[Issue #3685] Save search get opportunity id (#3704)

## Summary
Fixes [#{3685}](#3685)

### Time to review: __10 mins__

## Changes proposed
Query the search index limiting the results to only opportunity ID
Store the list of opportunity IDs returned to the DB as a list/array
(does not need to point to opportunity table as foreign keys)
Modify the users query to replace with static pagination 

## Additional information
Updated tests

api/src/adapters/search/opensearch_client.py

@@ -216,13 +216,18 @@ def search(
         search_query: dict,
         include_scores: bool = True,
         params: dict | None = None,
+        includes: list[str] | None = None,
         excludes: list[str] | None = None,
     ) -> SearchResponse:
         if params is None:
             params = {}
 
         response = self._client.search(
-            index=index_name, body=search_query, params=params, _source_excludes=excludes
+            index=index_name,
+            body=search_query,
+            params=params,
+            _source_includes=includes,
+            _source_excludes=excludes,
         )
         return SearchResponse.from_opensearch_response(response, include_scores)
 

api/src/api/users/user_routes.py

@@ -3,7 +3,8 @@
 
 import flask
 
-from src.adapters import db
+import src.adapters.search.flask_opensearch as flask_opensearch
+from src.adapters import db, search
 from src.adapters.db import flask_db
 from src.api import response
 from src.api.route_utils import raise_flask_error
@@ -249,8 +250,9 @@ def user_get_saved_opportunities(db_session: db.Session, user_id: UUID) -> respo
 @user_blueprint.doc(responses=[200, 401])
 @user_blueprint.auth_required(api_jwt_auth)
 @flask_db.with_db_session()
+@flask_opensearch.with_search_client()
 def user_save_search(
-    db_session: db.Session, user_id: UUID, json_data: dict
+    search_client: search.SearchClient, db_session: db.Session, user_id: UUID, json_data: dict
 ) -> response.ApiResponse:
     logger.info("POST /v1/users/:user_id/saved-searches")
 
@@ -261,7 +263,7 @@ def user_save_search(
         raise_flask_error(401, "Unauthorized user")
 
     with db_session.begin():
-        saved_search = create_saved_search(db_session, user_id, json_data)
+        saved_search = create_saved_search(search_client, db_session, user_id, json_data)
 
     logger.info(
         "Saved search for user",

api/src/services/opportunities_v1/search_opportunities.py

@@ -5,6 +5,7 @@
 from pydantic import BaseModel, Field
 
 import src.adapters.search as search
+from src.adapters.search.opensearch_response import SearchResponse
 from src.api.opportunities_v1.opportunity_schemas import OpportunityV1Schema
 from src.pagination.pagination_models import PaginationInfo, PaginationParams, SortDirection
 from src.search.search_config import get_search_config
@@ -53,6 +54,15 @@
     ScoringRule.DEFAULT: DEFAULT,
 }
 
+STATIC_PAGINATION = {
+    "pagination": {
+        "order_by": "post_date",
+        "page_offset": 1,
+        "page_size": 1000,
+        "sort_direction": "descending",
+    }
+}
+
 SCHEMA = OpportunityV1Schema()
 
 
@@ -154,7 +164,7 @@ def _add_aggregations(builder: search.SearchQueryBuilder) -> None:
     builder.aggregation_terms("agency", _adjust_field_name("agency_code"), size=1000)
 
 
-def _get_search_request(params: SearchOpportunityParams) -> dict:
+def _get_search_request(params: SearchOpportunityParams, aggregation: bool = True) -> dict:
     builder = search.SearchQueryBuilder()
 
     # Make sure total hit count gets counted for more than 10k records
@@ -176,25 +186,38 @@ def _get_search_request(params: SearchOpportunityParams) -> dict:
     # Filters
     _add_search_filters(builder, params.filters)
 
-    # Aggregations / Facet / Filter Counts
-    _add_aggregations(builder)
+    if aggregation:
+        # Aggregations / Facet / Filter Counts
+        _add_aggregations(builder)
 
     return builder.build()
 
 
-def search_opportunities(
-    search_client: search.SearchClient, raw_search_params: dict
-) -> Tuple[Sequence[dict], dict, PaginationInfo]:
-    search_params = SearchOpportunityParams.model_validate(raw_search_params)
-
+def _search_opportunities(
+    search_client: search.SearchClient,
+    search_params: SearchOpportunityParams,
+    includes: list | None = None,
+) -> SearchResponse:
     search_request = _get_search_request(search_params)
 
     index_alias = get_search_config().opportunity_search_index_alias
     logger.info(
         "Querying search index alias %s", index_alias, extra={"search_index_alias": index_alias}
     )
 
