fix(dataset): stop split mutation, determinize VOC, guard collisions (#2394)
- `train_test_split` seeded the global `random` module and shuffled the caller's list in place, so `DetectionDataset.split()` reordered its own `image_paths` and polluted process-wide randomness; use a local `random.Random` and shuffle a copy - Pascal VOC class ids were assigned in `set`-iteration and filesystem-glob order, so the same dataset produced different `class_id` values across runs; sort class names and the loaded file list - dataset exports keyed output files on basename, silently overwriting when two entries shared a name across directories (common after `merge()`); detect basename collisions and raise - add regression tests for split determinism, VOC id stability, and export collisions * fix(dataset): add LabelMe collision guard, hoist pre-flight checks, make guard private * test(dataset): add collision guard tests for as_yolo, as_pascal_voc, and boundary cases * docs(dataset): document ValueError raises, fix stale docstrings, add non-mutation guarantee --------- Co-authored-by: claude[bot] <209825114+claude[bot]@users.noreply.github.com> Co-authored-by: OpenAI Codex <codex@openai.com> Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
This commit is contained in:
Jirka Borovec
GitHub
parent
15dbbb5cb1
commit
f173905c8b
|
|
@ -37,6 +37,7 @@ from supervision.dataset.formats.yolo import (
|
|||
save_yolo_annotations,
|
||||
)
|
||||
from supervision.dataset.utils import (
|
||||
_check_no_basename_collisions,
|
||||
build_class_index_mapping,
|
||||
map_detections_class_id,
|
||||
merge_class_lists,
|
||||
|
|
@ -181,7 +182,7 @@ class DetectionDataset(BaseDataset):
|
|||
) -> tuple[DetectionDataset, DetectionDataset]:
|
||||
"""
|
||||
Splits the dataset into two parts (training and testing)
|
||||
using the provided split_ratio.
|
||||
using the provided split_ratio. The input dataset is not mutated.
|
||||
|
||||
Args:
|
||||
split_ratio: The ratio of the training
|
||||
|
|
@ -378,7 +379,28 @@ class DetectionDataset(BaseDataset):
|
|||
polygon points to be removed from the input polygon,
|
||||
in the range [0, 1). Argument is used only for segmentation datasets.
|
||||
show_progress: If True, display a progress bar during saving.
|
||||
|
||||
Raises:
|
||||
ValueError: If two image paths share the same basename (when
|
||||
images_directory_path is set) or the same stem (when
|
||||
annotations_directory_path is set), which would cause one
|
||||
output file to overwrite another. Rename images to ensure
|
||||
unique basenames before exporting a merged dataset.
|
||||
"""
|
||||
# Pre-flight: validate output uniqueness before writing any file
|
||||
if images_directory_path:
|
||||
_check_no_basename_collisions(
|
||||
image_paths=self.image_paths,
|
||||
key=lambda image_path: Path(image_path).name,
|
||||
output_kind="image",
|
||||
)
|
||||
if annotations_directory_path:
|
||||
_check_no_basename_collisions(
|
||||
image_paths=self.image_paths,
|
||||
key=lambda image_path: f"{Path(image_path).stem}.xml",
|
||||
output_kind="Pascal VOC annotation",
|
||||
)
|
||||
|
||||
if images_directory_path:
|
||||
save_dataset_images(
|
||||
dataset=self,
|
||||
|
|
@ -386,6 +408,11 @@ class DetectionDataset(BaseDataset):
|
|||
show_progress=show_progress,
|
||||
)
|
||||
if annotations_directory_path:
|
||||
_check_no_basename_collisions(
|
||||
image_paths=self.image_paths,
|
||||
key=lambda image_path: f"{Path(image_path).stem}.xml",
|
||||
output_kind="Pascal VOC annotation",
|
||||
)
|
||||
Path(annotations_directory_path).mkdir(parents=True, exist_ok=True)
|
||||
for image_path, image, annotations in tqdm(
|
||||
self,
|
||||
|
|
@ -580,6 +607,12 @@ class DetectionDataset(BaseDataset):
|
|||
`from_yolo(..., is_obb=True)`. Masks are ignored when
|
||||
`is_obb=True`.
