import sys
from pathlib import Path
# Add project root to path for benchmark imports
sys.path.insert(0, str(Path(__file__).parent.parent.parent))
from benchmark.cli.build_locbench import _build_target_ranges, _cluster_lines_to_ranges
from benchmark.schemas import DatasetCase, GroundTruthEntry
class TestClusterLinesToRanges:
def test_merges_within_gap(self) -> None:
assert _cluster_lines_to_ranges([10, 14], gap=3) == [(10, 14)]
def test_splits_beyond_gap(self) -> None:
assert _cluster_lines_to_ranges([10, 15], gap=3) == [(10, 10), (15, 15)]
def test_filters_invalid_lines(self) -> None:
assert _cluster_lines_to_ranges([0, -1, 3, 3], gap=3) == [(3, 3)]
class TestBuildTargetRanges:
def test_falls_back_when_too_many_clusters(self) -> None:
# With GAP=3, these become 3 clusters -> fall back to a single span.
result = _build_target_ranges([1, 10, 20], context_start=1, context_end=30)
assert result == [(1, 20)]
def test_clamps_to_context(self) -> None:
result = _build_target_ranges([1, 10, 20], context_start=5, context_end=15)
assert result == [(5, 15)]
class TestDatasetCaseGroundTruthFiles:
def test_prefers_target_ranges(self) -> None:
case = DatasetCase(
id="c1",
query="q",
repo="r",
base_commit="abc",
hard_gt=[
GroundTruthEntry(
path="a.py",
function="foo",
range=(100, 200),
target_ranges=[(110, 112), (150, 150)],
)
],
)
assert case.ground_truth_files == {"a.py": [(110, 112), (150, 150)]}
assert case.ground_truth_context_files == {"a.py": [(100, 200)]}
def test_falls_back_to_context_range(self) -> None:
case = DatasetCase(
id="c2",
query="q",
repo="r",
base_commit="abc",
hard_gt=[
GroundTruthEntry(
path="a.py",
function="foo",
range=(10, 20),
target_ranges=[],
)
],
)
assert case.ground_truth_files == {"a.py": [(10, 20)]}
