python - Raw queries and custom sql queries what is the fastest? -

assuming file models.py in django application (webapp) following :

from django.db import models django.db import connection  class foo(models.model):     name = models.charfield(...)     surname = models.charfield(...)  def dictfetchall(cursor):     "returns rows cursor dict"     desc = cursor.description     return [         dict(zip([col[0] col in desc], row))         row in cursor.fetchall()     ]  def get_foo():     cursor = connection.cursor()     cursor.execute('select * foo_table')     rows = dictfetchall(cursor)     return rows 

to access database content, have basicly 2 options :

option 1 :

from webapp.models import foo bar = foo.objects.raw('select * foo_table') 

option 2 :

from application.models import get_foo bar = get_foo() 

which option fastest in execution ? there better way want ?

there no direct , clear answer on approach better.

using manager.raw() still keeps within orm layer , while returns model instances still have nice database abstraction. but, while making raw query, django more cursor.execute in order translate results model instances (see happening in rawqueryset , rawquery classes).

but (quote docs):

sometimes manager.raw() isn’t quite enough: might need perform queries don’t map cleanly models, or directly execute update, insert, or delete queries.

so, speaking, choose depends on results going , going them.

see also:

• performing raw sql queries
• executing-custom-sql-directly
• raw sql queries in django views