1
0
mirror of https://github.com/django/django.git synced 2024-12-22 17:16:24 +00:00

Fixed #23646 -- Added QuerySet.bulk_update() to efficiently update many models.

This commit is contained in:
Tom Forbes 2018-09-18 21:14:44 +01:00 committed by Tim Graham
parent 7b159df942
commit 9cbdb44014
11 changed files with 359 additions and 8 deletions

View File

@ -265,6 +265,10 @@ class BaseDatabaseFeatures:
# INSERT?
supports_ignore_conflicts = True
# Does this backend require casting the results of CASE expressions used
# in UPDATE statements to ensure the expression has the correct type?
requires_casted_case_in_updates = False
def __init__(self, connection):
self.connection = connection

View File

@ -48,6 +48,7 @@ class DatabaseFeatures(BaseDatabaseFeatures):
V_I := P_I;
END;
$$ LANGUAGE plpgsql;"""
requires_casted_case_in_updates = True
supports_over_clause = True
supports_aggregate_filter_clause = True
supported_explain_formats = {'JSON', 'TEXT', 'XML', 'YAML'}

View File

@ -18,9 +18,9 @@ from django.db import (
from django.db.models import DateField, DateTimeField, sql
from django.db.models.constants import LOOKUP_SEP
from django.db.models.deletion import Collector
from django.db.models.expressions import F
from django.db.models.expressions import Case, Expression, F, Value, When
from django.db.models.fields import AutoField
from django.db.models.functions import Trunc
from django.db.models.functions import Cast, Trunc
from django.db.models.query_utils import FilteredRelation, InvalidQuery, Q
from django.db.models.sql.constants import CURSOR, GET_ITERATOR_CHUNK_SIZE
from django.db.utils import NotSupportedError
@ -473,6 +473,50 @@ class QuerySet:
return objs
def bulk_update(self, objs, fields, batch_size=None):
"""
Update the given fields in each of the given objects in the database.
"""
if batch_size is not None and batch_size < 0:
raise ValueError('Batch size must be a positive integer.')
if not fields:
raise ValueError('Field names must be given to bulk_update().')
objs = tuple(objs)
if not all(obj.pk for obj in objs):
raise ValueError('All bulk_update() objects must have a primary key set.')
fields = [self.model._meta.get_field(name) for name in fields]
if any(not f.concrete or f.many_to_many for f in fields):
raise ValueError('bulk_update() can only be used with concrete fields.')
if any(f.primary_key for f in fields):
raise ValueError('bulk_update() cannot be used with primary key fields.')
if not objs:
return
# PK is used twice in the resulting update query, once in the filter
# and once in the WHEN. Each field will also have one CAST.
max_batch_size = connections[self.db].ops.bulk_batch_size(['pk', 'pk'] + fields, objs)
batch_size = min(batch_size, max_batch_size) if batch_size else max_batch_size
requires_casting = connections[self.db].features.requires_casted_case_in_updates
batches = (objs[i:i + batch_size] for i in range(0, len(objs), batch_size))
updates = []
for batch_objs in batches:
update_kwargs = {}
for field in fields:
when_statements = []
for obj in batch_objs:
attr = getattr(obj, field.attname)
if not isinstance(attr, Expression):
attr = Value(attr, output_field=field)
when_statements.append(When(pk=obj.pk, then=attr))
case_statement = Case(*when_statements, output_field=field)
if requires_casting:
case_statement = Cast(case_statement, output_field=field)
update_kwargs[field.attname] = case_statement
updates.append(([obj.pk for obj in batch_objs], update_kwargs))
with transaction.atomic(using=self.db, savepoint=False):
for pks, update_kwargs in updates:
self.filter(pk__in=pks).update(**update_kwargs)
bulk_update.alters_data = True
def get_or_create(self, defaults=None, **kwargs):
"""
Look up an object with the given kwargs, creating one if necessary.