|
||||
show_progress: If True, display a progress bar during saving.
|
||||
|
||||
Raises:
|
||||
ValueError: If two image paths share the same basename (when
|
||||
images_directory_path is set) or the same annotation
|
||||
file name (when annotations_directory_path is set),
|
||||
which would cause one output file to overwrite another.
|
||||
"""
|
||||
if is_obb and (
|
||||
min_image_area_percentage != 0.0
|
||||
|
|
@ -595,6 +628,20 @@ class DetectionDataset(BaseDataset):
|
|||
UserWarning,
|
||||
stacklevel=2,
|
||||
)
|
||||
# Pre-flight: validate output uniqueness before writing any file
|
||||
if images_directory_path:
|
||||
_check_no_basename_collisions(
|
||||
image_paths=self.image_paths,
|
||||
key=lambda image_path: Path(image_path).name,
|
||||
output_kind="image",
|
||||
)
|
||||
if annotations_directory_path:
|
||||
_check_no_basename_collisions(
|
||||
image_paths=self.image_paths,
|
||||
key=lambda image_path: Path(image_path).stem + ".txt",
|
||||
output_kind="YOLO annotation",
|
||||
)
|
||||
|
||||
if images_directory_path is not None:
|
||||
save_dataset_images(
|
||||
dataset=self,
|
||||
|
|
|
|||
|
|
@ -7,6 +7,7 @@ from typing import TYPE_CHECKING, Any
|
|||
import numpy as np
|
||||
import numpy.typing as npt
|
||||
|
||||
from supervision.dataset.utils import _check_no_basename_collisions
|
||||
from supervision.detection.core import Detections
|
||||
from supervision.detection.utils.converters import (
|
||||
mask_to_polygons,
|
||||
|
|
@ -363,6 +364,10 @@ def save_labelme_annotations(
|
|||
annotations_directory_path: Directory where the LabelMe ``.json`` files
|
||||
are written (created if it does not exist).
|
||||
|
||||
Raises:
|
||||
ValueError: If two image paths map to the same output .json stem,
|
||||
which would cause one annotation file to overwrite another.
|
||||
|
||||
Examples:
|
||||
```python
|
||||
import supervision as sv
|
||||
|
|
@ -375,6 +380,11 @@ def save_labelme_annotations(
|
|||
)
|
||||
```
|
||||
"""
|
||||
_check_no_basename_collisions(
|
||||
image_paths=dataset.image_paths,
|
||||
key=lambda image_path: f"{Path(image_path).stem}.json",
|
||||
output_kind="LabelMe annotation",
|
||||
)
|
||||
Path(annotations_directory_path).mkdir(parents=True, exist_ok=True)
|
||||
for image_path, image, detections in dataset:
|
||||
image_height, image_width, _ = image.shape
|
||||
|
|
|
|||
|
|
@ -184,8 +184,7 @@ def load_pascal_voc_annotations(
|
|||
show_progress: bool = False,
|
||||
) -> tuple[list[str], list[str], dict[str, Detections]]:
|
||||
"""
|
||||
Loads PASCAL VOC XML annotations and returns the image name,
|
||||
a Detections instance, and a list of class names.
|
||||
Load Pascal VOC XML annotations in sorted image-path order.
|
||||
|
||||
Args:
|
||||
images_directory_path: The path to the directory containing the images.
|
||||
|
|
@ -196,17 +195,17 @@ def load_pascal_voc_annotations(
|
|||
show_progress: If True, display a progress bar during loading.
|
||||
|
||||
Returns:
|
||||
A tuple with a list
|
||||
of class names, a list of paths to images, and a dictionary with image
|
||||
paths as keys and corresponding Detections instances as values.