-    response = search_client.search(index_alias, search_request, excludes=["attachments"])
+    response = search_client.search(
+        index_alias, search_request, includes=includes, excludes=["attachments"]
+    )
+
+    return response
+
+
+def search_opportunities(
+    search_client: search.SearchClient, raw_search_params: dict
+) -> Tuple[Sequence[dict], dict, PaginationInfo]:
+
+    search_params = SearchOpportunityParams.model_validate(raw_search_params)
+    response = _search_opportunities(search_client, search_params)
 
     pagination_info = PaginationInfo(
         page_offset=search_params.pagination.page_offset,
@@ -213,3 +236,13 @@ def search_opportunities(
     records = SCHEMA.load(response.records, many=True)
 
     return records, response.aggregations, pagination_info
+
+
+def search_opportunities_id(search_client: search.SearchClient, search_query: dict) -> list:
+    # Override pagination when calling opensearch
+    updated_search_query = search_query | STATIC_PAGINATION
+    search_params = SearchOpportunityParams.model_validate(updated_search_query)
+
+    response = _search_opportunities(search_client, search_params, includes=["opportunity_id"])
+
+    return [opp["opportunity_id"] for opp in response.records]

api/src/services/users/create_saved_search.py

@@ -1,18 +1,25 @@
 import logging
 from uuid import UUID
 
-from src.adapters import db
+from src.adapters import db, search
 from src.db.models.user_models import UserSavedSearch
+from src.services.opportunities_v1.search_opportunities import search_opportunities_id
 
 logger = logging.getLogger(__name__)
 
 
-def create_saved_search(db_session: db.Session, user_id: UUID, json_data: dict) -> UserSavedSearch:
+def create_saved_search(
+    search_client: search.SearchClient, db_session: db.Session, user_id: UUID, json_data: dict
+) -> UserSavedSearch:
+
+    # Retrieve opportunity IDs
+    opportunity_ids = search_opportunities_id(search_client, json_data["search_query"])
+
     saved_search = UserSavedSearch(
         user_id=user_id,
         name=json_data["name"],
         search_query=json_data["search_query"],
-        searched_opportunity_ids=[],
+        searched_opportunity_ids=opportunity_ids,
     )
 
     db_session.add(saved_search)

api/tests/conftest.py

@@ -220,6 +220,14 @@ def opportunity_index_alias(search_client, monkeypatch_session):
     return alias
 
 
+@pytest.fixture(scope="class")
+def opportunity_search_index_class(search_client, monkeypatch):
+    # Note we don't actually create anything, this is just a random name
+    alias = f"test-opportunity-index-alias-{uuid.uuid4().int}"
+    monkeypatch.setenv("OPPORTUNITY_SEARCH_INDEX_ALIAS", alias)
+    return alias
+
+
 def _generate_rsa_key_pair():
     # Rather than define a private/public key, generate one for the tests
     key = rsa.generate_private_key(public_exponent=65537, key_size=2048)

api/tests/src/api/users/test_user_save_search_post.py

@@ -1,10 +1,57 @@
+import uuid
+from datetime import date
+
 import pytest
 
-from src.constants.lookup_constants import FundingInstrument
+from src.api.opportunities_v1.opportunity_schemas import OpportunityV1Schema
+from src.constants.lookup_constants import (
+    ApplicantType,
+    FundingCategory,
+    FundingInstrument,
+    OpportunityStatus,
+)
 from src.db.models.user_models import UserSavedSearch
 from tests.src.api.opportunities_v1.conftest import get_search_request
+from tests.src.api.opportunities_v1.test_opportunity_route_search import build_opp
 from tests.src.db.models.factories import UserFactory
 
