Files
notifications-api/app/dao/notifications_dao.py

811 lines
28 KiB
Python
Raw Normal View History

2021-03-10 13:55:06 +00:00
from datetime import datetime, timedelta
from itertools import groupby
from operator import attrgetter
from botocore.exceptions import ClientError
from flask import current_app
2021-03-10 13:55:06 +00:00
from notifications_utils.international_billing_rates import (
INTERNATIONAL_BILLING_RATES,
)
from notifications_utils.recipients import (
InvalidEmailError,
2021-03-10 13:55:06 +00:00
try_validate_and_format_phone_number,
validate_and_format_email_address,
)
from notifications_utils.timezones import convert_bst_to_utc, convert_utc_to_bst
from sqlalchemy import and_, asc, desc, func, or_, union
from sqlalchemy.orm import joinedload
from sqlalchemy.orm.exc import NoResultFound
from sqlalchemy.sql import functions
2018-12-10 16:27:59 +00:00
from sqlalchemy.sql.expression import case
from werkzeug.datastructures import MultiDict
from app import create_uuid, db, statsd_client
from app.dao.dao_utils import autocommit
from app.letters.utils import LetterPDFNotFound, find_letter_pdf_in_s3
2016-03-21 12:37:34 +00:00
from app.models import (
2021-03-10 13:55:06 +00:00
EMAIL_TYPE,
KEY_TYPE_NORMAL,
KEY_TYPE_TEST,
LETTER_TYPE,
NOTIFICATION_CREATED,
NOTIFICATION_DELIVERED,
NOTIFICATION_PENDING,
NOTIFICATION_PENDING_VIRUS_CHECK,
NOTIFICATION_PERMANENT_FAILURE,
2021-03-10 13:55:06 +00:00
NOTIFICATION_SENDING,
NOTIFICATION_SENT,
NOTIFICATION_STATUS_TYPES_COMPLETED,
2021-03-10 13:55:06 +00:00
NOTIFICATION_TEMPORARY_FAILURE,
SMS_TYPE,
2021-03-10 13:55:06 +00:00
FactNotificationStatus,
Notification,
NotificationHistory,
ProviderDetails,
)
from app.utils import (
escape_special_characters,
get_london_midnight_in_utc,
midnight_n_days_ago,
)
def dao_get_last_date_template_was_used(template_id, service_id):
last_date_from_notifications = db.session.query(
functions.max(Notification.created_at)
).filter(
Notification.service_id == service_id,
Notification.template_id == template_id,
Notification.key_type != KEY_TYPE_TEST
).scalar()
if last_date_from_notifications:
return last_date_from_notifications
last_date = db.session.query(
functions.max(FactNotificationStatus.bst_date)
).filter(
FactNotificationStatus.template_id == template_id,
FactNotificationStatus.key_type != KEY_TYPE_TEST
).scalar()
return last_date
@autocommit
def dao_create_notification(notification):
if not notification.id:
# need to populate defaulted fields before we create the notification history object
notification.id = create_uuid()
if not notification.status:
notification.status = NOTIFICATION_CREATED
db.session.add(notification)
def country_records_delivery(phone_prefix):
dlr = INTERNATIONAL_BILLING_RATES[phone_prefix]['attributes']['dlr']
return dlr and dlr.lower() == 'yes'
def _update_notification_status(notification, status, detailed_status_code=None):
# status = _decide_permanent_temporary_failure(
# status=status, notification=notification, detailed_status_code=detailed_status_code
# )
# notification.status = status
# dao_update_notification(notification)
return notification
@autocommit
def update_notification_status_by_id(notification_id, status, sent_by=None, detailed_status_code=None):
notification = Notification.query.with_for_update().filter(Notification.id == notification_id).first()
if not notification:
current_app.logger.info('notification not found for id {} (update to status {})'.format(
notification_id,
status
))
return None
if notification.status not in {
NOTIFICATION_CREATED,
NOTIFICATION_SENDING,
NOTIFICATION_PENDING,
NOTIFICATION_SENT,
NOTIFICATION_PENDING_VIRUS_CHECK
}:
_duplicate_update_warning(notification, status)
return None
if (
notification.notification_type == SMS_TYPE
and notification.international
and not country_records_delivery(notification.phone_prefix)
):
return None
if not notification.sent_by and sent_by:
notification.sent_by = sent_by
return _update_notification_status(
notification=notification,
status=status,
detailed_status_code=detailed_status_code
)
@autocommit
