On my Django site, I am saving a user's reactions so when a user clicks a button, I store it as a created time and when the user clicks it second time, I stored the time as a finish time and so forth. Here it is my model;
class UserStatus(models.Model):
STATUS_TYPES = (
('online', 'online'),
('offline', 'offline')
)
user = models.ForeignKey(User)
status_type = models.CharField(max_length=30, choices=STATUS_TYPES, default='online')
created_time = models.DateTimeField(auto_now_add=True)
finish_time = models.DateTimeField(blank=True, null=True)
time_diff = models.DateTimeField(blank=True, null=True)
I added time_diff to show the time difference between created_time and finish time. When I try an example in the shell, I use;
user_status.created_time
datetime.datetime(2016, 3, 31, 12, 50, 21, tzinfo=<UTC>)
user_status.finish_time
datetime.datetime(2016, 3, 31, 12, 51, 37, 998593, tzinfo=<UTC>)
user_status.finish_time - user_status.created_time
datetime.timedelta(0, 76, 998593)
Everything seems to be fine until now, however when I wrote user_status.save() it gave an error;
line 93, in parse_datetime
match = datetime_re.match(value)
TypeError: expected string or buffer
I did not understand why it gave such an error.
Thank you in advance.
Now you try to use DateTimeField, but this field can only be used for storing date and time (but not time difference). You should use DurationField for storing timedelta.
Related
I have a model which contains date range i want to filter the data based on the range date
that is i want the data who's date range is 90 days from today's date.
class MyModel(models.Model):
name = models.CharField(max_length=255)
start_end_date = ranges.DateTimeRangeField(validators=
[validate_range_date_time])
so when we select the start date on page the end date will popoulate the same date but i cannot concatenate filter just by today date + timedelta(days=90) this is one single date and the field is date range, so how can i filter the date range data which is 90 days from now.
the model stores start_end_date as
'start_end_date': DateTimeTZRange(datetime.datetime(2022, 11, 29, 9, 15), datetime.datetime(2022, 11, 29, 10, 0),
Mymodel.objects.filter(start_end_date__contains=timezone.now() + timezone.timedelta(days=90))
timezone.now() + timezone.timedelta(days=90) = datetime.datetime(2022, 11, 29, 22, 52, 7, 759648)
the query is giving empty set
I think you could design the model more easily.
class MyModel(models.Model):
name = models.CharField(max_length=255)
start_date = models.DateTimeField()
end_date = models.DateTimeField()
Then you can find objects like the following.
target_time = timezone.now() + timezone.timedelta(days=90)
MyModel.objects.filter(start_date__lte = target_time).filter(end_date__gte = target_time)
As it's a DateTimeRangeField, I think your result can be achieved by using startswith and endswith just like that:
max_date = timezone.now() + timezone.timedelta(days=90)
MyModel.objects.filter(start_end_date__startswith__gte=timezone.now(), start_end_date__endswith__lte=max_date)
Hope it helps!
I haven't used this field myself, but in base of what i read from documentaition, it should be like this:
from psycopg2.extras import DateTimeTZRange
Mymodel.objects.filter(
start_end_date__contained_by=DateTimeTZRange(
timezone.now(),
timezone.now() + timezone.timedelta(days=90)
)
)
to check if any start_end_date field is in 90 days from now, you should also pass a datetime range.
edited:
from psycopg2.extras import DateTimeTZRange
Mymodel.objects.filter(
start_end_date__contained_by=DateTimeTZRange(
timezone.now(),
timezone.now() + timezone.timedelta(days=90),
start_end_date__lower_inc=False
)
)
I need to make an operation only available in 30 seconds, and then it expires. I Have this model:
(operation)
status = CharField(max_length=10, default="INI")
created_at = DateTimeField(auto_now_add=True)
account = ForeignKey(Account, null=False, on_delete=CASCADE)
metadata = models.JSONField(default=dict)
external_ref = CharField(max_length=128)
expires_at = DateTimeField(default=timezone.now()+timezone.timedelta(seconds=30))
I wanna be able to create in the expires_at field, a timestamp with exactly 30 seconds from the created_at Date, it is like a timeout function, but when I run the test:
def test_timeout_is_30_seconds(self):
print(self.operation.created_at)
timer = self.operation.created_at + timezone.timedelta(seconds=30)
print(timer)
self.assertEqual(self.operation.expires_at, timer)
it fails with this message:
AssertionError: datetime.datetime(2021, 6, 22, 19, 0, 42, 537490, tzinfo=<UTC>) != datetime.datetime(2021, 6, 22, 19, 0, 45, 844588, tzinfo=<UTC>)
I dont know if I need to make an external function or method inside the class or directly in the View, but I would prefer this default behavior in the models so I dont need to worry about setting expiry dates
I would be very grateful if you could help me solve it! :D any tips and information is appreciated
This is a common error. Your expression timezone.now()+timezone.timedelta(seconds=30) is being evaluated once, when the class is defined, and that value is being used as the default for every instance.
What you actually want is for the expiration time to be freshly calculated each time a new instance is created. Which means that you want to set default to a function:
def half_minute_hence():
return timezone.now() + timezone.timedelta(seconds=30)
expires_at = DateTimeField(default=half_minute_hence)
My django project is correctly enable the timezone in settings.
However, the datetime field of Django ORM object is a naive datetime object as shown in Block 3.
The expected result should be same as the output of Block 4 without any manually conversion.
