I want to show a list of tags that is related to a particular tag(in an optimised way). I wonder why django-taggit does not provide an inbuilt functional开发者_如何学运维ity for this common task.
The solution I have to offer does a bit more than what you're asking for because it allows to find related tags for a set of given tags, not for only a single given tag. In reality, this is probably what you want to do though. I'm not sure if it's really optimal in term of performance since it uses a subquery, but it works and I find it easy to understand.
First, here is the test case:
from django.test import TestCase
from .models import Item, get_related_tags
class RelatedTagsTest(TestCase):
def setUp(self):
article1 = Item.objects.create(title='Python vs. COBOL')
article1.tags.add('programming', 'python', 'cobol')
article2 = Item.objects.create(title='Python vs. Boa Constrictor')
article2.tags.add('zoology', 'python', 'boa')
article3 = Item.objects.create(title='COBOL vs. FORTRAN')
article3.tags.add('cobol', 'fortran', 'programming')
def test_unique_tag(self):
self.assertEquals(get_related_tags('programming'),
['cobol', 'fortran', 'python'])
self.assertEquals(get_related_tags('python'),
['boa', 'cobol', 'programming', 'zoology'])
def test_multiple_tags(self):
self.assertEquals(get_related_tags('boa', 'fortran'),
['cobol', 'programming', 'python', 'zoology'])
As you can see, by "related tags" we mean the set of tags which are associated with items which are tagged with a set of given tags.
And here is our model with a function to get related tags:
from django.db import models
from taggit.managers import TaggableManager
from taggit.models import Tag
class Item(models.Model):
title = models.CharField(max_length=100)
tags = TaggableManager()
def get_related_tags(*tags):
# Get a QuerySet of related items
related_items = Item.objects.filter(tags__name__in=tags)
# Get tags for those related items (I found the name of the lookup field by
# reading taggit's source code)
qs = Tag.objects.filter(taggit_taggeditem_items__item__in=related_items)
# Exclude the tags we already have
qs = qs.exclude(name__in=tags)
# Order by name and remove duplicates
qs = qs.order_by('name').distinct()
# Return tag names to simplify test code, real code would probably return
# Tag objects
return [t.name for t in qs]
Note that you can easily add the number of items per tag using qs.annotate(count=Count('name'))
. It will be available as a count
attribute on each Tag
object.
精彩评论