Sith/galaxy/management/commands/rule_galaxy.py

64 lines
2.3 KiB
Python
Raw Normal View History

# -*- coding:utf-8 -*
#
# Copyright 2023
# - Skia <skia@hya.sk>
#
# Ce fichier fait partie du site de l'Association des Étudiants de l'UTBM,
# http://ae.utbm.fr.
#
# This program is free software; you can redistribute it and/or modify it under
# the terms of the GNU General Public License a published by the Free Software
# Foundation; either version 3 of the License, or (at your option) any later
# version.
#
# This program is distributed in the hope that it will be useful, but WITHOUT
# ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS
# FOR A PARTICULAR PURPOSE. See the GNU General Public License for more
# details.
#
# You should have received a copy of the GNU General Public License along with
# this program; if not, write to the Free Sofware Foundation, Inc., 59 Temple
# Place - Suite 330, Boston, MA 02111-1307, USA.
#
#
2024-06-24 11:07:36 +00:00
import logging
Galaxy improvements (#628) * galaxy: improve logging and performance reporting * galaxy: add a full galaxy state test * galaxy: optimize user self score computation * galaxy: add 'generate_galaxy_test_data' command for development at scale * galaxy: big refactor Main changes: - Multiple Galaxy objects can now exist at the same time in DB. This allows for ruling a new galaxy while still displaying the old one. - The criteria to quickly know whether a user is a possible citizen is now a simple query on picture count. This avoids a very complicated query to database, that could often result in huge working memory load. With this change, it should be possible to run the galaxy even on a vanilla Postgres that didn't receive fine tuning for the Sith's galaxy. * galaxy: template: make the galaxy graph work and be usable with a lot of stars - Display focused star and its connections clearly - Display star label faintly by default for other stars to avoid overloading the graph - Hide non-focused lanes - Avoid clicks on non-highlighted, too far stars - Make the canva adapt its width to initial screen size, doesn't work dynamically * galaxy: better docstrings * galaxy: use bulk_create whenever possible This is a big performance gain, especially for the tests. Examples: ---- `./manage.py test galaxy.tests.GalaxyTest.test_full_galaxy_state` Measurements averaged over 3 run on *my machine*™: Before: 2min15s After: 1m41s ---- `./manage.py generate_galaxy_test_data --user-pack-count 1` Before: 48s After: 25s ---- `./manage.py rule_galaxy` (for 600 citizen, corresponding to 1 user-pack) Before: 14m4s After: 12m34s * core: populate: use a less ambiguous 'timezone.now()' When running the tests around midnight, the day is changing, leading to some values being offset to the next day depending on the timezone, and making some tests to fail. This ensure to use a less ambiguous `now` when populating the database. * write more extensive documentation - add documentation to previously documented classes and functions and refactor some of the documented one, in accordance to the PEP257 and ReStructuredText standards ; - add some type hints ; - use a NamedTuple for the `Galaxy.compute_users_score` method instead of a raw tuple. Also change a little bit the logic in the function which call the latter ; - add some additional parameter checks on a few functions ; - change a little bit the logic of the log level setting for the galaxy related commands. * galaxy: tests: split Model and View for more efficient data usage --------- Co-authored-by: maréchal <thgirod@hotmail.com>
2023-05-10 10:47:02 +00:00
import warnings
from django.core.management.base import BaseCommand
from django.db import connection
from galaxy.models import Galaxy
class Command(BaseCommand):
help = (
"Rule the Galaxy! "
"Reset the whole galaxy and compute once again all the relation scores of all users. "
"As the sith's users are rather numerous, this command might be quite expensive in memory "
"and CPU time. Please keep this fact in mind when scheduling calls to this command in a production "
"environment."
