Skip to content

~95% on SimpleQA (e.g. Qwen3.6-27B on a 3090). Supports all local and cloud LLMs (llama.cpp, Ollama, Google, ...). 10+ search engines - arXiv, PubMed, your private documents. Everything Local & Encrypted.

test: dedupe 22 redundant tests across evaluator pure_logic + high_value (#4267)

LearningCircuitd0b9ee0d899a
diff --git a/tests/advanced_search_system/evidence/test_evaluator_high_value.py b/tests/advanced_search_system/evidence/test_evaluator_high_value.py
index b89198f47f..abe50e03e1 100644
--- a/tests/advanced_search_system/evidence/test_evaluator_high_value.py
+++ b/tests/advanced_search_system/evidence/test_evaluator_high_value.py
@@ -206,27 +206,6 @@ class TestAssessMatchQuality:
value=value,
)
- def test_exact_value_match_returns_1(self, evaluator):
- evidence = self._make_evidence("The lake was formed during ice age")
- constraint = self._make_constraint("formed during ice age")
- assert evaluator._assess_match_quality(evidence, constraint) == 1.0
-
- def test_partial_word_match_returns_0_8(self, evaluator):
- evidence = self._make_evidence("The trail has glacial features")
- constraint = self._make_constraint("glacial ice formations")
- # "glacial" appears in evidence claim
- assert evaluator._assess_match_quality(evidence, constraint) == 0.8
-
- def test_no_match_returns_0_6(self, evaluator):
- evidence = self._make_evidence("The river flows south")
- constraint = self._make_constraint("volcanic eruption")
- assert evaluator._assess_match_quality(evidence, constraint) == 0.6
-
- def test_case_insensitive_exact_match(self, evaluator):
- evidence = self._make_evidence("Contains BODY PART in name")
- constraint = self._make_constraint("body part")
- assert evaluator._assess_match_quality(evidence, constraint) == 1.0
-
def test_case_insensitive_partial_match(self, evaluator):
evidence = self._make_evidence("Located near MOUNTAINS")
constraint = self._make_constraint("rocky mountains area")
diff --git a/tests/advanced_search_system/evidence/test_evaluator_pure_logic.py b/tests/advanced_search_system/evidence/test_evaluator_pure_logic.py
index abb7505f90..a80b017fde 100644
--- a/tests/advanced_search_system/evidence/test_evaluator_pure_logic.py
+++ b/tests/advanced_search_system/evidence/test_evaluator_pure_logic.py
@@ -51,54 +51,6 @@ def _evidence(claim="test claim"):
class TestParseEvidenceResponse:
"""Verify LLM response parsing into evidence components."""
- def test_all_fields_parsed(self):
- """All standard fields are extracted."""
- ev = _evaluator()
- content = (
- "CLAIM: The mountain is tall\n"
- "TYPE: direct_statement\n"
- "SOURCE: Official survey\n"
- "CONFIDENCE: 0.85\n"
- "REASONING: Based on measurements\n"
- "QUOTE: It stands at 14411 feet"
- )
- result = ev._parse_evidence_response(content)
- assert result["claim"] == "The mountain is tall"
- assert result["type"] == "direct_statement"
- assert result["source"] == "Official survey"
- assert result["confidence"] == "0.85"
- assert result["reasoning"] == "Based on measurements"
- assert result["quote"] == "It stands at 14411 feet"
-
- def test_confidence_extracts_float_from_text(self):
- """Confidence value extracted from surrounding text."""
- ev = _evaluator()
- content = "CONFIDENCE: approximately 0.75 based on evidence"
- result = ev._parse_evidence_response(content)
- assert result["confidence"] == "0.75"
-
- def test_confidence_integer(self):
- """Integer confidence is extracted."""
- ev = _evaluator()
- content = "CONFIDENCE: 1"
- result = ev._parse_evidence_response(content)
- assert result["confidence"] == "1"
-
- def test_unknown_fields_ignored(self):
- """Fields not in the expected set are ignored."""
- ev = _evaluator()
- content = "CLAIM: test\nFOO: bar\nNOTES: something"
- result = ev._parse_evidence_response(content)
- assert "claim" in result
- assert "foo" not in result
- assert "notes" not in result
-
- def test_empty_content(self):
- """Empty content returns empty dict."""
- ev = _evaluator()
- result = ev._parse_evidence_response("")
- assert result == {}
-
def test_no_colon_lines_skipped(self):
"""Lines without colons are skipped."""
ev = _evaluator()
@@ -107,100 +59,12 @@ class TestParseEvidenceResponse:
assert result["claim"] == "test"
assert result["type"] == "inference"
- def test_colon_in_value(self):
- """Values containing colons are preserved."""
- ev = _evaluator()
- content = "CLAIM: the ratio is 3:1 for water"
- result = ev._parse_evidence_response(content)
- assert result["claim"] == "the ratio is 3:1 for water"
-
- def test_case_insensitive_keys(self):
- """Keys are matched case-insensitively."""
- ev = _evaluator()
- content = "Claim: hello\nType: inference"
- result = ev._parse_evidence_response(content)
- assert result["claim"] == "hello"
- assert result["type"] == "inference"
-
# ---------------------------------------------------------------------------
# _parse_evidence_type
# ---------------------------------------------------------------------------
-class TestParseEvidenceType:
- """Verify evidence type string mapping."""
-
- def test_direct_statement(self):
- ev = _evaluator()
- assert (
- ev._parse_evidence_type("direct_statement")
- == EvidenceType.DIRECT_STATEMENT
- )
-
- def test_official_record(self):
- ev = _evaluator()
- assert (
- ev._parse_evidence_type("official_record")
- == EvidenceType.OFFICIAL_RECORD
- )
-
- def test_research_finding(self):
- ev = _evaluator()
- assert (
- ev._parse_evidence_type("research_finding")
- == EvidenceType.RESEARCH_FINDING
- )
-
- def test_news_report(self):
- ev = _evaluator()
- assert (
- ev._parse_evidence_type("news_report") == EvidenceType.NEWS_REPORT
- )
-
- def test_statistical_data(self):
- ev = _evaluator()
- assert (
- ev._parse_evidence_type("statistical_data")
- == EvidenceType.STATISTICAL_DATA
- )
-
- def test_inference(self):
- ev = _evaluator()
- assert ev._parse_evidence_type("inference") == EvidenceType.INFERENCE
-
- def test_correlation(self):
- ev = _evaluator()
- assert (
- ev._parse_evidence_type("correlation") == EvidenceType.CORRELATION
- )
-
- def test_speculation(self):
- ev = _evaluator()
- assert (
- ev._parse_evidence_type("speculation") == EvidenceType.SPECULATION
- )
-
- def test_unknown_defaults_to_speculation(self):
- ev = _evaluator()
- assert (
- ev._parse_evidence_type("unknown_type") == EvidenceType.SPECULATION
- )
-
- def test_case_insensitive(self):
- ev = _evaluator()
- assert (
- ev._parse_evidence_type("DIRECT_STATEMENT")
- == EvidenceType.DIRECT_STATEMENT
- )
-
- def test_mixed_case(self):
- ev = _evaluator()
- assert (
- ev._parse_evidence_type("News_Report") == EvidenceType.NEWS_REPORT
- )
-
-
# ---------------------------------------------------------------------------
# _assess_match_quality
# ---------------------------------------------------------------------------