View File

@ -2089,6 +2089,42 @@ instance (if the database normally supports it).
The ``ignore_conflicts`` parameter was added.
``bulk_update()``
~~~~~~~~~~~~~~~~~
.. versionadded:: 2.2
.. method:: bulk_update(objs, fields, batch_size=None)
This method efficiently updates the given fields on the provided model
instances, generally with one query::
>>> objs = [
... Entry.objects.create(headline='Entry 1'),
... Entry.objects.create(headline='Entry 2'),
... ]
>>> objs[0].headline = 'This is entry 1'
>>> objs[1].headline = 'This is entry 2'
>>> Entry.objects.bulk_update(objs, ['headline'])
:meth:`.QuerySet.update` is used to save the changes, so this is more efficient
than iterating through the list of models and calling ``save()`` on each of
them, but it has a few caveats:
* You cannot update the model's primary key.
* Each model's ``save()`` method isn't called, and the
:attr:`~django.db.models.signals.pre_save` and
:attr:`~django.db.models.signals.post_save` signals aren't sent.
* If updating a large number of columns in a large number of rows, the SQL
generated can be very large. Avoid this by specifying a suitable
``batch_size``.
* Updating fields defined on multi-table inheritance ancestors will incur an
extra query per ancestor.
The ``batch_size`` parameter controls how many objects are saved in a single
query. The default is to create all objects in one batch, except for SQLite
and Oracle which have restrictions on the number of variables used in a query.
``count()``
~~~~~~~~~~~