)
def handle(self, *args, **options):
logger = logging.getLogger("main")
Galaxy improvements (#628) * galaxy: improve logging and performance reporting * galaxy: add a full galaxy state test * galaxy: optimize user self score computation * galaxy: add 'generate_galaxy_test_data' command for development at scale * galaxy: big refactor Main changes: - Multiple Galaxy objects can now exist at the same time in DB. This allows for ruling a new galaxy while still displaying the old one. - The criteria to quickly know whether a user is a possible citizen is now a simple query on picture count. This avoids a very complicated query to database, that could often result in huge working memory load. With this change, it should be possible to run the galaxy even on a vanilla Postgres that didn't receive fine tuning for the Sith's galaxy. * galaxy: template: make the galaxy graph work and be usable with a lot of stars - Display focused star and its connections clearly - Display star label faintly by default for other stars to avoid overloading the graph - Hide non-focused lanes - Avoid clicks on non-highlighted, too far stars - Make the canva adapt its width to initial screen size, doesn't work dynamically * galaxy: better docstrings * galaxy: use bulk_create whenever possible This is a big performance gain, especially for the tests. Examples: ---- `./manage.py test galaxy.tests.GalaxyTest.test_full_galaxy_state` Measurements averaged over 3 run on *my machine*™: Before: 2min15s After: 1m41s ---- `./manage.py generate_galaxy_test_data --user-pack-count 1` Before: 48s After: 25s ---- `./manage.py rule_galaxy` (for 600 citizen, corresponding to 1 user-pack) Before: 14m4s After: 12m34s * core: populate: use a less ambiguous 'timezone.now()' When running the tests around midnight, the day is changing, leading to some values being offset to the next day depending on the timezone, and making some tests to fail. This ensure to use a less ambiguous `now` when populating the database. * write more extensive documentation - add documentation to previously documented classes and functions and refactor some of the documented one, in accordance to the PEP257 and ReStructuredText standards ; - add some type hints ; - use a NamedTuple for the `Galaxy.compute_users_score` method instead of a raw tuple. Also change a little bit the logic in the function which call the latter ; - add some additional parameter checks on a few functions ; - change a little bit the logic of the log level setting for the galaxy related commands. * galaxy: tests: split Model and View for more efficient data usage --------- Co-authored-by: maréchal <thgirod@hotmail.com>
2023-05-10 10:47:02 +00:00
if options["verbosity"] < 0 or 2 < options["verbosity"]:
warnings.warn("verbosity level should be between 0 and 2 included")
if options["verbosity"] == 2:
logger.setLevel(logging.DEBUG)
Galaxy improvements (#628) * galaxy: improve logging and performance reporting * galaxy: add a full galaxy state test * galaxy: optimize user self score computation * galaxy: add 'generate_galaxy_test_data' command for development at scale * galaxy: big refactor Main changes: - Multiple Galaxy objects can now exist at the same time in DB. This allows for ruling a new galaxy while still displaying the old one. - The criteria to quickly know whether a user is a possible citizen is now a simple query on picture count. This avoids a very complicated query to database, that could often result in huge working memory load. With this change, it should be possible to run the galaxy even on a vanilla Postgres that didn't receive fine tuning for the Sith's galaxy. * galaxy: template: make the galaxy graph work and be usable with a lot of stars - Display focused star and its connections clearly - Display star label faintly by default for other stars to avoid overloading the graph - Hide non-focused lanes - Avoid clicks on non-highlighted, too far stars - Make the canva adapt its width to initial screen size, doesn't work dynamically * galaxy: better docstrings * galaxy: use bulk_create whenever possible This is a big performance gain, especially for the tests. Examples: ---- `./manage.py test galaxy.tests.GalaxyTest.test_full_galaxy_state` Measurements averaged over 3 run on *my machine*™: Before: 2min15s After: 1m41s ---- `./manage.py generate_galaxy_test_data --user-pack-count 1` Before: 48s After: 25s ---- `./manage.py rule_galaxy` (for 600 citizen, corresponding to 1 user-pack) Before: 14m4s After: 12m34s * core: populate: use a less ambiguous 'timezone.now()' When running the tests around midnight, the day is changing, leading to some values being offset to the next day depending on the timezone, and making some tests to fail. This ensure to use a less ambiguous `now` when populating the database. * write more extensive documentation - add documentation to previously documented classes and functions and refactor some of the documented one, in accordance to the PEP257 and ReStructuredText standards ; - add some type hints ; - use a NamedTuple for the `Galaxy.compute_users_score` method instead of a raw tuple. Also change a little bit the logic in the function which call the latter ; - add some additional parameter checks on a few functions ; - change a little bit the logic of the log level setting for the galaxy related commands. * galaxy: tests: split Model and View for more efficient data usage --------- Co-authored-by: maréchal <thgirod@hotmail.com>
2023-05-10 10:47:02 +00:00
elif options["verbosity"] == 1:
logger.setLevel(logging.INFO)
else:
Galaxy improvements (#628) * galaxy: improve logging and performance reporting * galaxy: add a full galaxy state test * galaxy: optimize user self score computation * galaxy: add 'generate_galaxy_test_data' command for development at scale * galaxy: big refactor Main changes: - Multiple Galaxy objects can now exist at the same time in DB. This allows for ruling a new galaxy while still displaying the old one. - The criteria to quickly know whether a user is a possible citizen is now a simple query on picture count. This avoids a very complicated query to database, that could often result in huge working memory load. With this change, it should be possible to run the galaxy even on a vanilla Postgres that didn't receive fine tuning for the Sith's galaxy. * galaxy: template: make the galaxy graph work and be usable with a lot of stars - Display focused star and its connections clearly - Display star label faintly by default for other stars to avoid overloading the graph - Hide non-focused lanes - Avoid clicks on non-highlighted, too far stars - Make the canva adapt its width to initial screen size, doesn't work dynamically * galaxy: better docstrings * galaxy: use bulk_create whenever possible This is a big performance gain, especially for the tests. Examples: ---- `./manage.py test galaxy.tests.GalaxyTest.test_full_galaxy_state` Measurements averaged over 3 run on *my machine*™: Before: 2min15s After: 1m41s ---- `./manage.py generate_galaxy_test_data --user-pack-count 1` Before: 48s After: 25s ---- `./manage.py rule_galaxy` (for 600 citizen, corresponding to 1 user-pack) Before: 14m4s After: 12m34s * core: populate: use a less ambiguous 'timezone.now()' When running the tests around midnight, the day is changing, leading to some values being offset to the next day depending on the timezone, and making some tests to fail. This ensure to use a less ambiguous `now` when populating the database. * write more extensive documentation - add documentation to previously documented classes and functions and refactor some of the documented one, in accordance to the PEP257 and ReStructuredText standards ; - add some type hints ; - use a NamedTuple for the `Galaxy.compute_users_score` method instead of a raw tuple. Also change a little bit the logic in the function which call the latter ; - add some additional parameter checks on a few functions ; - change a little bit the logic of the log level setting for the galaxy related commands. * galaxy: tests: split Model and View for more efficient data usage --------- Co-authored-by: maréchal <thgirod@hotmail.com>
2023-05-10 10:47:02 +00:00
logger.setLevel(logging.ERROR)
logger.info("The Galaxy is being ruled by the Sith.")
Galaxy improvements (#628) * galaxy: improve logging and performance reporting * galaxy: add a full galaxy state test * galaxy: optimize user self score computation * galaxy: add 'generate_galaxy_test_data' command for development at scale * galaxy: big refactor Main changes: - Multiple Galaxy objects can now exist at the same time in DB. This allows for ruling a new galaxy while still displaying the old one. - The criteria to quickly know whether a user is a possible citizen is now a simple query on picture count. This avoids a very complicated query to database, that could often result in huge working memory load. With this change, it should be possible to run the galaxy even on a vanilla Postgres that didn't receive fine tuning for the Sith's galaxy. * galaxy: template: make the galaxy graph work and be usable with a lot of stars - Display focused star and its connections clearly - Display star label faintly by default for other stars to avoid overloading the graph - Hide non-focused lanes - Avoid clicks on non-highlighted, too far stars - Make the canva adapt its width to initial screen size, doesn't work dynamically * galaxy: better docstrings * galaxy: use bulk_create whenever possible This is a big performance gain, especially for the tests. Examples: ---- `./manage.py test galaxy.tests.GalaxyTest.test_full_galaxy_state` Measurements averaged over 3 run on *my machine*™: Before: 2min15s After: 1m41s ---- `./manage.py generate_galaxy_test_data --user-pack-count 1` Before: 48s After: 25s ---- `./manage.py rule_galaxy` (for 600 citizen, corresponding to 1 user-pack) Before: 14m4s After: 12m34s * core: populate: use a less ambiguous 'timezone.now()' When running the tests around midnight, the day is changing, leading to some values being offset to the next day depending on the timezone, and making some tests to fail. This ensure to use a less ambiguous `now` when populating the database. * write more extensive documentation - add documentation to previously documented classes and functions and refactor some of the documented one, in accordance to the PEP257 and ReStructuredText standards ; - add some type hints ; - use a NamedTuple for the `Galaxy.compute_users_score` method instead of a raw tuple. Also change a little bit the logic in the function which call the latter ; - add some additional parameter checks on a few functions ; - change a little bit the logic of the log level setting for the galaxy related commands. * galaxy: tests: split Model and View for more efficient data usage --------- Co-authored-by: maréchal <thgirod@hotmail.com>
2023-05-10 10:47:02 +00:00
galaxy = Galaxy.objects.create()
galaxy.rule()
logger.info("Sending old galaxies' remains to garbage.")
Galaxy.objects.filter(state__isnull=True).delete()
logger.info("Ruled the galaxy in {} queries.".format(len(connection.queries)))
if options["verbosity"] > 2:
for q in connection.queries:
logger.debug(q)