lib.itmens/catalog/search/models.py

212 lines
6.3 KiB
Python
Raw Normal View History

2023-12-31 08:32:19 -05:00
# pyright: reportFunctionMemberAccess=false
import hashlib
2024-12-08 19:33:05 +00:00
from urllib.parse import quote_plus
2023-06-03 11:10:48 -04:00
import django_rq
2023-06-19 16:37:35 -04:00
from auditlog.context import set_actor
from django.conf import settings
2023-07-08 17:30:56 -04:00
from django.core.cache import cache
2024-05-25 23:38:11 -04:00
from loguru import logger
from rq.job import Job
from catalog.common.downloaders import RESPONSE_CENSORSHIP, DownloadError
2024-12-08 19:33:05 +00:00
from catalog.common.models import ItemCategory, SiteName
from catalog.common.sites import SiteManager
from ..models import Item, TVSeason
from .typesense import Indexer as TypeSenseIndexer
2023-06-03 11:10:48 -04:00
# from .meilisearch import Indexer as MeiliSearchIndexer
class DbIndexer:
@classmethod
def check(cls):
pass
2023-09-25 23:22:34 +00:00
@classmethod
def init(cls):
pass
2023-06-03 11:10:48 -04:00
@classmethod
2023-07-12 01:11:15 -04:00
def search(cls, q, page=1, categories=None, tag=None, sort=None):
2024-04-06 00:13:50 -04:00
result = lambda: None # noqa
2023-06-03 11:10:48 -04:00
result.items = Item.objects.filter(title__contains=q)[:10]
result.num_pages = 1
result.count = len(result.items)
return result
@classmethod
def update_model_indexable(cls, model):
pass
2023-06-26 08:12:56 -04:00
@classmethod
def register_list_model(cls, list_model):
pass
@classmethod
def register_piece_model(cls, model):
pass
2023-06-03 11:10:48 -04:00
2024-12-08 19:33:05 +00:00
class ExternalSearchResultItem:
def __init__(
self,
category: ItemCategory | None,
source_site: SiteName,
source_url: str,
title: str,
subtitle: str,
brief: str,
cover_url: str,
):
self.class_name = "base"
self.category = category
self.external_resources = {
"all": [
{
"url": source_url,
"site_name": source_site,
"site_label": source_site,
}
]
}
self.source_site = source_site
self.source_url = source_url
self.display_title = title
self.subtitle = subtitle
self.display_description = brief
self.cover_image_url = cover_url
2025-01-04 11:23:07 -05:00
def __repr__(self):
return f"[{self.category}] {self.display_title} {self.url}"
2024-12-08 19:33:05 +00:00
@property
def verbose_category_name(self):
return self.category.label if self.category else ""
@property
def url(self):
return f"/search?q={quote_plus(self.source_url)}"
@property
def scraped(self):
return False
2023-06-03 11:10:48 -04:00
# if settings.SEARCH_BACKEND == "MEILISEARCH":
#
# el
if settings.SEARCH_BACKEND == "TYPESENSE":
Indexer = TypeSenseIndexer
else:
Indexer = DbIndexer
2023-07-12 01:11:15 -04:00
def query_index(keywords, categories=None, tag=None, page=1, prepare_external=True):
2023-06-25 11:48:09 -04:00
if (
page < 1
or page > 99
or (not tag and isinstance(keywords, str) and len(keywords) < 2)
):
2023-06-10 22:44:50 -04:00
return [], 0, 0, []
2023-07-12 01:11:15 -04:00
result = Indexer.search(keywords, page=page, categories=categories, tag=tag)
2023-06-08 19:21:02 -04:00
keys = set()
2023-06-03 11:10:48 -04:00
items = []
2023-06-08 19:21:02 -04:00
duplicated_items = []
2023-06-03 11:10:48 -04:00
urls = []
for i in result.items:
2023-06-08 19:21:02 -04:00
if i.is_deleted or i.merged_to_item: # only happen if index is delayed
continue
2023-06-09 12:24:44 -04:00
if i.class_name == "work": # TODO: add searchable_item_class global config
continue