+SPORTS = build_opp(
+    opportunity_title="Research into Sports administrator industry",
+    opportunity_number="EFG8532950",
+    agency="USAID",
+    summary_description="HHS-CDC is looking to further investigate this topic. Car matter style top quality generation effort. Computer purpose while consumer left.",
+    opportunity_status=OpportunityStatus.FORECASTED,
+    assistance_listings=[("79.718", "Huff LLC")],
+    applicant_types=[ApplicantType.OTHER],
+    funding_instruments=[FundingInstrument.GRANT],
+    funding_categories=[FundingCategory.SCIENCE_TECHNOLOGY_AND_OTHER_RESEARCH_AND_DEVELOPMENT],
+    post_date=date(2019, 12, 8),
+    close_date=date(2024, 12, 28),
+    is_cost_sharing=True,
+    expected_number_of_awards=1,
+    award_floor=402500,
+    award_ceiling=8050000,
+    estimated_total_program_funding=5000,
+)
+MEDICAL_LABORATORY = build_opp(
+    opportunity_title="Research into Medical laboratory scientific officer industry",
+    opportunity_number="AO-44-EMC-878",
+    agency="USAID",
+    summary_description="HHS-CDC is looking to further investigate this topic. Car matter style top quality generation effort. Computer purpose while consumer left.",
+    opportunity_status=OpportunityStatus.FORECASTED,
+    assistance_listings=[("43.012", "Brown LLC")],
+    applicant_types=[ApplicantType.OTHER],
+    funding_instruments=[FundingInstrument.GRANT],
+    funding_categories=[FundingCategory.SCIENCE_TECHNOLOGY_AND_OTHER_RESEARCH_AND_DEVELOPMENT],
+    post_date=date(2025, 1, 25),  #
+    close_date=date(2025, 6, 4),  #
+    is_cost_sharing=True,  #
+    expected_number_of_awards=1,
+    award_floor=402500,
+    award_ceiling=8050000,
+    estimated_total_program_funding=5000,
+)
+
 
 @pytest.fixture(autouse=True, scope="function")
 def clear_saved_searches(db_session):
@@ -13,6 +60,14 @@ def clear_saved_searches(db_session):
     yield
 
 
+@pytest.fixture
+def opportunity_search_index_alias(search_client, monkeypatch):
+    # Note we don't actually create anything, this is just a random name
+    alias = f"test-opportunity-search-index-alias-{uuid.uuid4().int}"
+    monkeypatch.setenv("OPPORTUNITY_SEARCH_INDEX_ALIAS", alias)
+    return alias
+
+
 def test_user_save_search_post_unauthorized_user(client, db_session, user, user_auth_token):
     # Try to save a search for a different user ID
     different_user = UserFactory.create()
@@ -71,14 +126,31 @@ def test_user_save_search_post_invalid_request(client, user, user_auth_token, db
     assert len(saved_searches) == 0
 
 
-def test_user_save_search_post(client, user, user_auth_token, enable_factory_create, db_session):
+def test_user_save_search_post(
+    client,
+    opportunity_index,
+    search_client,
+    user,
+    user_auth_token,
+    enable_factory_create,
+    db_session,
+    opportunity_search_index_alias,
+    monkeypatch,
+):
     # Test data
     search_name = "Test Search"
     search_query = get_search_request(
         funding_instrument_one_of=[FundingInstrument.GRANT],
-        agency_one_of=["LOC"],
+        agency_one_of=["USAID"],
     )
 
+    # Load into the search index
+    schema = OpportunityV1Schema()
+    json_records = [schema.dump(opp) for opp in [SPORTS, MEDICAL_LABORATORY]]
+    search_client.bulk_upsert(opportunity_index, json_records, "opportunity_id")
+
+    search_client.swap_alias_index(opportunity_index, opportunity_search_index_alias)
+
     # Make the request to save a search
     response = client.post(
         f"/v1/users/{user.user_id}/saved-searches",
@@ -88,18 +160,23 @@ def test_user_save_search_post(client, user, user_auth_token, enable_factory_cre
 
     assert response.status_code == 200
     assert response.json["message"] == "Success"
-
     # Verify the search was saved in the database
     saved_search = db_session.query(UserSavedSearch).one()
+
     assert saved_search.user_id == user.user_id
     assert saved_search.name == search_name
     assert saved_search.search_query == {
         "format": "json",
-        "filters": {"agency": {"one_of": ["LOC"]}, "funding_instrument": {"one_of": ["grant"]}},
+        "filters": {"agency": {"one_of": ["USAID"]}, "funding_instrument": {"one_of": ["grant"]}},
         "pagination": {
             "order_by": "opportunity_id",
             "page_size": 25,
             "page_offset": 1,
             "sort_direction": "ascending",
         },
     }
+    # Verify pagination for the query was over-written. searched_opportunity_ids should be ordered by "post_date"
+    assert saved_search.searched_opportunity_ids == [
+        MEDICAL_LABORATORY.opportunity_id,
+        SPORTS.opportunity_id,
+    ]