alex/ArchiveBox
1
0
Fork 0
mirror of https://github.com/ArchiveBox/ArchiveBox.git synced 2026-04-05 23:37:58 +10:00
Code Issues Packages Projects Releases Wiki Activity

Use QuerySets for search backend API instead of pks

Browse Source
This commit is contained in:
JDC
2020-11-19 18:19:33 -05:00
committed by Nick Sweeting
parent f383648ffc
commit 823df34080
3 changed files with 16 additions and 9 deletions
Download Patch File Download Diff File Expand all files Collapse all files

19
archivebox/search/__init__.py
View File

@@ -1,7 +1,8 @@
from typing import List, Union, Generator
from typing import List, Union
from pathlib import Path
from importlib import import_module
from django.db.models import QuerySet
from archivebox.index.schema import Link
from archivebox.util import enforce_types
@@ -39,16 +40,24 @@ def write_search_index(link: Link, texts: Union[List[str], None]=None, out_dir:
backend.index(snapshot_id=str(snap.id), texts=texts)
@enforce_types
def query_search_index(text: str) -> List[str]:
def query_search_index(query: str, out_dir: Path=OUTPUT_DIR) -> QuerySet:
if search_backend_enabled():
setup_django(out_dir, check_db=True)
from core.models import Snapshot
backend = import_backend()
return backend.search(text)
snapshot_ids = backend.search(query)
# TODO preserve ordering from backend
qsearch = Snapshot.objects.filter(pk__in=snapshot_ids)
return qsearch
else:
return []
@enforce_types
def flush_search_index(snapshot_ids: Generator[str, None, None]):
if not indexing_enabled() or not snapshot_ids:
def flush_search_index(snapshots: QuerySet):
if not indexing_enabled() or not snapshots:
return
backend = import_backend()
snapshot_ids=(str(pk) for pk in snapshots.values_list('pk',flat=True))
backend.flush(snapshot_ids)