def update_notification_status_by_reference(reference, status):
# this is used to update letters and emails
notification = Notification.query.filter(Notification.reference == reference).first()
if not notification:
current_app.logger.error('notification not found for reference {} (update to {})'.format(reference, status))
return None
if notification.status not in {
NOTIFICATION_SENDING,
NOTIFICATION_PENDING
}:
_duplicate_update_warning(notification, status)
return None
return _update_notification_status(
notification=notification,
status=status
)
@autocommit
def dao_update_notification(notification):
notification.updated_at = datetime.utcnow()
db.session.add(notification)
def get_notifications_for_job(service_id, job_id, filter_dict=None, page=1, page_size=None):
if page_size is None:
page_size = current_app.config['PAGE_SIZE']
2016-03-21 12:37:34 +00:00
query = Notification.query.filter_by(service_id=service_id, job_id=job_id)
query = _filter_query(query, filter_dict)
return query.order_by(asc(Notification.job_row_number)).paginate(
page=page,
per_page=page_size
)
def dao_get_notification_count_for_job_id(*, job_id):
return Notification.query.filter_by(job_id=job_id).count()
def get_notification_with_personalisation(service_id, notification_id, key_type):
filter_dict = {'service_id': service_id, 'id': notification_id}
if key_type:
filter_dict['key_type'] = key_type
return Notification.query.filter_by(**filter_dict).options(joinedload('template')).one()
def get_notification_by_id(notification_id, service_id=None, _raise=False):
filters = [Notification.id == notification_id]
if service_id:
filters.append(Notification.service_id == service_id)
query = Notification.query.filter(*filters)
return query.one() if _raise else query.first()
def get_notifications_for_service(
service_id,
filter_dict=None,
page=1,
page_size=None,
count_pages=True,
limit_days=None,
key_type=None,
personalisation=False,
include_jobs=False,
include_from_test_key=False,
older_than=None,
client_reference=None,
include_one_off=True,
error_out=True
):
if page_size is None:
page_size = current_app.config['PAGE_SIZE']
filters = [Notification.service_id == service_id]
if limit_days is not None:
filters.append(Notification.created_at >= midnight_n_days_ago(limit_days))
if older_than is not None:
older_than_created_at = db.session.query(
Notification.created_at).filter(Notification.id == older_than).as_scalar()
filters.append(Notification.created_at < older_than_created_at)
if not include_jobs:
filters.append(Notification.job_id == None) # noqa
if not include_one_off:
filters.append(Notification.created_by_id == None) # noqa
if key_type is not None:
filters.append(Notification.key_type == key_type)
elif not include_from_test_key:
filters.append(Notification.key_type != KEY_TYPE_TEST)
if client_reference is not None:
filters.append(Notification.client_reference == client_reference)
query = Notification.query.filter(*filters)
query = _filter_query(query, filter_dict)
if personalisation:
query = query.options(
joinedload('template')
)
return query.order_by(desc(Notification.created_at)).paginate(
page=page,
per_page=page_size,
count=count_pages,
error_out=error_out,
)
2016-03-21 12:37:34 +00:00
def _filter_query(query, filter_dict=None):
if filter_dict is None:
2016-11-22 17:30:03 +00:00
return query
multidict = MultiDict(filter_dict)
# filter by status
statuses = multidict.getlist('status')
if statuses:
statuses = Notification.substitute_status(statuses)
query = query.filter(Notification.status.in_(statuses))
2016-11-22 17:30:03 +00:00
# filter by template
template_types = multidict.getlist('template_type')
if template_types:
query = query.filter(Notification.notification_type.in_(template_types))
2016-11-22 17:30:03 +00:00
return query
@autocommit
def insert_notification_history_delete_notifications(
notification_type, service_id, timestamp_to_delete_backwards_from, qry_limit=50000
):
2021-12-01 14:28:08 +00:00
"""
Delete up to 50,000 notifications that are past retention for a notification type and service.
