[DOP-19781] Add full-text search to users

Author: Ilyas Gasanov

docs/changelog/next_release/98.feature.rst

@@ -0,0 +1 @@
+Add full-text search for **users**

syncmaster/backend/api/v1/users.py

@@ -22,11 +22,17 @@ async def get_users(
     page_size: int = Query(gt=0, le=200, default=20),
     current_user: User = Depends(get_user(is_active=True)),
     unit_of_work: UnitOfWork = Depends(UnitOfWorkMarker),
+    search_query: str | None = Query(
+        None,
+        title="Search Query",
+        description="full-text search for users",
+    ),
 ) -> UserPageSchema:
     pagination = await unit_of_work.user.paginate(
         page=page,
         page_size=page_size,
         is_superuser=current_user.is_superuser,
+        search_query=search_query,
     )
     return UserPageSchema.from_pagination(pagination)
 

syncmaster/db/migrations/versions/2024-10-07_1df3778daf4f_add_search_vector_to_user_table.py

@@ -0,0 +1,38 @@
+# SPDX-FileCopyrightText: 2023-2024 MTS PJSC
+# SPDX-License-Identifier: Apache-2.0
+"""add search_vector to user table
+
+Revision ID: 1df3778daf4f
+Revises: b9f5c4315bb2
+Create Date: 2024-10-07 06:59:12.667067
+
+"""
+import sqlalchemy as sa
+from alembic import op
+from sqlalchemy.dialects import postgresql
+
+# revision identifiers, used by Alembic.
+revision = "1df3778daf4f"
+down_revision = "b9f5c4315bb2"
+branch_labels = None
+depends_on = None
+
+
+def upgrade() -> None:
+    # ### commands auto generated by Alembic - please adjust! ###
+    op.add_column(
+        "user",
+        sa.Column(
+            "search_vector",
+            postgresql.TSVECTOR(),
+            sa.Computed("to_tsvector('english'::regconfig, username)", persisted=True),
+            nullable=False,
+        ),
+    )
+    # ### end Alembic commands ###
+
+
+def downgrade() -> None:
+    # ### commands auto generated by Alembic - please adjust! ###
+    op.drop_column("user", "search_vector")
+    # ### end Alembic commands ###

syncmaster/db/models.py

@@ -37,6 +37,16 @@ class User(Base, TimestampMixin, DeletableMixin):
     username: Mapped[str] = mapped_column(String(256), nullable=False, unique=True, index=True)
     is_superuser: Mapped[bool] = mapped_column(Boolean, default=False, nullable=False)
     is_active: Mapped[bool] = mapped_column(Boolean, default=False, nullable=False)
+    search_vector: Mapped[str] = mapped_column(
+        TSVECTOR,
+        Computed(
+            "to_tsvector('english'::regconfig, username)",
+            persisted=True,
+        ),
+        nullable=False,
+        deferred=True,
+        doc="Full-text search vector",
+    )
 
     def __repr__(self) -> str:
         return f"User(username={self.username}, is_superuser={self.is_superuser}, is_active={self.is_active})"

syncmaster/db/repositories/user.py

@@ -2,7 +2,7 @@
 # SPDX-License-Identifier: Apache-2.0
 from typing import Any, NoReturn
 
-from sqlalchemy import ScalarResult, insert, select
+from sqlalchemy import ScalarResult, func, insert, select
 from sqlalchemy.exc import DBAPIError, IntegrityError, NoResultFound
 from sqlalchemy.ext.asyncio import AsyncSession
 
@@ -17,11 +17,24 @@ class UserRepository(Repository[User]):
     def __init__(self, session: AsyncSession) -> None:
         super().__init__(model=User, session=session)
 
-    async def paginate(self, page: int, page_size: int, is_superuser: bool) -> Pagination:
+    async def paginate(
+        self,
+        page: int,
+        page_size: int,
+        is_superuser: bool,
+        search_query: str | None = None,
+    ) -> Pagination:
         stmt = select(User).where(User.is_deleted.is_(False))
 