View File

@ -199,6 +199,9 @@ Models
:class:`~django.db.models.DateTimeField`, and the new :lookup:`iso_year`
lookup allows querying by an ISO-8601 week-numbering year.
* The new :meth:`.QuerySet.bulk_update` method allows efficiently updating
specific fields on multiple model instances.
Requests and Responses
~~~~~~~~~~~~~~~~~~~~~~

View File

@ -532,6 +532,7 @@ class ManagerTest(SimpleTestCase):
'update_or_create',
'create',
'bulk_create',
'bulk_update',
'filter',
'aggregate',
'annotate',

View File

@ -56,9 +56,9 @@ class Migration(migrations.Migration):
name='OtherTypesArrayModel',
fields=[
('id', models.AutoField(verbose_name='ID', serialize=False, auto_created=True, primary_key=True)),
('ips', ArrayField(models.GenericIPAddressField(), size=None)),
('uuids', ArrayField(models.UUIDField(), size=None)),
('decimals', ArrayField(models.DecimalField(max_digits=5, decimal_places=2), size=None)),
('ips', ArrayField(models.GenericIPAddressField(), size=None, default=list)),
('uuids', ArrayField(models.UUIDField(), size=None, default=list)),
('decimals', ArrayField(models.DecimalField(max_digits=5, decimal_places=2), size=None, default=list)),
('tags', ArrayField(TagField(), blank=True, null=True, size=None)),
('json', ArrayField(JSONField(default={}), default=[])),
('int_ranges', ArrayField(IntegerRangeField(), null=True, blank=True)),

View File

@ -63,9 +63,9 @@ class NestedIntegerArrayModel(PostgreSQLModel):
class OtherTypesArrayModel(PostgreSQLModel):
ips = ArrayField(models.GenericIPAddressField())
uuids = ArrayField(models.UUIDField())
decimals = ArrayField(models.DecimalField(max_digits=5, decimal_places=2))
ips = ArrayField(models.GenericIPAddressField(), default=list)
uuids = ArrayField(models.UUIDField(), default=list)
decimals = ArrayField(models.DecimalField(max_digits=5, decimal_places=2), default=list)
tags = ArrayField(TagField(), blank=True, null=True)
json = ArrayField(JSONField(default=dict), default=list)
int_ranges = ArrayField(IntegerRangeField(), blank=True, null=True)

View File

@ -0,0 +1,34 @@
from datetime import date
from . import PostgreSQLTestCase
from .models import (
HStoreModel, IntegerArrayModel, JSONModel, NestedIntegerArrayModel,
NullableIntegerArrayModel, OtherTypesArrayModel, RangesModel,
)
try:
from psycopg2.extras import NumericRange, DateRange
except ImportError:
pass # psycopg2 isn't installed.
class BulkSaveTests(PostgreSQLTestCase):
def test_bulk_update(self):
test_data = [
(IntegerArrayModel, 'field', [], [1, 2, 3]),
(NullableIntegerArrayModel, 'field', [1, 2, 3], None),
(JSONModel, 'field', {'a': 'b'}, {'c': 'd'}),
(NestedIntegerArrayModel, 'field', [], [[1, 2, 3]]),
(HStoreModel, 'field', {}, {1: 2}),
(RangesModel, 'ints', None, NumericRange(lower=1, upper=10)),
(RangesModel, 'dates', None, DateRange(lower=date.today(), upper=date.today())),
(OtherTypesArrayModel, 'ips', [], ['1.2.3.4']),
(OtherTypesArrayModel, 'json', [], [{'a': 'b'}])
]
for Model, field, initial, new in test_data:
with self.subTest(model=Model, field=field):
instances = Model.objects.bulk_create(Model(**{field: initial}) for _ in range(20))
for instance in instances:
setattr(instance, field, new)
Model.objects.bulk_update(instances, [field])
self.assertSequenceEqual(Model.objects.filter(**{field: new}), instances)

View File

@ -718,3 +718,8 @@ class RelatedIndividual(models.Model):
class Meta:
db_table = 'RelatedIndividual'
class CustomDbColumn(models.Model):
custom_column = models.IntegerField(db_column='custom_name', null=True)
ip_address = models.GenericIPAddressField(null=True)

View File

@ -0,0 +1,223 @@
import datetime
from django.core.exceptions import FieldDoesNotExist
from django.db.models import F
from django.db.models.functions import Lower
from django.test import TestCase
from .models import (
Article, CustomDbColumn, CustomPk, Detail, Individual, Member, Note,
Number, Paragraph, SpecialCategory, Tag, Valid,
)
class BulkUpdateNoteTests(TestCase):
def setUp(self):
self.notes = [
Note.objects.create(note=str(i), misc=str(i))
for i in range(10)
]
def create_tags(self):
self.tags = [
Tag.objects.create(name=str(i))
for i in range(10)
]
def test_simple(self):
for note in self.notes:
note.note = 'test-%s' % note.id
with self.assertNumQueries(1):
Note.objects.bulk_update(self.notes, ['note'])
self.assertCountEqual(
Note.objects.values_list('note', flat=True),
[cat.note for cat in self.notes]
)
def test_multiple_fields(self):
for note in self.notes:
note.note = 'test-%s' % note.id
note.misc = 'misc-%s' % note.id
with self.assertNumQueries(1):
Note.objects.bulk_update(self.notes, ['note', 'misc'])
self.assertCountEqual(
Note.objects.values_list('note', flat=True),
[cat.note for cat in self.notes]
)
self.assertCountEqual(
Note.objects.values_list('misc', flat=True),
[cat.misc for cat in self.notes]
)
def test_batch_size(self):
with self.assertNumQueries(len(self.notes)):
Note.objects.bulk_update(self.notes, fields=['note'], batch_size=1)
def test_unsaved_models(self):
objs = self.notes + [Note(note='test', misc='test')]
msg = 'All bulk_update() objects must have a primary key set.'
with self.assertRaisesMessage(ValueError, msg):
Note.objects.bulk_update(objs, fields=['note'])
def test_foreign_keys_do_not_lookup(self):
self.create_tags()
for note, tag in zip(self.notes, self.tags):
note.tag = tag
with self.assertNumQueries(1):