Steps are as follows:
Create a temporary notifications table
Populate that table with up to 50k notifications that are to be deleted. (Note: no specified order)
Insert everything in the temp table into notification history
Delete from notifications if notification id is in the temp table
Drop the temp table (automatically when the transaction commits)
Temporary tables are in a separate postgres schema, and only visible to the current session (db connection,
in a celery task there's one connection per thread.)
"""
# Setting default query limit to 50,000 which take about 48 seconds on current table size
# 10, 000 took 11s and 100,000 took 1 min 30 seconds.
select_into_temp_table = """
2021-12-01 14:28:08 +00:00
CREATE TEMP TABLE NOTIFICATION_ARCHIVE ON COMMIT DROP AS
SELECT id, job_id, job_row_number, service_id, template_id, template_version, api_key_id,
key_type, notification_type, created_at, sent_at, sent_by, updated_at, reference, billable_units,
client_reference, international, phone_prefix, rate_multiplier, notification_status,
created_by_id, postage, document_download_count
FROM notifications
WHERE service_id = :service_id
AND notification_type = :notification_type
AND created_at < :timestamp_to_delete_backwards_from
AND key_type in ('normal', 'team')
limit :qry_limit
"""
select_into_temp_table_for_letters = """
2021-12-01 14:28:08 +00:00
CREATE TEMP TABLE NOTIFICATION_ARCHIVE ON COMMIT DROP AS
SELECT id, job_id, job_row_number, service_id, template_id, template_version, api_key_id,
key_type, notification_type, created_at, sent_at, sent_by, updated_at, reference, billable_units,
client_reference, international, phone_prefix, rate_multiplier, notification_status,
created_by_id, postage, document_download_count
FROM notifications
WHERE service_id = :service_id
AND notification_type = :notification_type
AND created_at < :timestamp_to_delete_backwards_from
AND notification_status NOT IN ('pending-virus-check', 'created', 'sending')
AND key_type in ('normal', 'team')
limit :qry_limit
"""
# Insert into NotificationHistory if the row already exists do nothing.
insert_query = """
insert into notification_history
SELECT * from NOTIFICATION_ARCHIVE
ON CONFLICT ON CONSTRAINT notification_history_pkey
DO NOTHING
"""
delete_query = """
DELETE FROM notifications
where id in (select id from NOTIFICATION_ARCHIVE)
"""
input_params = {
"service_id": service_id,
"notification_type": notification_type,
"timestamp_to_delete_backwards_from": timestamp_to_delete_backwards_from,
"qry_limit": qry_limit
}
2020-03-25 08:08:33 +00:00
select_to_use = select_into_temp_table_for_letters if notification_type == 'letter' else select_into_temp_table
db.session.execute(select_to_use, input_params)
result = db.session.execute("select count(*) from NOTIFICATION_ARCHIVE").fetchone()[0]
db.session.execute(insert_query)
db.session.execute(delete_query)
return result
def move_notifications_to_notification_history(
notification_type,
service_id,
timestamp_to_delete_backwards_from,
qry_limit=50000
):
deleted = 0
if notification_type == LETTER_TYPE:
_delete_letters_from_s3(
notification_type, service_id, timestamp_to_delete_backwards_from, qry_limit
)
delete_count_per_call = 1
while delete_count_per_call > 0:
delete_count_per_call = insert_notification_history_delete_notifications(
notification_type=notification_type,
service_id=service_id,
timestamp_to_delete_backwards_from=timestamp_to_delete_backwards_from,
qry_limit=qry_limit
)
deleted += delete_count_per_call
# Deleting test Notifications, test notifications are not persisted to NotificationHistory
Notification.query.filter(
Notification.notification_type == notification_type,
Notification.service_id == service_id,
Notification.created_at < timestamp_to_delete_backwards_from,
Notification.key_type == KEY_TYPE_TEST
).delete(synchronize_session=False)
2018-12-21 13:57:35 +00:00
db.session.commit()
return deleted
def _delete_letters_from_s3(
notification_type, service_id, date_to_delete_from, query_limit
):
letters_to_delete_from_s3 = db.session.query(
Notification
).filter(
Notification.notification_type == notification_type,
Notification.created_at < date_to_delete_from,
Notification.service_id == service_id,
# although letters in non completed statuses do have PDFs in s3, they do not exist in the
# production-letters-pdf bucket as they never made it that far so we do not try and delete
# them from it
Notification.status.in_(NOTIFICATION_STATUS_TYPES_COMPLETED)
).limit(query_limit).all()
for letter in letters_to_delete_from_s3:
try:
letter_pdf = find_letter_pdf_in_s3(letter)
letter_pdf.delete()
except ClientError:
current_app.logger.exception(
"Error deleting S3 object for letter: {}".format(letter.id))
except LetterPDFNotFound:
current_app.logger.warning(
"No S3 object to delete for letter: {}".format(letter.id))
@autocommit
def dao_delete_notifications_by_id(notification_id):
db.session.query(Notification).filter(
Notification.id == notification_id
).delete(synchronize_session='fetch')
def dao_timeout_notifications(cutoff_time, limit=100000):
"""
Set email and SMS notifications (only) to "temporary-failure" status
if they're still sending from before the specified cutoff_time.