+        if search_query:
+            ts_query = func.plainto_tsquery("english", search_query)
+            stmt = stmt.where(User.search_vector.op("@@")(ts_query))
+            stmt = stmt.add_columns(func.ts_rank(User.search_vector, ts_query).label("rank"))
+            stmt = stmt.order_by(func.ts_rank(User.search_vector, ts_query).desc())
+
         if not is_superuser:
             stmt = stmt.where(User.is_active.is_(True))
+
         return await self._paginate_scalar_result(query=stmt.order_by(User.username), page=page, page_size=page_size)
 
     async def read_by_id(self, user_id: int, **kwargs: Any) -> User:

tests/test_unit/tests_users/__init__.py

@@ -6,13 +6,13 @@
 pytestmark = [pytest.mark.asyncio, pytest.mark.backend]
 
 
-async def test_get_users(
+async def test_get_user(
     client: AsyncClient,
     simple_user: MockUser,
     inactive_user: MockUser,
     deleted_user: MockUser,
 ):
-    response = await client.get("v1/users")
+    response = await client.get(f"/v1/users/{simple_user.id}")
     assert response.status_code == 401
     assert response.json() == {
         "error": {
@@ -22,72 +22,35 @@ async def test_get_users(
         },
     }
 
+    # check from simple user
     response = await client.get(
-        "v1/users",
+        f"v1/users/{simple_user.id}",
         headers={"Authorization": f"Bearer {simple_user.token}"},
     )
     assert response.status_code == 200
-    result = response.json()
-    assert result.keys() == {"items", "meta"}
-    assert result["items"][0].keys() == {"id", "is_superuser", "username"}
-    assert result["meta"] == {
-        "page": 1,
-        "pages": 1,
-        "total": len(result["items"]),
-        "page_size": 20,
-        "has_next": False,
-        "has_previous": False,
-        "next_page": None,
-        "previous_page": None,
+    assert response.json() == {
+        "id": simple_user.id,
+        "is_superuser": simple_user.is_superuser,
+        "username": simple_user.username,
     }
-    for user_data in result["items"]:
-        assert user_data["username"] != deleted_user.username
 
+    # check from simple user deleted user
     response = await client.get(
-        "v1/users",
-        headers={"Authorization": f"Bearer {inactive_user.token}"},
+        f"v1/users/{deleted_user.id}",
+        headers={"Authorization": f"Bearer {simple_user.token}"},
     )
-    assert response.status_code == 403
-    assert response.json() == {
-        "error": {
-            "code": "forbidden",
-            "message": "Inactive user",
-            "details": None,
-        },
-    }
-
-
-async def test_get_current_user(
-    client: AsyncClient,
-    simple_user: MockUser,
-    inactive_user: MockUser,
-):
-    # not authenticated user
-    response = await client.get("/v1/users/me")
-    assert response.status_code == 401
+    assert response.status_code == 404
     assert response.json() == {
         "error": {
-            "code": "unauthorized",
-            "message": "Not authenticated",
+            "code": "not_found",
+            "message": "User not found",
             "details": None,
         },
     }
 
-    # active user
-    response = await client.get(
-        "/v1/users/me",
-        headers={"Authorization": f"Bearer {simple_user.token}"},
-    )
-    assert response.status_code == 200
-    assert response.json() == {
-        "id": simple_user.id,
-        "is_superuser": simple_user.is_superuser,
-        "username": simple_user.username,
-    }
-
-    # inactive user
+    # check from inactive user
     response = await client.get(
-        "/v1/users/me",
+        f"v1/users/{simple_user.id}",
         headers={"Authorization": f"Bearer {inactive_user.token}"},
     )
     assert response.status_code == 403
@@ -100,13 +63,13 @@ async def test_get_current_user(
     }
 
 
-async def test_get_user(
+async def test_get_current_user(
     client: AsyncClient,
     simple_user: MockUser,
     inactive_user: MockUser,
-    deleted_user: MockUser,
 ):
-    response = await client.get(f"/v1/users/{simple_user.id}")
+    # not authenticated user
+    response = await client.get("/v1/users/me")
     assert response.status_code == 401
     assert response.json() == {
         "error": {
@@ -116,9 +79,9 @@ async def test_get_user(
         },
     }
 
