x = polySphere(name='earth')[0]
>>> x.attr('myAttr')
Traceback (most recent call last):
...
MayaAttributeError: Maya Attribute does not exist: u'earth.myAttr'
>>> x = polySphere(name='moon')[0]
>>> x.myAttr
Traceback (most recent call last):
...
AttributeError: Transform(u'moon') has no attribute or method named 'myAttr'
>>> if objExists( 'fooBar' ):
... print "It Exists"
... else:
... print "It Doesn't Exist"
It Doesn't Exist
>>> try:
... PyNode( 'fooBar' )
... print "It Exists"
... except MayaObjectError:
... print "It Doesn't Exist"
It Doesn't Exist
>>> if objExists( 'persp.spangle' ):
... print "Attribute Exists"
... else:
... print "Attribute Doesn't Exist"
Attribute Doesn't Exist
>>> x = PyNode('persp')
>>> if x.hasAttr('spangle'):
... print "Attribute Exists"
... else:
... print "Attribute Doesn't Exist"
Attribute Doesn't Exist
>>> try:
... PyNode( 'persp.spangle' )
... print "Attribute Exists"
... except MayaAttributeError:
... print "Attribute Doesn't Exist"
Attribute Doesn't Exist
>>> x = PyNode('persp')
>>> try:
... x.spangle
... print "Attribute Exists"
... except AttributeError:
... print "Attribute Doesn't Exist"
Attribute Doesn't Exist
Two other PyMEL idioms have been removed as a result of this change. Attribute.add has been removed because the attribute has to exist in order to successfully get an Attribute instance. Instead, you should use the addAttr method on the node:
>>> PyNode('persp').addAttr( 'myNewFloatAttr', at=float )
Similarly, the force flag for setAttr functions, which creates the attribute before setting if it does not exist, can only be safely used from the node class and not the attribute class:
Still supported:
>>> PyNode('persp').setAttr( 'myNewIntAttr', 2, force=1 )
New construct:
>>> PyNode('persp').setDynamicAttr( 'myNewIntAttr', 2 )
One advantage of the old way of dealing with non-existent objects was that you could use the name parsing methods of the PyNode classes to manipulate the object’s name until you found what you were looking for. To allow for this, we’ve added several classes which operate on non-existent nodes and contain only methods for string parsing and existence testing. These nodes can be found in the other module and are named other.NameParser, other.AttributeName, other.DependNodeName, and other.DagNodeName.
While PyNode serves to easily cast any string to its proper class in the node hierarchy, other nodes in the hierarchy can achieve the same effect:
>>> PyNode('lambert1')
nt.Lambert(u'lambert1')
>>> DependNode('lambert1')
nt.Lambert(u'lambert1')
If the determined type does not match the requested type, an error will be raised. For example, a lambert node is not a DAG node:
>>> nt.DagNode( 'lambert1' )
Traceback (most recent call last):
...
TypeError: Determined type is Lambert, which is not a subclass of desired type DagNode
This is useful because it can be used as a quick way to assert that a given node is of the desire type.
>>> select( 'lambert1' ) # this line represents user action
>>> try:
... nt.DagNode( selected()[0] )
... except TypeError:
... print "Please select a DAG node"
Please select a DAG node