2023-06-08 19:21:02 -04:00
my_key = (
[i.isbn]
2023-06-03 11:10:48 -04:00
if hasattr(i, "isbn")
2023-06-08 19:21:02 -04:00
else ([i.imdb_code] if hasattr(i, "imdb_code") else [])
2023-06-03 11:10:48 -04:00
)
2023-06-08 19:21:02 -04:00
if hasattr(i, "works"):
my_key += [i[0] for i in i.works.all().values_list("id")]
if len(my_key):
2024-04-06 00:13:50 -04:00
sl = len(keys) + len(my_key)
2023-06-08 19:21:02 -04:00
keys.update(my_key)
# check and skip dup with same imdb or isbn or works id
2024-04-06 00:13:50 -04:00
if len(keys) < sl:
2023-06-08 19:21:02 -04:00
duplicated_items.append(i)
else:
items.append(i)
else:
2023-06-03 11:10:48 -04:00
items.append(i)
for res in i.external_resources.all():
urls.append(res.url)
2023-06-08 19:21:02 -04:00
# hide show if its season exists
seasons = [i for i in items if i.__class__ == TVSeason]
for season in seasons:
if season.show in items:
duplicated_items.append(season.show)
items.remove(season.show)
2023-06-03 11:10:48 -04:00
if prepare_external:
# store site url to avoid dups in external search
2023-07-12 01:11:15 -04:00
cache_key = f"search_{','.join(categories or [])}_{keywords}"
2023-06-03 11:10:48 -04:00
urls = list(set(cache.get(cache_key, []) + urls))
cache.set(cache_key, urls, timeout=300)
2023-06-08 19:21:02 -04:00
return items, result.num_pages, result.count, duplicated_items
2023-06-03 11:10:48 -04:00
2023-12-04 09:11:32 -05:00
def get_fetch_lock(user, url):
if user and user.is_authenticated:
_fetch_lock_key = f"_fetch_lock:{user.id}"
_fetch_lock_ttl = 1 if settings.DEBUG else 3
else:
_fetch_lock_key = "_fetch_lock"
_fetch_lock_ttl = 1 if settings.DEBUG else 15
if cache.get(_fetch_lock_key):
return False
cache.set(_fetch_lock_key, 1, timeout=_fetch_lock_ttl)
# do not fetch the same url twice in 2 hours
_fetch_lock_key = f"_fetch_lock:{url}"
_fetch_lock_ttl = 1 if settings.DEBUG else 7200
2023-07-08 17:30:56 -04:00
if cache.get(_fetch_lock_key):
return False
cache.set(_fetch_lock_key, 1, timeout=_fetch_lock_ttl)
return True
2023-06-19 16:37:35 -04:00
def enqueue_fetch(url, is_refetch, user=None):
2023-06-03 11:10:48 -04:00
job_id = "fetch_" + hashlib.md5(url.encode()).hexdigest()
in_progress = False
try:
job = Job.fetch(id=job_id, connection=django_rq.get_connection("fetch"))
in_progress = job.get_status() in ["queued", "started"]
2024-04-06 00:13:50 -04:00
except Exception:
2023-06-03 11:10:48 -04:00
in_progress = False
if not in_progress:
django_rq.get_queue("fetch").enqueue(
2023-06-19 16:37:35 -04:00
_fetch_task, url, is_refetch, user, job_id=job_id
2023-06-03 11:10:48 -04:00
)
return job_id
2023-06-19 16:37:35 -04:00
def _fetch_task(url, is_refetch, user):
2023-06-03 11:10:48 -04:00
item_url = "-"
2023-08-05 15:59:37 -04:00
with set_actor(user if user and user.is_authenticated else None):
2023-06-19 16:37:35 -04:00
try:
site = SiteManager.get_site_by_url(url)
if not site:
return None
site.get_resource_ready(ignore_existing_content=is_refetch)
item = site.get_item()
if item:
2024-05-25 23:38:11 -04:00
logger.info(f"fetched {url} {item.url} {item}")
2023-06-19 16:37:35 -04:00
item_url = item.url
else:
2024-05-25 23:38:11 -04:00
logger.error(f"fetch {url} failed")
except DownloadError as e:
if e.response_type != RESPONSE_CENSORSHIP:
logger.error(f"fetch {url} error", extra={"exception": e})
2023-06-19 16:37:35 -04:00
except Exception as e:
logger.error(f"parse {url} error", extra={"exception": e})
2023-06-19 16:37:35 -04:00
return item_url