(UNDER DEVELOPMENT)
Hierarchical db model (Zope inspired) implementation for Django.
For example, let's build a naive retail model hierarchy example:
Store
|
Department
/ \
PriceUtil Section
/ \
PriceUtil RetailItem
We do this by specifying parents and children constraints. (Assuming app label is 'mystore'.)
from ztree import TreeContent
from django.db import models
class Store(TreeContent):
PARENTS_CONSTRAINT = ({'content_type_name': '<root>'},)
CHILDREN_CONSTRAINT = ({'content_type_name': 'mystore.department',)
name = models.TextField()
class Department(TreeContent):
PARENTS_CONSTRAINT = ({'content_type_name': 'mystore.store'},)
# quantifier indicating max num of objects allowed
CHILDREN_CONSTRAINT = ({'content_type_name': 'mystore.priceutil', 'quantifier': 1},
{'content_type_name': 'mystore.section'} )
name = models.TextField()
class Section(TreeContent):
PARENTS_CONSTRAINT = ({'content_type_name': 'mystore.department'},)
CHILDREN_CONSTRAINT = ({'content_type_name': 'mystore.priceutil', 'quantifier': 1},
{'content_type_name': 'mystore.retailitem'} )
code = models.CharField(max_length=50)
class PriceUtil(TreeContent):
PARENTS_CONSTRAINT = ({'content_type_name': 'mystore.department'},
{'content_type_name': 'mystore.section'})
discount_percent = models.IntegerField()
class RetailItem(TreeContent):
PARENTS_CONSTRAINT = ({'content_type_name': 'mystore.section'},)
name = models.TextField()
cost = models.FloatField()
price = models.FloatField()
The following methods are available for tree traversal (search).
get_node - retrieves tree node given the context path
filter_children - filters children of a given node, if no filter specified all children returned
lookup - searches up tree branch (towards root) starting from context node returning the first node
satisfying the filter
lookup_all - same as lookup but returns all nodes found
filter_descendents - descendents search from a given context node
count_children - count children of a context node satisfying the filter
For example, if we had a node:
/superstore/mensdept/shoes/hushpuppies123
So, the RetailItem 'hushpuppies' are in the 'shoes' Section in 'mensdept' Department of a 'superstore' Store.
Assume we also created a PriceUtil at the 'shoes' Section.
/superstore/mensdept/shoes/shoespriceutil
Now if we do a PriceUtil lookup (content type name 'mystore.priceutil') from the 'hushpuppies123' (context). (For example, we are displaying the shoes detail view and need applicable price info.)
from ztree.query.manager import TreeQueryManager
tqm = TreeQueryManager()
price_util = tqm.lookup(context_path='/superstore/mensdept/shoes/hushpuppies123', ct='mystore.priceutil')
We will get the '/superstore/mensdept/shoes/shoespriceutil' utility.
This could be used to override objects (pricing configuration in this case) in more local parts of the site tree. For example we could set up some generic PricUtil at the department level and override it by creating one at the 'shoes' section level. For example, department discount is 10%, but only for short time there is a 20% discount at Mens Shoes section. Lookup always retrieving more local object.
ztree extends on Django's auth app by creating a concept of LocalUser whose permissions are local to the tree branch the user is created in.
For example if we created a user 'joeb' in Department 'mensdept' with some department manager permissions. The user permissions would only apply in the branch with the root of 'mensdept', but not outside this branch (in other parts of the 'superstore' or other departments).
It is possible to set up ztree as a back-end web service allowing for easier caching, better performance and separation from the front-end.