Reputation: 1913
I have following models:
class TestCaseStatus(models.Model):
name = models.CharField(default='n/a', max_length=50)
def __str__(self):
return self.name
class TestCase(models.Model):
id = models.CharField(unique=True, max_length=100, primary_key=True)
name = models.CharField(default='n/a', max_length=200)
def __str__(self):
return self.name
class TestRun(models.Model):
test_id = models.ForeignKey(TestCase)
run_id = models.CharField(max_length=100, default='0')
datetime = models.DateTimeField()
status = models.ForeignKey(TestCaseStatus)
messages = models.CharField(default='n/a', max_length=500)
def __str__(self):
return self.run_id, self.test_id, self.status
I want to execute following SQL query:
select a.test_id_id, a.status_id, b.status_id
from runscompare_testrun as a
left join runscompare_testrun as b on a.test_id_id = b.test_id_id
where a.run_id = 1 and b.run_id=2
SQL output:
I assume that there is a django way to do it to avoid using raw() call, but I can't seems to find it.
Upvotes: 1
Views: 84
Reputation: 2651
If I'm understanding your query correctly, you want all the details where run_id="1"
, with additional details for any runs where the TestCase
is the same, but run_id="2"
.
With that assumption, you could do something like this:
# Start with run_id=1
a = TestRun.objects.filter(run_id="1").select_related(
'test_id', 'status').prefetch_related(
'test_id__testrun_set', 'test_id__testrun_set__status')
# For each, get related
for i in a:
b = i.test_id.testrun_set.filter(run_id="2")
print(i.test_id, i.status, [j.status for j in b])
This won't be a single database query, as both select_related
(docs) and prefetch_related
(docs) will make a database call, but it will reduce the number of calls made.
Upvotes: 1