In [1]: from django.conf import settings
...: settings.USE_TZ, settings.TIME_ZONE
Out[1]: (True, 'Asia/Hong_Kong')
In [2]: from qms.models import Quota
In [3]: q = Quota.objects.get(pk=1)
...: q.create_date, q.write_date
Out[3]:
(datetime.datetime(2021, 3, 10, 17, 37, 42, 489818),
datetime.datetime(2021, 3, 10, 17, 37, 42, 489818))
In [4]: from django.utils import timezone
...: timezone.make_aware(q.create_date,timezone.utc), \
...: timezone.make_aware(q.write_date, timezone.utc)
Out[4]:
(datetime.datetime(2021, 3, 10, 17, 37, 42, 489818, tzinfo=<UTC>),
datetime.datetime(2021, 3, 10, 17, 37, 42, 489818, tzinfo=<UTC>))
Record in SQL
Column
value
id
1
create_date
2021-03-10 17:37:42.489818+00
write_date
2021-03-10 17:37:42.489818+00
name
email
Django Model Definition
class Quota(models.models):
name = models.CharField(max_length=255)
create_date = models.DateTimeField(auto_now_add=True)
write_date = models.DateTimeField(auto_now=True)
The PostgreSQL database schema and settings, Table "public.qms_quota"
Column
Type
Modifiers
id
integer
not null default nextval('qms_quota_id_seq'::regclass)
create_date
timestamp with time zone
not null
write_date
timestamp with time zone
not null
name
character varying(255)
not null
SHOW TIMEZONE;
TimeZone
----------
UTC
Questions
How can I get the timezone-aware datetime object directly without any conversion?
Or the manual conversion is expected ?
The root-cause is a bug from a connection pool library django_postgrespool2==2.0.1.
When you use the your connection engine with "django_postgrespool2", it will NOT correctly handle the timezone settings. Releated Issue
TLDR: use engine django.db.backends.postgresql
You can use django.utils.timezone.localtime to convert the datetime received from the DB to localtime:
from django.utils.timezone import localtime
q = Quota.objects.get(pk=1)
print(localtime(q.create_date), localtime(q.write_date))
I'm actually using python 3.7 and Django 3.0.4.
I using in my app models.py like this for a simple system of messages.
from torii.models import User
class Room(models.Model):
users = models.ManyToManyField(User,
related_name='rooms',
blank=True)
name = models.CharField(max_length=100)
class Message(models.Model):
room = models.ForeignKey(Room,
on_delete=models.CASCADE,
related_name='messages')
user = models.ForeignKey(User,
on_delete=models.CASCADE)
content = models.TextField()
date = models.DateTimeField(auto_now_add=True)
So when my user send a message I create a new Message object attached to my Room.
I want to query all Room for a given User and annotate the date of the most recent message and the last message in my query.
No problem to get the most recent date in my related messages using Max like this:
for room in Room.objects.filter(users=user).annotate(last_message_date=Max('messages__date')).order_by('-last_message_date'):
print(room.__dict__)
{'id': 7, 'name': 'room-7', 'last_message_date': datetime.datetime(2020, 3, 20, 14, 0, 2, 118190, tzinfo=<UTC>)}
{'id': 9, 'name': 'room-9', 'last_message_date': datetime.datetime(2020, 3, 8, 15, 19, 52, 343780, tzinfo=<UTC>)}
{'id': 8, 'name': 'room-8', 'last_message_date': datetime.datetime(2020, 3, 7, 17, 18, 32, 648093, tzinfo=<UTC>)}
But I don't find any way to simply annotate the content of the last message. I have tried Max('messages__content') but in fact, it's order message by alphabetic order and it's always the same who is returned... I tried several subqueries with F and Q but that it didn’t work very well.
How can I annotate the result of room.messages.last().content directly with my query?
Finally, I solve this problem using OuterRef and Subquery documented there.
from django.db.models import Max, OuterRef, Subquery
newest = Message.objects.filter(room_id=OuterRef('pk')).order_by('-date')
for room in Room.objects.filter(users=user)
.annotate(last_message_date=Max('messages__date'),
last_message=Subquery(newest.values('content')[:1]))
.order_by('-last_message_date'):
print(r.__dict__)
As I understand the behavior, we prepare beforehand Subquery by saying to use id of the element who call the subquery to annotate the filtred elements and order then by -date. So it returns the most recent element and we take the value of field content to annotate it.
This will create a complex but unique request using all the performance of the database.
Is it possible to filter a models.DateTimeField but only get the month in the filter object?
The field is:
time_stamp = models.DateTimeField(
default=timezone.now)
When I filter it, this is what I get:
[datetime.datetime(2016, 9, 22, 15, 2, 48, 867473, tzinfo=),
datetime.datetime(2016, 9, 22, 15, 4, 22, 618675, tzinfo=),
datetime.datetime(2016, 9, 22, 15, 5, 20, 939593, tzinfo=)]
The filter returns 3 rows, but clearly there is too much information. I only require the months, and maybe the year.
How can I achieve this?
Any help or direction would be appreciated,
Thanks
If you are using django 1.10.x there is Extract db function
from django.db.models.functions import Extract
months = MyModel.objects.annotate(month_stamp=Extract('time_stamp', 'month')).values_list('month_stamp', flat=True)
For django 1.9.x
from django.db.models import Func
def Extract(field, date_field='DOW'):
template = "EXTRACT({} FROM %(expressions)s::timestamp)".format(date_field)
return Func(field, template=template)
months = MyModel.objects.annotate(month_stamp=Extract('time_stamp', 'month')).values_list('month_stamp', flat=True)
You can use propety:
Class your_model(models.Model):
time_stamp = models.DateTimeField(
default=timezone.now)
#property
def _get_year(self):
return self.time_stamp.strftime("%Y-%m")
year = property(_get_year) #EDIT