From 7a9448a3e52e216f4f11b996be12ab87b99fe4bc Mon Sep 17 00:00:00 2001
From: moson-mo <mo-son@mailbox.org>
Date: Tue, 29 Nov 2022 14:45:24 +0100
Subject: [PATCH] perf: improve packages search-query
Improves performance for queries with large result sets.
The "group by" clause can be removed for all search types but the keywords.
Signed-off-by: moson-mo <mo-son@mailbox.org>
---
aurweb/packages/search.py | 5 ++++-
aurweb/routers/packages.py | 28 ++++++++++++----------------
test/test_packages_routes.py | 17 +++++++++++++++++
3 files changed, 33 insertions(+), 17 deletions(-)
diff --git a/aurweb/packages/search.py b/aurweb/packages/search.py
index c0740cda8..d5e00110f 100644
--- a/aurweb/packages/search.py
+++ b/aurweb/packages/search.py
@@ -136,7 +136,10 @@ class PackageSearch:
self._join_user()
self._join_keywords()
keywords = set(k.lower() for k in keywords)
- self.query = self.query.filter(PackageKeyword.Keyword.in_(keywords))
+ self.query = self.query.filter(PackageKeyword.Keyword.in_(keywords)).group_by(
+ models.Package.Name
+ )
+
return self
def _search_by_maintainer(self, keywords: str) -> orm.Query:
diff --git a/aurweb/routers/packages.py b/aurweb/routers/packages.py
index a4aac4966..6a943dbfb 100644
--- a/aurweb/routers/packages.py
+++ b/aurweb/routers/packages.py
@@ -93,22 +93,18 @@ async def packages_get(
search.sort_by(sort_by, sort_order)
# Insert search results into the context.
- results = (
- search.results()
- .with_entities(
- models.Package.ID,
- models.Package.Name,
- models.Package.PackageBaseID,
- models.Package.Version,
- models.Package.Description,
- models.PackageBase.Popularity,
- models.PackageBase.NumVotes,
- models.PackageBase.OutOfDateTS,
- models.User.Username.label("Maintainer"),
- models.PackageVote.PackageBaseID.label("Voted"),
- models.PackageNotification.PackageBaseID.label("Notify"),
- )
- .group_by(models.Package.Name)
+ results = search.results().with_entities(
+ models.Package.ID,
+ models.Package.Name,
+ models.Package.PackageBaseID,
+ models.Package.Version,
+ models.Package.Description,
+ models.PackageBase.Popularity,
+ models.PackageBase.NumVotes,
+ models.PackageBase.OutOfDateTS,
+ models.User.Username.label("Maintainer"),
+ models.PackageVote.PackageBaseID.label("Voted"),
+ models.PackageNotification.PackageBaseID.label("Notify"),
)
packages = results.limit(per_page).offset(offset)
diff --git a/test/test_packages_routes.py b/test/test_packages_routes.py
index bf1799639..f9cea694b 100644
--- a/test/test_packages_routes.py
+++ b/test/test_packages_routes.py
@@ -740,6 +740,23 @@ def test_packages_search_by_keywords(client: TestClient, packages: list[Package]
rows = root.xpath('//table[@class="results"]/tbody/tr')
assert len(rows) == 1
+ # Now let's add another keyword to the same package
+ with db.begin():
+ db.create(
+ PackageKeyword, PackageBase=package.PackageBase, Keyword="testKeyword2"
+ )
+
+ # And request packages with both keywords, we should still get 1 result.
+ with client as request:
+ response = request.get(
+ "/packages", params={"SeB": "k", "K": "testKeyword testKeyword2"}
+ )
+ assert response.status_code == int(HTTPStatus.OK)
+
+ root = parse_root(response.text)
+ rows = root.xpath('//table[@class="results"]/tbody/tr')
+ assert len(rows) == 1
+
def test_packages_search_by_maintainer(
client: TestClient, maintainer: User, package: Package
--
GitLab