"""
updated_at = datetime.utcnow()
current_statuses = [NOTIFICATION_SENDING, NOTIFICATION_PENDING]
new_status = NOTIFICATION_TEMPORARY_FAILURE
notifications = Notification.query.filter(
Notification.created_at < cutoff_time,
Notification.status.in_(current_statuses),
Notification.notification_type.in_([SMS_TYPE, EMAIL_TYPE])
).limit(limit).all()
Notification.query.filter(
Notification.id.in_([n.id for n in notifications]),
).update(
{'status': new_status, 'updated_at': updated_at},
synchronize_session=False
)
db.session.commit()
return notifications
def is_delivery_slow_for_providers(
created_at,
threshold,
delivery_time,
):
"""
Returns a dict of providers and whether they are currently slow or not. eg:
{
'mmg': True,
'firetext': False
}
"""
slow_notification_counts = db.session.query(
ProviderDetails.identifier,
case(
[(
Notification.status == NOTIFICATION_DELIVERED,
(Notification.updated_at - Notification.sent_at) >= delivery_time
)],
else_=(datetime.utcnow() - Notification.sent_at) >= delivery_time
).label("slow"),
func.count().label('count')
).select_from(
ProviderDetails
).outerjoin(
Notification, and_(
Notification.notification_type == SMS_TYPE,
Notification.sent_by == ProviderDetails.identifier,
Notification.created_at >= created_at,
Notification.sent_at.isnot(None),
Notification.status.in_([NOTIFICATION_DELIVERED, NOTIFICATION_PENDING, NOTIFICATION_SENDING]),
Notification.key_type != KEY_TYPE_TEST
)
).filter(
ProviderDetails.notification_type == 'sms',
ProviderDetails.active
).order_by(
ProviderDetails.identifier
).group_by(
ProviderDetails.identifier,
"slow"
)
slow_providers = {}
for provider, rows in groupby(slow_notification_counts, key=attrgetter('identifier')):
rows = list(rows)
total_notifications = sum(row.count for row in rows)
slow_notifications = sum(row.count for row in rows if row.slow)
slow_providers[provider] = (slow_notifications / total_notifications >= threshold)
statsd_client.gauge(f'slow-delivery.{provider}.ratio', slow_notifications / total_notifications)
return slow_providers
@autocommit
def dao_update_notifications_by_reference(references, update_dict):
updated_count = Notification.query.filter(
Notification.reference.in_(references)
).update(
update_dict,
synchronize_session=False
)
2017-04-07 11:22:03 +01:00
updated_history_count = 0
if updated_count != len(references):
updated_history_count = NotificationHistory.query.filter(
NotificationHistory.reference.in_(references)
).update(
update_dict,
synchronize_session=False
)
return updated_count, updated_history_count
def dao_get_notifications_by_recipient_or_reference(
service_id,
search_term,
notification_type=None,
statuses=None,
page=1,
page_size=None,
error_out=True,
):
if notification_type == SMS_TYPE:
normalised = try_validate_and_format_phone_number(search_term)
for character in {'(', ')', ' ', '-'}:
normalised = normalised.replace(character, '')
normalised = normalised.lstrip('+0')
elif notification_type == EMAIL_TYPE:
try:
normalised = validate_and_format_email_address(search_term)
except InvalidEmailError:
normalised = search_term.lower()
elif notification_type in {LETTER_TYPE, None}:
# For letters, we store the address without spaces, so we need
# to removes spaces from the search term to match. We also do
# this when a notification type isnt provided (this will
# happen if a user doesnt have permission to see the dashboard)
# because email addresses and phone numbers will never be stored
# with spaces either.
normalised = ''.join(search_term.split()).lower()
else:
raise TypeError(
f'Notification type must be {EMAIL_TYPE}, {SMS_TYPE}, {LETTER_TYPE} or None'
)
normalised = escape_special_characters(normalised)
search_term = escape_special_characters(search_term)
filters = [
Notification.service_id == service_id,
or_(
Notification.normalised_to.like("%{}%".format(normalised)),
Notification.client_reference.ilike("%{}%".format(search_term)),
),
Notification.key_type != KEY_TYPE_TEST,
]
if statuses:
filters.append(Notification.status.in_(statuses))
if notification_type:
filters.append(Notification.notification_type == notification_type)
results = db.session.query(Notification)\
.filter(*filters)\
.order_by(desc(Notification.created_at))\
.paginate(page=page, per_page=page_size, count=False, error_out=error_out)
return results
def dao_get_notification_by_reference(reference):
return Notification.query.filter(
Notification.reference == reference
).one()
2020-03-27 15:48:54 +00:00
def dao_get_notification_or_history_by_reference(reference):
try:
# This try except is necessary because in test keys and research mode does not create notification history.
# Otherwise we could just search for the NotificationHistory object
return Notification.query.filter(
Notification.reference == reference
).one()
except NoResultFound:
return NotificationHistory.query.filter(
NotificationHistory.reference == reference
).one()
def dao_get_notifications_processing_time_stats(start_date, end_date):
"""
For a given time range, returns the number of notifications sent and the number of
those notifications that we processed within 10 seconds
SELECT
count(notifications),
coalesce(sum(CASE WHEN sent_at - created_at <= interval '10 seconds' THEN 1 ELSE 0 END), 0)
FROM notifications
WHERE
created_at > 'START DATE' AND
created_at < 'END DATE' AND
api_key_id IS NOT NULL AND
key_type != 'test' AND
notification_type != 'letter';
"""
under_10_secs = Notification.sent_at - Notification.created_at <= timedelta(seconds=10)
sum_column = functions.coalesce(functions.sum(
case(
[
(under_10_secs, 1)
],
else_=0
)
), 0)
return db.session.query(
func.count(Notification.id).label('messages_total'),
sum_column.label('messages_within_10_secs')
).filter(
Notification.created_at >= start_date,
Notification.created_at < end_date,
Notification.api_key_id.isnot(None),
Notification.key_type != KEY_TYPE_TEST,
Notification.notification_type != LETTER_TYPE
).one()
def dao_get_last_notification_added_for_job_id(job_id):
last_notification_added = Notification.query.filter(
Notification.job_id == job_id
).order_by(
Notification.job_row_number.desc()
).first()
return last_notification_added
def notifications_not_yet_sent(should_be_sending_after_seconds, notification_type):
older_than_date = datetime.utcnow() - timedelta(seconds=should_be_sending_after_seconds)
notifications = Notification.query.filter(
Notification.created_at <= older_than_date,
Notification.notification_type == notification_type,
Notification.status == NOTIFICATION_CREATED
).all()
return notifications
use yield_per instead of limit limit means we only return 50k letters, if there are more than that for a service we'll skip them and they won't be picked up until the next day. If you remove the limit, sqlalchemy prefetches query results so it can build up ORM results, for example collapsing joined rows into single objects with chidren. SQLAlchemy streams the data into a buffer, and normally will still prefetch the entire resultset so it can ensure integrity of the session, (so that if you modify one result that is duplicated further down in the results, both rows are updated in the session for example). However, we don't care about that, but we do care about preventing the result set taking up too much memory. We can use `yield_per` to yield from sqlalchemy to the iterator (in this case the `for letter in letters_awaiting_sending` loop in letters_pdf_tasks.py) - this means every time we hit 10000 rows, we go back to the database to get the next 10k. This way, we only ever need 10k rows in memory at a time. This has some caveats, mostly around how we handle the data the query returns. They're a bit hard to parse but I'm pretty sure the notable limitations are: * It's dangerous to modify ORM objects returned by yield_per queries * It's dangerous to join in a yield_per query if you think there will be more than one row per item (for example, if you join from notification to service, there'll be multiple result rows containing the same service, and if these are split over different yield chunks, then we may experience undefined behaviour. These two limitations are focused around there being no guarantee of having one unique row per item. For more reading: https://docs.sqlalchemy.org/en/13/orm/query.html?highlight=yield_per#sqlalchemy.orm.query.Query.yield_per https://www.mail-archive.com/sqlalchemy@googlegroups.com/msg12443.html
