zope.interface compliant ordered dictionary.
Project description
zodict
Ordered dictionary which implements the corresponding zope.interface.common.mapping interface.
>>> from zope.interface.common.mapping import IFullMapping >>> from zodict import zodict >>> zod = zodict() >>> IFullMapping.providedBy(zod) True
Node
This is a zodict which provides a location. Location the zope way means each item in the node-tree knows its parent and its own name.
>>> from zope.location.interface import ILocation >>> from zodict import Node >>> root = Node('root') >>> ILocation.providedBy(Node) True >>> root['child'] = Node() >>> root['child'].path ['root', 'child'] >>> child = root['child'] >>> child.__name__ 'child' >>> child.__parent__ <Node object 'root' at ...>
The filtereditems function.
>>> from zope.interface import Interface >>> from zope.interface import alsoProvides >>> class IMarker(Interface): pass >>> alsoProvides(root['child']['subchild'], IMarker) >>> IMarker.providedBy(root['child']['subchild']) True >>> for item in root['child'].filtereditems(IMarker): ... print item.path ['root', 'child', 'subchild']
UUID related operations on Node.
>>> uuid = root['child']['subchild'].uuid >>> uuid UUID('...') >>> root.node(uuid).path ['root', 'child', 'subchild'] >>> root.uuid = uuid Traceback (most recent call last): ... ValueError: Given uuid was already used for another Node >>> import uuid >>> newuuid = uuid.uuid4() >>> root.uuid = newuuid >>> root['child'].node(newuuid).path ['root']
Node insertion (an insertafter function exist as well).
>>> root['child1'] = Node() >>> root['child2'] = Node() >>> node = Node('child3') >>> root.insertbefore(node, root['child2']) >>> root.printtree() <class 'zodict.node.Node'>: root <class 'zodict.node.Node'>: child1 <class 'zodict.node.Node'>: child3 <class 'zodict.node.Node'>: child2
Move a node. Therefor we first need to detach the node we want to move from tree. Then insert the detached node elsewhere. In general, you can insert the detached node or subtree to a complete different tree.
>>> len(root._index.keys()) 6 >>> node = root.detach('child4') >>> node <Node object 'child4' at ...> >>> len(node._index.keys()) 1 >>> len(root._index.keys()) 5 >>> len(root.values()) 4 >>> root.insertbefore(node, root['child1']) >>> root.printtree() <class 'zodict.node.Node'>: root <class 'zodict.node.Node'>: child4 <class 'zodict.node.Node'>: child1 <class 'zodict.node.Node'>: child3 <class 'zodict.node.Node'>: child5 <class 'zodict.node.Node'>: child2
Merge 2 Node Trees.
>>> tree1 = Node() >>> tree1['a'] = Node() >>> tree1['b'] = Node() >>> tree2 = Node() >>> tree2['d'] = Node() >>> tree2['e'] = Node() >>> tree1._index is tree2._index False >>> len(tree1._index.keys()) 3 >>> tree1.printtree() <class 'zodict.node.Node'>: None <class 'zodict.node.Node'>: a <class 'zodict.node.Node'>: b >>> len(tree2._index.keys()) 3 >>> tree2.printtree() <class 'zodict.node.Node'>: None <class 'zodict.node.Node'>: d <class 'zodict.node.Node'>: e >>> tree1['c'] = tree2 >>> len(tree1._index.keys()) 6 >> sorted(tree1._index.values(), key=lambda x: x.__name__) >>> tree1._index is tree2._index True >>> tree1.printtree() <class 'zodict.node.Node'>: None <class 'zodict.node.Node'>: a <class 'zodict.node.Node'>: b <class 'zodict.node.Node'>: c <class 'zodict.node.Node'>: d <class 'zodict.node.Node'>: e
Changes
Version 1.7.0
- -Add Node.detach function. Needed for node or subtree moving. This is
done due to performance reasons. rnix, 2009-12-18
- -Node.index returns now a NodeIndex object, which implements
zope.interface.common.mapping.IReadMapping. This functions convert uuid instances to integers before node lookup. So we still fit the contract of returning nodes from index by uuid. rnix, 2009-12-18
- -Change type of keys of Node._index to int. uuid.UUID.__hash__
function was called too often jensens, rnix, 2009-12-18
- -make Node thread safe.
jensens, rnix, 2009-12-18
Version 1.6.1
- -make Node trees merge properly.
rnix, 2009-12-15
- -make getter and setter functions of uuid property private.
rnix, 2009-12-15
Version 1.6.0
- -remove the traverser module.
rnix, 2009-11-28
- -improve insertbefore and insertafter a little bit.
rnix, 2009-11-28
- -add index Attribute to Node. Allows access to the internal
_index attribute. rnix, 2009-11-28
- -remove @accept and @return decorators. Just overhead.
rnix, 2009-11-28
Version 1.5.0
- -add insertbefore and insertafter function to Node.
rnix, 2009-11-27
- -fix printtree if Node.__name__ is None.
rnix, 2009-11-20
- -add printtree debug helper function to Node.
rnix, 2009-11-09
- -define own Traverser interface and reduce dependencies.
rnix, 2009-10-28
- -removed import of tests from zodicts __init__. this caused import errors
if interlude wasnt installed. jensens, 2009-07-16
Version 1.4.0
- -Don’t allow classes as values of a Node. Attribute __name__
conflicts. jensens, 2009-05-06
- -repr(nodeobj) now returns the real classname and not fixed
<Node object this helps a lot while testing and using classes inheriting from Node! jensens, 2009-05-06
- -Make tests run with python setup.py test.
Removed superflous dependency on zope.testing. jensens, 2009-05-06
Version 1.3.3
-Fix ITraverser interface import including BBB.
Version 1.3.2
- -Add root property to Node.
thet, 2009-04-24
Version 1.3.1
- -Add __delitem__ function to Node.
rnix, 2009-04-16
Version 1.3
- -Add uuid Attribute and node function to Node.
rnix, 2009-03-23
Version 1.2
- -Add filtereditems function to Node.
rnix, 2009-03-22
Version 1.1
- -Add INode interface and implementation.
rnix, 2009-03-18
Credits
- -Written by Robert Niederreiter <rnix@squarewave.at>
2009-03-17
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.