|
||||
A tuple with a list of class names, a sorted list of paths to images,
|
||||
and a dictionary with image paths as keys and corresponding
|
||||
Detections instances as values.
|
||||
"""
|
||||
|
||||
image_paths = [
|
||||
image_paths = sorted(
|
||||
str(path)
|
||||
for path in list_files_with_extensions(
|
||||
directory=images_directory_path, extensions=["jpg", "jpeg", "png"]
|
||||
)
|
||||
]
|
||||
)
|
||||
|
||||
classes: list[str] = []
|
||||
annotations = {}
|
||||
|
|
@ -270,6 +269,14 @@ def detections_from_xml_obj(
|
|||
</object>
|
||||
</annotation>
|
||||
|
||||
Args:
|
||||
root: Parsed Pascal VOC ``<annotation>`` XML element.
|
||||
classes: Existing class names used to assign stable class ids.
|
||||
resolution_wh: Image resolution as ``(width, height)`` for mask
|
||||
rasterization.
|
||||
force_masks: If True, returns a mask array for every object even when
|
||||
no ``<polygon>`` element is present.
|
||||
|
||||
Returns:
|
||||
A tuple containing a Detections object and an
|
||||
updated list of class names, extended with the class names
|
||||
|
|
@ -322,7 +329,7 @@ def detections_from_xml_obj(
|
|||
# https://github.com/roboflow/supervision/issues/144
|
||||
xyxy_arr -= 1
|
||||
|
||||
for k in set(class_names):
|
||||
for k in sorted(set(class_names)):
|
||||
if k not in extended_classes:
|
||||
extended_classes.append(k)
|
||||
class_id = np.array(
|
||||
|
|
|
|||
|
|
@ -12,7 +12,10 @@ from PIL import Image
|
|||
from tqdm.auto import tqdm
|
||||
|
||||
from supervision.config import ORIENTED_BOX_COORDINATES
|
||||
from supervision.dataset.utils import approximate_mask_with_polygons
|
||||
from supervision.dataset.utils import (
|
||||
_check_no_basename_collisions,
|
||||
approximate_mask_with_polygons,
|
||||
)
|
||||
from supervision.detection.core import Detections
|
||||
from supervision.detection.utils._typing import _DetectionDataType
|
||||
from supervision.detection.utils.converters import polygon_to_mask, polygon_to_xyxy
|
||||
|
|
@ -470,6 +473,11 @@ def save_yolo_annotations(
|
|||
>>> dataset = DetectionDataset(classes=["cat"], images={}, annotations={})
|
||||
>>> save_yolo_annotations(dataset, "/tmp/labels")
|
||||
"""
|
||||
_check_no_basename_collisions(
|
||||
image_paths=dataset.image_paths,
|
||||
key=lambda image_path: _image_name_to_annotation_name(Path(image_path).name),
|
||||
output_kind="YOLO annotation",
|
||||
)
|
||||
Path(annotations_directory_path).mkdir(parents=True, exist_ok=True)
|
||||
for image_path, image, annotation in tqdm(
|
||||
dataset,
|
||||
|
|
|
|||
|
|
@ -4,6 +4,7 @@ import copy
|
|||
import os
|
||||
import random
|
||||
import shutil
|
||||
from collections.abc import Callable
|
||||
from pathlib import Path
|
||||
from typing import TYPE_CHECKING, TypeVar, cast
|
||||
|
||||
|
|
@ -126,6 +127,51 @@ def map_detections_class_id(
|
|||
return detections_copy
|
||||
|
||||
|
||||
def _check_no_basename_collisions(
|
||||
image_paths: list[str],
|
||||
key: Callable[[str], str],
|
||||
output_kind: str,
|
||||
) -> None:
|
||||
"""Raise if two image paths would be written to the same output file.
|
||||
|
||||
Dataset image paths may share a basename when they originate from different
|
||||
directories (a legal, common state after :meth:`DetectionDataset.merge`).
