test(analyzer,serp): add tests for caching, single query, and parallel processing

- Add TestSingleQueryBugFix: verify SERP.query called once per analysis - Add TestPatentCaching: DB cache hit/miss, SERP query cache hit/miss - Add TestDynamicDateRange: rolling window, days_back param - Add TestFilesystemPDFCaching: skip download, redownload empty files - Add autouse mock_db fixture to prevent real DB connections in all tests
2026-03-24 14:39:09 -04:00
parent 1a297eb60b
commit 6f0b448044
2 changed files with 277 additions and 2 deletions
@@ -1,11 +1,22 @@
 """Tests for the high-level company analyzer orchestration."""

 import pytest
-from unittest.mock import Mock, patch, call
+from unittest.mock import Mock, patch, call, MagicMock
 from SPARC.analyzer import CompanyAnalyzer
 from SPARC.types import Patent, Patents, CompanyAnalysisResult, BatchAnalysisResult


+@pytest.fixture(autouse=True)
+def mock_db(mocker):
+    """Mock DatabaseClient for all tests so no real DB connection is needed."""
+    mock_db_cls = mocker.patch("SPARC.analyzer.DatabaseClient")
+    mock_db_instance = MagicMock()
+    mock_db_instance.get_cached_patent.return_value = None
+    mock_db_instance.get_cached_serp_query.return_value = None
+    mock_db_cls.return_value = mock_db_instance
+    return mock_db_instance
+
+
 class TestCompanyAnalyzer:
    """Test the CompanyAnalyzer orchestration logic."""

@@ -17,7 +28,7 @@ class TestCompanyAnalyzer:

        mock_llm.assert_called_once_with(api_key="test-key")

-    def test_analyze_company_full_pipeline(self, mocker):
+    def test_analyze_company_full_pipeline(self, mocker, mock_db):
        """Test complete company analysis pipeline."""
        # Mock all the dependencies
        mock_query = mocker.patch("SPARC.analyzer.SERP.query")
@@ -178,6 +189,180 @@ class TestCompanyAnalyzer:
        assert "PDF not found" in result