2020-10-23 20:06:24 +01:00
def dao_get_letters_to_be_printed(print_run_deadline, postage, query_limit=10000):
"""
use yield_per instead of limit limit means we only return 50k letters, if there are more than that for a service we'll skip them and they won't be picked up until the next day. If you remove the limit, sqlalchemy prefetches query results so it can build up ORM results, for example collapsing joined rows into single objects with chidren. SQLAlchemy streams the data into a buffer, and normally will still prefetch the entire resultset so it can ensure integrity of the session, (so that if you modify one result that is duplicated further down in the results, both rows are updated in the session for example). However, we don't care about that, but we do care about preventing the result set taking up too much memory. We can use `yield_per` to yield from sqlalchemy to the iterator (in this case the `for letter in letters_awaiting_sending` loop in letters_pdf_tasks.py) - this means every time we hit 10000 rows, we go back to the database to get the next 10k. This way, we only ever need 10k rows in memory at a time. This has some caveats, mostly around how we handle the data the query returns. They're a bit hard to parse but I'm pretty sure the notable limitations are: * It's dangerous to modify ORM objects returned by yield_per queries * It's dangerous to join in a yield_per query if you think there will be more than one row per item (for example, if you join from notification to service, there'll be multiple result rows containing the same service, and if these are split over different yield chunks, then we may experience undefined behaviour. These two limitations are focused around there being no guarantee of having one unique row per item. For more reading: https://docs.sqlalchemy.org/en/13/orm/query.html?highlight=yield_per#sqlalchemy.orm.query.Query.yield_per https://www.mail-archive.com/sqlalchemy@googlegroups.com/msg12443.html
2020-10-23 20:06:24 +01:00
Return all letters created before the print run deadline that have not yet been sent. This yields in batches of 10k
to prevent the query taking too long and eating up too much memory. As each 10k batch is yielded, the
get_key_and_size_of_letters_to_be_sent_to_print function will go and fetch the s3 data, andhese start sending off
tasks to the notify-ftp app to send them.
CAUTION! Modify this query with caution. Modifying filters etc is fine, but if we join onto another table, then
there may be undefined behaviour. Essentially we need each ORM object returned for each row to be unique,
and we should avoid modifying state of returned objects.
For more reading:
https://docs.sqlalchemy.org/en/13/orm/query.html?highlight=yield_per#sqlalchemy.orm.query.Query.yield_per
https://www.mail-archive.com/sqlalchemy@googlegroups.com/msg12443.html
"""
notifications = Notification.query.filter(
Notification.created_at < convert_bst_to_utc(print_run_deadline),
Notification.notification_type == LETTER_TYPE,
Notification.status == NOTIFICATION_CREATED,
Notification.key_type == KEY_TYPE_NORMAL,
Notification.postage == postage,
Notification.billable_units > 0
).order_by(
Notification.service_id,
Notification.created_at
use yield_per instead of limit limit means we only return 50k letters, if there are more than that for a service we'll skip them and they won't be picked up until the next day. If you remove the limit, sqlalchemy prefetches query results so it can build up ORM results, for example collapsing joined rows into single objects with chidren. SQLAlchemy streams the data into a buffer, and normally will still prefetch the entire resultset so it can ensure integrity of the session, (so that if you modify one result that is duplicated further down in the results, both rows are updated in the session for example). However, we don't care about that, but we do care about preventing the result set taking up too much memory. We can use `yield_per` to yield from sqlalchemy to the iterator (in this case the `for letter in letters_awaiting_sending` loop in letters_pdf_tasks.py) - this means every time we hit 10000 rows, we go back to the database to get the next 10k. This way, we only ever need 10k rows in