|
||||
Exporting them into a single flat output directory keyed on the basename or
|
||||
stem would silently overwrite one file with another and mispair images with
|
||||
their annotations. This guard detects such collisions before any file is
|
||||
written and names the colliding source paths.
|
||||
|
||||
Args:
|
||||
image_paths: The dataset image paths about to be written.
|
||||
key: Maps an image path to the output file name it would be written to.
|
||||
output_kind: Human-readable description of the output (e.g. ``"image"``
|
||||
or ``"YOLO annotation"``) used in the error message.
|
||||
|
||||
Raises:
|
||||
ValueError: If two image paths map to the same output file name.
|
||||
|
||||
Examples:
|
||||
>>> from pathlib import Path
|
||||
>>> from supervision.dataset.utils import _check_no_basename_collisions
|
||||
>>> _check_no_basename_collisions(
|
||||
... ["a/img.jpg", "b/img.jpg"], lambda p: Path(p).name, "image"
|
||||
... )
|
||||
Traceback (most recent call last):
|
||||
...
|
||||
ValueError: Cannot export dataset: image paths 'a/img.jpg' and ...
|
||||
"""
|
||||
seen: dict[str, str] = {}
|
||||
for image_path in image_paths:
|
||||
output_name = key(image_path)
|
||||
if output_name in seen:
|
||||
raise ValueError(
|
||||
f"Cannot export dataset: image paths {seen[output_name]!r} and "
|
||||
f"{image_path!r} both map to {output_kind} file {output_name!r}. "
|
||||
"Ensure all image basenames are unique before exporting."
|
||||
)
|
||||
seen[output_name] = image_path
|
||||
|
||||
|
||||
def save_dataset_images(
|
||||
dataset: DetectionDataset,
|
||||
images_directory_path: str,
|
||||
|
|
@ -149,6 +195,11 @@ def save_dataset_images(
|
|||
>>> dataset = DetectionDataset(classes=["cat"], images={}, annotations={})
|
||||
>>> save_dataset_images(dataset, "/tmp/images")
|
||||
"""
|
||||
_check_no_basename_collisions(
|
||||
image_paths=dataset.image_paths,
|
||||
key=lambda image_path: Path(image_path).name,
|
||||
output_kind="image",
|
||||
)
|
||||
Path(images_directory_path).mkdir(parents=True, exist_ok=True)
|
||||
for image_path in tqdm(
|
||||
dataset.image_paths,
|
||||
|
|
@ -179,13 +230,21 @@ def train_test_split(
|
|||
shuffle: Whether to shuffle the data before splitting.
|
||||
|
||||
Returns:
|
||||
The split data.
|
||||
"""
|
||||
if random_state is not None:
|
||||
random.seed(random_state)
|
||||
The split data. The input list is copied and never mutated.
|
||||
|
||||
Examples:
|
||||
>>> train, test = train_test_split(
|
||||
... [1, 2, 3, 4, 5], train_ratio=0.6, random_state=0
|
||||
... )
|
||||
>>> len(train), len(test)
|
||||
(3, 2)
|
||||
"""
|
||||
rng = random.Random(random_state) # noqa: S311 — dataset split, not cryptographic
|
||||
if shuffle:
|
||||
random.shuffle(data)
|
||||
data = list(data)
|
||||
rng.shuffle(data)
|
||||
else:
|
||||
data = list(data) # copy to guarantee non-mutation
|
||||
|
||||
split_index = int(len(data) * train_ratio)
|
||||
return data[:split_index], data[split_index:]
|
||||
|
|
|
|||
|
|
@ -1,5 +1,7 @@
|
|||
from contextlib import ExitStack as DoesNotRaise
|
||||
from pathlib import Path
|
||||
|
||||
import cv2
|
||||
import numpy as np
|
||||
import pytest
|
||||
from defusedxml import ElementTree
|
||||
|
|
@ -7,6 +9,7 @@ from defusedxml import ElementTree
|
|||
from supervision.dataset.formats.pascal_voc import (
|
||||
detections_from_xml_obj,
|
||||
detections_to_pascal_voc,
|
||||
load_pascal_voc_annotations,
|
||||
object_to_pascal_voc,
|
||||
parse_polygon_points,
|
||||
)
|
||||
|
|
@ -242,3 +245,58 @@ def test_detections_from_xml_obj_mixed_polygon_and_bbox_masks_aligned(
|
|||
assert detections.mask.shape == (2, 30, 30)
|
||||
assert detections.mask[0].any()
|
||||
assert not detections.mask[1].any()
|
||||
|
||||
|
||||
def _write_voc_sample(
|
||||
images_dir: Path, annotations_dir: Path, stem: str, class_names: list[str]
|
||||
) -> None:
|
||||
"""Write one VOC image plus its bbox-only XML annotation to disk."""