+class TestSingleQueryBugFix:
+    """Test that SERP.query is only called once per company analysis."""
+
+    def test_analyze_company_safe_calls_query_once(self, mocker, mock_db):
+        """_analyze_company_safe should call SERP.query exactly once."""
+        mock_query = mocker.patch("SPARC.analyzer.SERP.query")
+        mock_save = mocker.patch("SPARC.analyzer.SERP.save_patents")
+        mock_parse = mocker.patch("SPARC.analyzer.SERP.parse_patent_pdf")
+        mock_minimize = mocker.patch("SPARC.analyzer.SERP.minimize_patent_for_llm")
+        mock_llm = mocker.patch("SPARC.analyzer.LLMAnalyzer")
+
+        patent = Patent(patent_id="US123", pdf_link="http://example.com/test.pdf")
+        mock_query.return_value = Patents(patents=[patent])
+
+        def save_side_effect(p):
+            p.pdf_path = "patents/US123.pdf"
+            return p
+
+        mock_save.side_effect = save_side_effect
+        mock_parse.return_value = {"abstract": "Test"}
+        mock_minimize.return_value = "Content"
+
+        mock_llm_instance = Mock()
+        mock_llm_instance.analyze_patent_portfolio.return_value = "Analysis"
+        mock_llm.return_value = mock_llm_instance
+
+        analyzer = CompanyAnalyzer()
+        analyzer._analyze_company_safe("TestCorp")
+
+        # The key assertion: SERP.query called exactly once, not twice
+        mock_query.assert_called_once_with("TestCorp")
+
+    def test_analyze_company_with_prefetched_patents_skips_query(self, mocker):
+        """analyze_company should not call SERP.query when patents are provided."""
+        mock_query = mocker.patch("SPARC.analyzer.SERP.query")
+        mock_save = mocker.patch("SPARC.analyzer.SERP.save_patents")
+        mock_parse = mocker.patch("SPARC.analyzer.SERP.parse_patent_pdf")
+        mock_minimize = mocker.patch("SPARC.analyzer.SERP.minimize_patent_for_llm")
+        mock_llm = mocker.patch("SPARC.analyzer.LLMAnalyzer")
+
+        patent = Patent(patent_id="US123", pdf_link="http://example.com/test.pdf")
+        prefetched = Patents(patents=[patent])
+
+        def save_side_effect(p):
+            p.pdf_path = "patents/US123.pdf"
+            return p
+
+        mock_save.side_effect = save_side_effect
+        mock_parse.return_value = {"abstract": "Test"}
+        mock_minimize.return_value = "Content"
+
+        mock_llm_instance = Mock()
+        mock_llm_instance.analyze_patent_portfolio.return_value = "Analysis"
+        mock_llm.return_value = mock_llm_instance
+
+        analyzer = CompanyAnalyzer()
+        analyzer.analyze_company("TestCorp", patents=prefetched)
+
+        # SERP.query should never be called
+        mock_query.assert_not_called()
+
+
+class TestPatentCaching:
+    """Test patent-level DB caching in the pipeline."""
+
+    def test_process_single_patent_uses_db_cache(self, mocker, mock_db):
+        """_process_single_patent returns cached content when available."""
+        mock_save = mocker.patch("SPARC.analyzer.SERP.save_patents")
+
+        mock_db.get_cached_patent.return_value = {
+            "patent_id": "US123",
+            "minimized_content": "Cached minimized content",
+        }
+
+        patent = Patent(patent_id="US123", pdf_link="http://example.com/test.pdf")
+        result = CompanyAnalyzer._process_single_patent(patent, "TestCorp", mock_db)
+
+        assert result == {"patent_id": "US123", "content": "Cached minimized content"}
+        # Should NOT download since cache hit
+        mock_save.assert_not_called()
+
+    def test_process_single_patent_stores_to_db_cache(self, mocker, mock_db):
+        """_process_single_patent stores result in DB after processing."""
+        mock_save = mocker.patch("SPARC.analyzer.SERP.save_patents")
+        mock_parse = mocker.patch("SPARC.analyzer.SERP.parse_patent_pdf")
+        mock_minimize = mocker.patch("SPARC.analyzer.SERP.minimize_patent_for_llm")
+
+        # No cache hit
+        mock_db.get_cached_patent.return_value = None
+
+        patent = Patent(patent_id="US123", pdf_link="http://example.com/test.pdf")
+
+        def save_side_effect(p):
+            p.pdf_path = "patents/US123.pdf"
+            return p
+
+        mock_save.side_effect = save_side_effect
+        mock_parse.return_value = {"abstract": "Test abstract"}
+        mock_minimize.return_value = "Minimized content"
+
+        result = CompanyAnalyzer._process_single_patent(patent, "TestCorp", mock_db)
+
+        assert result == {"patent_id": "US123", "content": "Minimized content"}
+        mock_db.store_patent.assert_called_once_with(
+            patent_id="US123",
+            company_name="TestCorp",
+            pdf_link="http://example.com/test.pdf",
+            raw_sections={"abstract": "Test abstract"},
+            minimized_content="Minimized content",
+        )
+
+    def test_serp_query_cache_hit_skips_api(self, mocker, mock_db):
+        """When SERP query is cached, API call is skipped."""
+        mock_query = mocker.patch("SPARC.analyzer.SERP.query")
+        mock_save = mocker.patch("SPARC.analyzer.SERP.save_patents")
+        mock_parse = mocker.patch("SPARC.analyzer.SERP.parse_patent_pdf")
+        mock_minimize = mocker.patch("SPARC.analyzer.SERP.minimize_patent_for_llm")
+        mock_llm = mocker.patch("SPARC.analyzer.LLMAnalyzer")
+
+        # Simulate SERP cache hit
+        mock_db.get_cached_serp_query.return_value = ["US123"]
+        # Simulate patent cache hit too
+        mock_db.get_cached_patent.return_value = {
+            "patent_id": "US123",
+            "minimized_content": "Cached content",
+        }
+
+        mock_llm_instance = Mock()
+        mock_llm_instance.analyze_patent_portfolio.return_value = "Analysis"
+        mock_llm.return_value = mock_llm_instance
+
+        analyzer = CompanyAnalyzer()
+        result = analyzer.analyze_company("TestCorp")
+
+        assert result == "Analysis"
+        # SERP.query should NOT be called
+        mock_query.assert_not_called()
+        # No downloads should happen
+        mock_save.assert_not_called()
+
+    def test_serp_query_cache_miss_stores_result(self, mocker, mock_db):
+        """When SERP query cache misses, result is stored after API call."""
+        mock_query = mocker.patch("SPARC.analyzer.SERP.query")
+        mock_save = mocker.patch("SPARC.analyzer.SERP.save_patents")
+        mock_parse = mocker.patch("SPARC.analyzer.SERP.parse_patent_pdf")
+        mock_minimize = mocker.patch("SPARC.analyzer.SERP.minimize_patent_for_llm")
+        mock_llm = mocker.patch("SPARC.analyzer.LLMAnalyzer")
+
+        mock_db.get_cached_serp_query.return_value = None
+
+        patent = Patent(patent_id="US123", pdf_link="http://example.com/test.pdf")
+        mock_query.return_value = Patents(patents=[patent])
+
+        def save_side_effect(p):
+            p.pdf_path = "patents/US123.pdf"
+            return p
+
+        mock_save.side_effect = save_side_effect
+        mock_parse.return_value = {"abstract": "Test"}
+        mock_minimize.return_value = "Content"
+
+        mock_llm_instance = Mock()
+        mock_llm_instance.analyze_patent_portfolio.return_value = "Analysis"
+        mock_llm.return_value = mock_llm_instance
+
+        analyzer = CompanyAnalyzer()
+        analyzer.analyze_company("TestCorp")
+
+        mock_db.store_serp_query.assert_called_once()
+        call_kwargs = mock_db.store_serp_query.call_args[1]
+        assert call_kwargs["company_name"] == "TestCorp"
+        assert call_kwargs["patent_ids"] == ["US123"]
+
+
 class TestBatchProcessing:
    """Test multi-company batch processing functionality."""