memory at a time. This has some caveats, mostly around how we handle the data the query returns. They're a bit hard to parse but I'm pretty sure the notable limitations are: * It's dangerous to modify ORM objects returned by yield_per queries * It's dangerous to join in a yield_per query if you think there will be more than one row per item (for example, if you join from notification to service, there'll be multiple result rows containing the same service, and if these are split over different yield chunks, then we may experience undefined behaviour. These two limitations are focused around there being no guarantee of having one unique row per item. For more reading: https://docs.sqlalchemy.org/en/13/orm/query.html?highlight=yield_per#sqlalchemy.orm.query.Query.yield_per https://www.mail-archive.com/sqlalchemy@googlegroups.com/msg12443.html
2020-10-23 20:06:24 +01:00
).yield_per(query_limit)
return notifications
def dao_get_letters_and_sheets_volume_by_postage(print_run_deadline):
notifications = db.session.query(
func.count(Notification.id).label('letters_count'),
func.sum(Notification.billable_units).label('sheets_count'),
Notification.postage
).filter(
Notification.created_at < convert_bst_to_utc(print_run_deadline),
Notification.notification_type == LETTER_TYPE,
Notification.status == NOTIFICATION_CREATED,
Notification.key_type == KEY_TYPE_NORMAL,
Notification.billable_units > 0
).group_by(
Notification.postage
).order_by(
Notification.postage
).all()
return notifications
def dao_old_letters_with_created_status():
yesterday_bst = convert_utc_to_bst(datetime.utcnow()) - timedelta(days=1)
last_processing_deadline = yesterday_bst.replace(hour=17, minute=30, second=0, microsecond=0)
notifications = Notification.query.filter(
Notification.created_at < convert_bst_to_utc(last_processing_deadline),
Notification.notification_type == LETTER_TYPE,
Notification.status == NOTIFICATION_CREATED
).order_by(
Notification.created_at
).all()
return notifications
def letters_missing_from_sending_bucket(seconds_to_subtract):
older_than_date = datetime.utcnow() - timedelta(seconds=seconds_to_subtract)
# We expect letters to have a `created` status, updated_at timestamp and billable units greater than zero.
notifications = Notification.query.filter(
Notification.billable_units == 0,
Notification.updated_at == None, # noqa
Notification.status == NOTIFICATION_CREATED,
Notification.created_at <= older_than_date,
Notification.notification_type == LETTER_TYPE,
Notification.key_type == KEY_TYPE_NORMAL
).order_by(
Notification.created_at
).all()
return notifications
def dao_precompiled_letters_still_pending_virus_check():
ninety_minutes_ago = datetime.utcnow() - timedelta(seconds=5400)
notifications = Notification.query.filter(
Notification.created_at < ninety_minutes_ago,
Notification.status == NOTIFICATION_PENDING_VIRUS_CHECK
).order_by(
Notification.created_at
).all()
return notifications
def _duplicate_update_warning(notification, status):
current_app.logger.info(
(
'Duplicate callback received for service {service_id}. '
'Notification ID {id} with type {type} sent by {sent_by}. '
'New status was {new_status}, current status is {old_status}. '
'This happened {time_diff} after being first set.'
).format(
id=notification.id,
old_status=notification.status,
new_status=status,
time_diff=datetime.utcnow() - (notification.updated_at or notification.created_at),
type=notification.notification_type,
sent_by=notification.sent_by,
service_id=notification.service_id
)
)
def get_service_ids_with_notifications_before(notification_type, timestamp):
return {
row.service_id
for row in db.session.query(
Notification.service_id
).filter(
Notification.notification_type == notification_type,
Notification.created_at < timestamp
).distinct()
}
def get_service_ids_with_notifications_on_date(notification_type, date):
start_date = get_london_midnight_in_utc(date)
end_date = get_london_midnight_in_utc(date + timedelta(days=1))
notification_table_query = db.session.query(
Notification.service_id.label('service_id')
).filter(
Notification.notification_type == notification_type,
# using >= + < is much more efficient than date(created_at)
Notification.created_at >= start_date,
Notification.created_at < end_date,
)
# Looking at this table is more efficient for historical notifications,
# provided the task to populate it has run before they were archived.
ft_status_table_query = db.session.query(
FactNotificationStatus.service_id.label('service_id')
).filter(
FactNotificationStatus.notification_type == notification_type,
FactNotificationStatus.bst_date == date,
)
return {
row.service_id for row in db.session.query(union(
notification_table_query, ft_status_table_query
).subquery()).distinct()
}