|
||||
cv2.imwrite(str(images_dir / f"{stem}.png"), np.zeros((20, 20, 3), dtype=np.uint8))
|
||||
objects = "".join(
|
||||
f"<object><name>{name}</name><bndbox><xmin>1</xmin><ymin>1</ymin>"
|
||||
f"<xmax>10</xmax><ymax>10</ymax></bndbox></object>"
|
||||
for name in class_names
|
||||
)
|
||||
(annotations_dir / f"{stem}.xml").write_text(f"<annotation>{objects}</annotation>")
|
||||
|
||||
|
||||
class TestLoadPascalVocDeterministicClasses:
|
||||
"""Regression tests for deterministic VOC class ordering (DAT-03)."""
|
||||
|
||||
def test_classes_sorted_within_file(self, tmp_path: Path) -> None:
|
||||
"""Class names from one file are assigned ids in sorted, stable order."""
|
||||
images_dir = tmp_path / "images"
|
||||
images_dir.mkdir()
|
||||
annotations_dir = tmp_path / "annotations"
|
||||
annotations_dir.mkdir()
|
||||
_write_voc_sample(images_dir, annotations_dir, "img", ["zebra", "ant", "mango"])
|
||||
|
||||
classes, _, _ = load_pascal_voc_annotations(
|
||||
images_directory_path=str(images_dir),
|
||||
annotations_directory_path=str(annotations_dir),
|
||||
)
|
||||
|
||||
assert classes == sorted(classes) == ["ant", "mango", "zebra"]
|
||||
|
||||
def test_repeated_loads_give_identical_class_ids(self, tmp_path: Path) -> None:
|
||||
"""Two loads of the same multi-file VOC set produce identical class ids."""
|
||||
images_dir = tmp_path / "images"
|
||||
images_dir.mkdir()
|
||||
annotations_dir = tmp_path / "annotations"
|
||||
annotations_dir.mkdir()
|
||||
_write_voc_sample(images_dir, annotations_dir, "a_img", ["zebra"])
|
||||
_write_voc_sample(images_dir, annotations_dir, "b_img", ["ant", "mango"])
|
||||
|
||||
first = load_pascal_voc_annotations(
|
||||
images_directory_path=str(images_dir),
|
||||
annotations_directory_path=str(annotations_dir),
|
||||
)
|
||||
second = load_pascal_voc_annotations(
|
||||
images_directory_path=str(images_dir),
|
||||
annotations_directory_path=str(annotations_dir),
|
||||
)
|
||||
|
||||
assert first[0] == second[0]
|
||||
assert {p: d.class_id.tolist() for p, d in first[2].items()} == {
|
||||
p: d.class_id.tolist() for p, d in second[2].items()
|
||||
}
|
||||
|
|
|
|||
|
|
@ -383,3 +383,88 @@ class TestDetectionDatasetInMemoryImages:
|
|||
ds_b = self._build_dataset({"img1.jpg": _create_image(fill_value=2)})
|
||||
|
||||
assert ds_a != ds_b
|
||||
|
||||
|
||||
class TestDetectionDatasetExportCollisions:
|
||||
"""Regression tests for the basename-collision guard on export (DAT-04)."""