-    # check from simple user
+    # active user
     response = await client.get(
-        f"v1/users/{simple_user.id}",
+        "/v1/users/me",
         headers={"Authorization": f"Bearer {simple_user.token}"},
     )
     assert response.status_code == 200
@@ -128,23 +91,9 @@ async def test_get_user(
         "username": simple_user.username,
     }
 
-    # check from simple user deleted user
-    response = await client.get(
-        f"v1/users/{deleted_user.id}",
-        headers={"Authorization": f"Bearer {simple_user.token}"},
-    )
-    assert response.status_code == 404
-    assert response.json() == {
-        "error": {
-            "code": "not_found",
-            "message": "User not found",
-            "details": None,
-        },
-    }
-
-    # check from inactive user
+    # inactive user
     response = await client.get(
-        f"v1/users/{simple_user.id}",
+        "/v1/users/me",
         headers={"Authorization": f"Bearer {inactive_user.token}"},
     )
     assert response.status_code == 403

tests/test_unit/test_users.py …/test_unit/tests_users/test_read_user.pytests/test_unit/test_users.py renamed to tests/test_unit/tests_users/test_read_user.py tests/test_unit/test_users.py renamed to tests/test_unit/tests_users/test_read_user.py

@@ -0,0 +1,120 @@
+import random
+import string
+
+import pytest
+from httpx import AsyncClient
+
+from tests.mocks import MockUser
+
+pytestmark = [pytest.mark.asyncio, pytest.mark.backend]
+
+
+async def test_get_users(
+    client: AsyncClient,
+    simple_user: MockUser,
+    inactive_user: MockUser,
+    deleted_user: MockUser,
+):
+    response = await client.get("v1/users")
+    assert response.status_code == 401
+    assert response.json() == {
+        "error": {
+            "code": "unauthorized",
+            "message": "Not authenticated",
+            "details": None,
+        },
+    }
+
+    response = await client.get(
+        "v1/users",
+        headers={"Authorization": f"Bearer {simple_user.token}"},
+    )
+    assert response.status_code == 200
+    result = response.json()
+    assert result.keys() == {"items", "meta"}
+    assert result["items"][0].keys() == {"id", "is_superuser", "username"}
+    assert result["meta"] == {
+        "page": 1,
+        "pages": 1,
+        "total": len(result["items"]),
+        "page_size": 20,
+        "has_next": False,
+        "has_previous": False,
+        "next_page": None,
+        "previous_page": None,
+    }
+    for user_data in result["items"]:
+        assert user_data["username"] != deleted_user.username
+
+    response = await client.get(
+        "v1/users",
+        headers={"Authorization": f"Bearer {inactive_user.token}"},
+    )
+    assert response.status_code == 403
+    assert response.json() == {
+        "error": {
+            "code": "forbidden",
+            "message": "Inactive user",
+            "details": None,
+        },
+    }
+
+
+@pytest.mark.parametrize(
+    "search_value_extractor",
+    [
+        lambda user: user.username,
+    ],
+    ids=["search_by_username"],
+)
+async def test_search_users_with_query(
+    client: AsyncClient,
+    superuser: MockUser,
+    simple_user: MockUser,
+    search_value_extractor,
+):
+    user = simple_user
+    search_query = search_value_extractor(user)
+
+    result = await client.get(
+        "v1/users",
+        headers={"Authorization": f"Bearer {superuser.token}"},
+        params={"search_query": search_query},
+    )
+
+    assert result.json() == {
+        "meta": {
+            "page": 1,
+            "pages": 1,
+            "total": 1,
+            "page_size": 20,
+            "has_next": False,
+            "has_previous": False,
+            "next_page": None,
+            "previous_page": None,
+        },
+        "items": [
+            {
+                "id": user.id,
+                "username": user.username,
+                "is_superuser": user.is_superuser,
+            },
+        ],
+    }
+    assert result.status_code == 200
+
+
+async def test_search_users_with_nonexistent_query(
+    client: AsyncClient,
+    superuser: MockUser,
+):
+    random_search_query = "".join(random.choices(string.ascii_lowercase + string.digits, k=12))
+
+    result = await client.get(
+        "v1/users",
+        headers={"Authorization": f"Bearer {superuser.token}"},
+        params={"search_query": random_search_query},
+    )
+
+    assert result.status_code == 200
+    assert result.json()["items"] == []