test: 补充知识处理、检索与 Agent 回归用例

覆盖切块、向量检索、工具调用和历史预算；模型相关用例使用桩。

backend/tests/repositories/test_knowledge_source_limit.py

@@ -0,0 +1,117 @@
+"""
+Copyright 2024-2026 ChatterMate
+
+Licensed under the Apache License, Version 2.0 (the "License");
+you may not use this file except in compliance with the License.
+You may obtain a copy of the License at
+
+    http://www.apache.org/licenses/LICENSE-2.0
+
+Unless required by applicable law or agreed to in writing, software
+distributed under the License is distributed on an "AS IS" BASIS,
+WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+See the License for the specific language governing permissions and
+limitations under the License.
+"""
+
+from datetime import datetime, timedelta, timezone
+from uuid import uuid4
+
+import pytest
+
+from app.models.knowledge import Knowledge, SourceType
+from app.models.knowledge_queue import KnowledgeQueue, QueueStatus
+from app.repositories.knowledge import KnowledgeRepository
+
+
+@pytest.fixture
+def repo(db):
+    return KnowledgeRepository(db)
+
+
+def _index(db, org_id, source: str) -> Knowledge:
+    """A finished crawl: the row the source ends up as."""
+    row = Knowledge(
+        source=source,
+        source_type=SourceType.FILE,
+        organization_id=org_id,
+        created_at=datetime.now(timezone.utc),
+    )
+    db.add(row)
+    db.commit()
+    return row
+
+
+def _queue(db, org_id, source: str, status=QueueStatus.PENDING, age=None) -> KnowledgeQueue:
+    """A crawl that has been accepted but has produced nothing yet."""
+    started = datetime.now(timezone.utc) - (age or timedelta(0))
+    item = KnowledgeQueue(
+        organization_id=org_id,
+        source_type="website",
+        source=source,
+        status=status,
+        created_at=started,
+        updated_at=started,
+    )
+    db.add(item)
+    db.commit()
+    return item
+
+
+class TestCountSourcesInUse:
+    def test_counts_indexed_sources(self, repo, db, test_organization):
+        assert repo.count_sources_in_use(test_organization.id) == 0
+        _index(db, test_organization.id, "https://a.example")
+        assert repo.count_sources_in_use(test_organization.id) == 1
+
+    @pytest.mark.parametrize("status", [QueueStatus.PENDING, QueueStatus.PROCESSING])
+    def test_a_queued_crawl_holds_its_slot(self, repo, db, test_organization, status):
+        """The bug: three sites queued at once each saw a count of zero, so a
+        one-source plan ended up with three."""
+        _queue(db, test_organization.id, "https://a.example", status)
+        _queue(db, test_organization.id, "https://b.example", status)
+        assert repo.count_sources_in_use(test_organization.id) == 2
+
+    def test_a_finished_crawl_is_not_counted_twice(self, repo, db, test_organization):
+        """The queue row stays behind once the source is indexed."""
+        _queue(db, test_organization.id, "https://a.example", QueueStatus.COMPLETED)
+        _index(db, test_organization.id, "https://a.example")
+        assert repo.count_sources_in_use(test_organization.id) == 1
+
+    def test_the_same_url_queued_twice_holds_two_slots(self, repo, db, test_organization):
+        """Each queue row becomes its own knowledge row - the worker does not
+        merge repeats - so counting distinct sources let a customer who kept
+        pressing Add walk past the limit."""
+        _queue(db, test_organization.id, "https://a.example")
+        _queue(db, test_organization.id, "https://a.example")
+        assert repo.count_sources_in_use(test_organization.id) == 2
+
+    def test_a_crawl_abandoned_mid_flight_stops_holding_its_slot(self, repo, db, test_organization):
+        """Only PENDING rows are ever retried, so a row left PROCESSING by a
+        restart would otherwise cost a slot for ever."""
+        _queue(db, test_organization.id, "https://a.example", QueueStatus.PROCESSING,
+               age=KnowledgeRepository.STALE_PROCESSING_AFTER + timedelta(minutes=1))
+        assert repo.count_sources_in_use(test_organization.id) == 0
+
+    def test_a_long_running_crawl_keeps_its_slot(self, repo, db, test_organization):
+        _queue(db, test_organization.id, "https://a.example", QueueStatus.PROCESSING,
+               age=KnowledgeRepository.STALE_PROCESSING_AFTER - timedelta(minutes=5))
+        assert repo.count_sources_in_use(test_organization.id) == 1
+
+    def test_failed_and_completed_crawls_hold_nothing(self, repo, db, test_organization):
+        _queue(db, test_organization.id, "https://a.example", QueueStatus.FAILED)
+        _queue(db, test_organization.id, "https://b.example", QueueStatus.COMPLETED)
+        assert repo.count_sources_in_use(test_organization.id) == 0
+
+    def test_another_organization_is_not_counted(self, repo, db, test_organization):
+        other_org = uuid4()
+        _queue(db, other_org, "https://a.example")
+        _index(db, other_org, "https://b.example")
+        assert repo.count_sources_in_use(test_organization.id) == 0
+
+    def test_indexed_count_still_excludes_the_queue(self, repo, db, test_organization):
+        """count_by_organization backs the paginated list, so it must keep
+        counting only what the list can actually show."""
+        _queue(db, test_organization.id, "https://a.example")
+        assert repo.count_by_organization(test_organization.id) == 0
+        assert repo.count_sources_in_use(test_organization.id) == 1