|
||||
|
||||
def test_as_yolo_raises_on_same_basename_images(self, tmp_path: Path) -> None:
|
||||
"""Same-basename images from different directories must not overwrite."""
|
||||
dataset = DetectionDataset(
|
||||
classes=["cat"],
|
||||
images=["dir_a/img.png", "dir_b/img.png"],
|
||||
annotations={
|
||||
"dir_a/img.png": _create_detections(
|
||||
xyxy=[[0, 0, 10, 10]], class_id=[0]
|
||||
),
|
||||
"dir_b/img.png": _create_detections(
|
||||
xyxy=[[0, 0, 10, 10]], class_id=[0]
|
||||
),
|
||||
},
|
||||
)
|
||||
|
||||
with pytest.raises(ValueError, match="both map to image file"):
|
||||
dataset.as_yolo(images_directory_path=str(tmp_path / "images"))
|
||||
|
||||
def test_as_yolo_raises_on_same_stem_annotations(self, tmp_path: Path) -> None:
|
||||
"""Same-stem images must not overwrite annotations."""
|
||||
dataset = DetectionDataset(
|
||||
classes=["cat"],
|
||||
images=["dir_a/img.jpg", "dir_b/img.png"],
|
||||
annotations={
|
||||
"dir_a/img.jpg": _create_detections(
|
||||
xyxy=[[0, 0, 10, 10]], class_id=[0]
|
||||
),
|
||||
"dir_b/img.png": _create_detections(
|
||||
xyxy=[[0, 0, 10, 10]], class_id=[0]
|
||||
),
|
||||
},
|
||||
)
|
||||
|
||||
with pytest.raises(ValueError, match="both map to YOLO annotation file"):
|
||||
dataset.as_yolo(
|
||||
images_directory_path=str(tmp_path / "images"),
|
||||
annotations_directory_path=str(tmp_path / "labels"),
|
||||
)
|
||||
|
||||
def test_as_pascal_voc_raises_on_same_basename_images(self, tmp_path: Path) -> None:
|
||||
"""Same-basename images must not overwrite image files on export."""
|
||||
dataset = DetectionDataset(
|
||||
classes=["cat"],
|
||||
images=["dir_a/img.jpg", "dir_b/img.jpg"],
|
||||
annotations={
|
||||
"dir_a/img.jpg": _create_detections(
|
||||
xyxy=[[0, 0, 10, 10]], class_id=[0]
|
||||
),
|
||||
"dir_b/img.jpg": _create_detections(
|
||||
xyxy=[[0, 0, 10, 10]], class_id=[0]
|
||||
),
|
||||
},
|
||||
)
|
||||
|
||||
with pytest.raises(ValueError, match="both map to image file"):
|
||||
dataset.as_pascal_voc(
|
||||
images_directory_path=str(tmp_path / "images"),
|
||||
)
|
||||
|
||||
def test_as_pascal_voc_raises_on_same_stem_annotations(
|
||||
self, tmp_path: Path
|
||||
) -> None:
|
||||
"""Same-stem images must not overwrite annotations."""