Note.objects.bulk_update(self.notes, ['tag'])
self.assertSequenceEqual(Note.objects.filter(tag__isnull=False), self.notes)
def test_set_field_to_null(self):
self.create_tags()
Note.objects.update(tag=self.tags[0])
for note in self.notes:
note.tag = None
Note.objects.bulk_update(self.notes, ['tag'])
self.assertCountEqual(Note.objects.filter(tag__isnull=True), self.notes)
def test_set_mixed_fields_to_null(self):
self.create_tags()
midpoint = len(self.notes) // 2
top, bottom = self.notes[:midpoint], self.notes[midpoint:]
for note in top:
note.tag = None
for note in bottom:
note.tag = self.tags[0]
Note.objects.bulk_update(self.notes, ['tag'])
self.assertCountEqual(Note.objects.filter(tag__isnull=True), top)
self.assertCountEqual(Note.objects.filter(tag__isnull=False), bottom)
def test_functions(self):
Note.objects.update(note='TEST')
for note in self.notes:
note.note = Lower('note')
Note.objects.bulk_update(self.notes, ['note'])
self.assertEqual(set(Note.objects.values_list('note', flat=True)), {'test'})
# Tests that use self.notes go here, otherwise put them in another class.
class BulkUpdateTests(TestCase):
def test_no_fields(self):
msg = 'Field names must be given to bulk_update().'
with self.assertRaisesMessage(ValueError, msg):
Note.objects.bulk_update([], fields=[])
def test_invalid_batch_size(self):
msg = 'Batch size must be a positive integer.'
with self.assertRaisesMessage(ValueError, msg):
Note.objects.bulk_update([], fields=['note'], batch_size=-1)
def test_nonexistent_field(self):
with self.assertRaisesMessage(FieldDoesNotExist, "Note has no field named 'nonexistent'"):
Note.objects.bulk_update([], ['nonexistent'])
pk_fields_error = 'bulk_update() cannot be used with primary key fields.'
def test_update_primary_key(self):
with self.assertRaisesMessage(ValueError, self.pk_fields_error):
Note.objects.bulk_update([], ['id'])
def test_update_custom_primary_key(self):
with self.assertRaisesMessage(ValueError, self.pk_fields_error):
CustomPk.objects.bulk_update([], ['name'])
def test_empty_objects(self):
with self.assertNumQueries(0):
Note.objects.bulk_update([], ['note'])
def test_large_batch(self):
Note.objects.bulk_create([
Note(note=str(i), misc=str(i))
for i in range(0, 2000)
])
notes = list(Note.objects.all())
Note.objects.bulk_update(notes, ['note'])
def test_only_concrete_fields_allowed(self):
obj = Valid.objects.create(valid='test')
detail = Detail.objects.create(data='test')
paragraph = Paragraph.objects.create(text='test')
Member.objects.create(name='test', details=detail)
msg = 'bulk_update() can only be used with concrete fields.'
with self.assertRaisesMessage(ValueError, msg):
Detail.objects.bulk_update([detail], fields=['member'])
with self.assertRaisesMessage(ValueError, msg):
Paragraph.objects.bulk_update([paragraph], fields=['page'])
with self.assertRaisesMessage(ValueError, msg):
Valid.objects.bulk_update([obj], fields=['parent'])
def test_custom_db_columns(self):
model = CustomDbColumn.objects.create(custom_column=1)
model.custom_column = 2
CustomDbColumn.objects.bulk_update([model], fields=['custom_column'])
model.refresh_from_db()
self.assertEqual(model.custom_column, 2)
def test_custom_pk(self):
custom_pks = [
CustomPk.objects.create(name='pk-%s' % i, extra='')
for i in range(10)
]
for model in custom_pks:
model.extra = 'extra-%s' % model.pk
CustomPk.objects.bulk_update(custom_pks, ['extra'])
self.assertCountEqual(
CustomPk.objects.values_list('extra', flat=True),
[cat.extra for cat in custom_pks]
)
def test_inherited_fields(self):
special_categories = [
SpecialCategory.objects.create(name=str(i), special_name=str(i))
for i in range(10)
]
for category in special_categories:
category.name = 'test-%s' % category.id
category.special_name = 'special-test-%s' % category.special_name
SpecialCategory.objects.bulk_update(special_categories, ['name', 'special_name'])
self.assertCountEqual(
SpecialCategory.objects.values_list('name', flat=True),
[cat.name for cat in special_categories]
)
self.assertCountEqual(
SpecialCategory.objects.values_list('special_name', flat=True),
[cat.special_name for cat in special_categories]
)
def test_field_references(self):
numbers = [Number.objects.create(num=0) for _ in range(10)]
for number in numbers:
number.num = F('num') + 1
Number.objects.bulk_update(numbers, ['num'])
self.assertCountEqual(Number.objects.filter(num=1), numbers)
def test_booleanfield(self):
individuals = [Individual.objects.create(alive=False) for _ in range(10)]
for individual in individuals:
individual.alive = True
Individual.objects.bulk_update(individuals, ['alive'])
self.assertCountEqual(Individual.objects.filter(alive=True), individuals)
def test_ipaddressfield(self):
for ip in ('2001::1', '1.2.3.4'):
with self.subTest(ip=ip):
models = [
CustomDbColumn.objects.create(ip_address='0.0.0.0')
for _ in range(10)
]
for model in models:
model.ip_address = ip
CustomDbColumn.objects.bulk_update(models, ['ip_address'])
self.assertCountEqual(CustomDbColumn.objects.filter(ip_address=ip), models)
def test_datetime_field(self):
articles = [
Article.objects.create(name=str(i), created=datetime.datetime.today())
for i in range(10)
]
point_in_time = datetime.datetime(1991, 10, 31)
for article in articles:
article.created = point_in_time
Article.objects.bulk_update(articles, ['created'])
self.assertCountEqual(Article.objects.filter(created=point_in_time), articles)