|
||||
dataset = DetectionDataset(
|
||||
classes=["cat"],
|
||||
images=["dir_a/img.jpg", "dir_b/img.png"],
|
||||
annotations={
|
||||
"dir_a/img.jpg": _create_detections(
|
||||
xyxy=[[0, 0, 10, 10]], class_id=[0]
|
||||
),
|
||||
"dir_b/img.png": _create_detections(
|
||||
xyxy=[[0, 0, 10, 10]], class_id=[0]
|
||||
),
|
||||
},
|
||||
)
|
||||
|
||||
with pytest.raises(ValueError, match="both map to Pascal VOC annotation file"):
|
||||
dataset.as_pascal_voc(
|
||||
annotations_directory_path=str(tmp_path / "annotations"),
|
||||
)
|
||||
|
|
|
|||
|
|
@ -1,10 +1,13 @@
|
|||
import random
|
||||
from contextlib import ExitStack as DoesNotRaise
|
||||
from pathlib import Path
|
||||
from typing import TypeVar
|
||||
|
||||
import pytest
|
||||
|
||||
from supervision import Detections
|
||||
from supervision.dataset.utils import (
|
||||
_check_no_basename_collisions,
|
||||
build_class_index_mapping,
|
||||
map_detections_class_id,
|
||||
merge_class_lists,
|
||||
|
|
@ -229,3 +232,71 @@ def test_map_detections_class_id(
|
|||
source_to_target_mapping=source_to_target_mapping, detections=detections
|
||||
)
|
||||
assert result == expected_result
|
||||
|
||||
|
||||
class TestTrainTestSplitRngIsolation:
|
||||
"""Regression tests for train_test_split RNG isolation (DAT-02)."""
|
||||
|
||||
def test_does_not_mutate_input_list(self) -> None:
|
||||
"""split() must not reorder the caller's list in place."""
|
||||
data = list(range(10))
|
||||
original = data.copy()
|
||||
train_test_split(data=data, train_ratio=0.5, random_state=42, shuffle=True)
|
||||
assert data == original
|
||||
|
||||
def test_does_not_pollute_global_rng(self) -> None:
|
||||
"""split() must not disturb the process-global random state."""
|
||||
state_before = random.getstate()
|
||||
train_test_split(
|
||||
data=list(range(10)), train_ratio=0.5, random_state=42, shuffle=True
|
||||
)
|
||||
assert random.getstate() == state_before
|
||||
|
||||
def test_result_independent_of_global_rng(self) -> None:
|
||||
"""A fixed random_state yields the same split regardless of global RNG."""
|
||||
first = train_test_split(
|
||||
data=list(range(10)), train_ratio=0.5, random_state=42, shuffle=True
|
||||
)
|
||||
for _ in range(5):
|
||||
random.random() # noqa: S311 — perturb global RNG; split must ignore it
|
||||
second = train_test_split(
|
||||
data=list(range(10)), train_ratio=0.5, random_state=42, shuffle=True
|
||||
)
|
||||
assert first == second
|
||||
|
||||
|
||||
class TestCheckNoBasenameCollisions:
|
||||
"""Regression tests for export basename collision detection (DAT-04)."""
|
||||
|
||||
def test_empty_list_does_not_raise(self) -> None:
|
||||
"""Empty image_paths must not raise."""
|
||||
_check_no_basename_collisions(
|
||||
image_paths=[],
|
||||
key=lambda image_path: Path(image_path).name,
|
||||
output_kind="image",
|
||||
)
|
||||
|
||||
def test_single_path_does_not_raise(self) -> None:
|
||||
"""Single image path cannot collide; must not raise."""
|
||||
_check_no_basename_collisions(
|
||||
image_paths=["a/img.jpg"],
|
||||
key=lambda image_path: Path(image_path).name,
|
||||
output_kind="image",
|
||||
)
|
||||
|
||||
def test_raises_on_colliding_output_names(self) -> None:
|
||||
"""Two source paths mapping to one output name must raise ValueError."""
|
||||
with pytest.raises(ValueError, match="both map to image file"):
|
||||
_check_no_basename_collisions(
|
||||
image_paths=["a/img.jpg", "b/img.jpg"],
|
||||
key=lambda image_path: Path(image_path).name,
|
||||
output_kind="image",
|
||||
)
|
||||
|
||||
def test_passes_on_unique_output_names(self) -> None:
|
||||
"""Distinct output names must not raise."""
|
||||
_check_no_basename_collisions(
|
||||
image_paths=["a/img1.jpg", "b/img2.jpg"],
|
||||
key=lambda image_path: Path(image_path).name,
|
||||
output_kind="image",
|
||||
)
|
||||
|
|
|
|||
Loading…
Reference in New Issue