EwaldandClaude Sonnet 5
019a5c9172
Defining NODE_CLASS_MAPPINGS made ComfyUI skip comfy_entrypoint, so /imageops/viewmedia and node replacements were never registered. Load as a V3 extension only, move server.py to routes.py so it stops shadowing ComfyUI's server module, and transcode animated previews with PyAV instead of an external ffmpeg. - Require ComfyUI 0.19.0 (has_intermediate_output); CI floor updated. - Process 11 nodes on ComfyUI's torch device when the frames fit in free memory (IMAGEOPS_COMPUTE_DEVICE=cpu disables it). - Add core CURVE input to Color Correct, BOUNDING_BOX output/input to Crop and Crop Stitch, LAYERS output to Comp (ComfyUI 0.31+), essentials_category, and Merge blend modes with golden parity. - Report progress through the public set_progress API and drop the hidden unique_id input. - Add ImageOps settings and a refresh-previews command. - Fix ImageOpsComp rotation and preview saving (undefined to_display_range) and masked Transform edge colors. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
93 lines
2.9 KiB
Python
93 lines
2.9 KiB
Python
"""Unit tests for ImageOpsCrop including mask synchronization."""
|
|
from __future__ import annotations
|
|
|
|
import pytest
|
|
import torch
|
|
|
|
from nodes.crop import ImageOpsCrop
|
|
|
|
|
|
def _unwrap(result):
|
|
if isinstance(result, dict):
|
|
return result["result"]
|
|
return result
|
|
|
|
|
|
class TestImageOpsCrop:
|
|
def test_crop_with_mask_synchronized(self):
|
|
node = ImageOpsCrop()
|
|
image = torch.zeros(1, 100, 100, 3, dtype=torch.float32)
|
|
# Put distinct values in quadrant
|
|
image[0, :50, :50, 0] = 1.0
|
|
|
|
mask = torch.zeros(1, 100, 100, dtype=torch.float32)
|
|
mask[0, :50, :50] = 1.0
|
|
|
|
result = _unwrap(node.execute(
|
|
image=image,
|
|
mask=mask,
|
|
aspect_ratio="1:1",
|
|
width=50,
|
|
height=50,
|
|
crop_center_x=0.25,
|
|
crop_center_y=0.25,
|
|
crop_scale=0.5,
|
|
))
|
|
|
|
out_image = result[0]
|
|
out_mask = result[1]
|
|
|
|
assert out_image.shape == (1, 50, 50, 3)
|
|
assert out_mask.shape == (1, 50, 50)
|
|
# Cropped region should capture the white area in both image and mask
|
|
assert (out_image[..., 0] > 0.9).all()
|
|
assert (out_mask > 0.9).all()
|
|
|
|
def test_crop_bypass(self):
|
|
node = ImageOpsCrop()
|
|
image = torch.rand(1, 80, 80, 3, dtype=torch.float32)
|
|
mask = torch.rand(1, 80, 80, dtype=torch.float32)
|
|
|
|
result = _unwrap(node.execute(image=image, mask=mask, bypass=True))
|
|
out_image = result[0]
|
|
out_mask = result[1]
|
|
|
|
assert torch.allclose(out_image, image)
|
|
assert torch.allclose(out_mask, mask)
|
|
|
|
|
|
class TestCropBoundingBox:
|
|
def _crop(self):
|
|
image = torch.rand(1, 100, 120, 3, dtype=torch.float32)
|
|
result = _unwrap(ImageOpsCrop().execute(
|
|
image=image,
|
|
aspect_ratio="1:1",
|
|
width=40,
|
|
height=40,
|
|
crop_center_x=0.25,
|
|
crop_center_y=0.75,
|
|
crop_scale=0.5,
|
|
))
|
|
return image, result
|
|
|
|
def test_bounding_box_output_is_core_bbox_and_appended_last(self):
|
|
_, result = self._crop()
|
|
|
|
assert [type(item).__name__ for item in result[:3]] == ["Tensor", "Tensor", "dict"]
|
|
assert set(result[3]) == {"x", "y", "width", "height"}
|
|
frame = result[2]["bbox"]
|
|
assert result[3] == {key: frame[key] for key in ("x", "y", "width", "height")}
|
|
assert all(isinstance(value, int) for value in result[3].values())
|
|
|
|
def test_crop_stitch_accepts_core_bounding_box(self):
|
|
from nodes.crop_stitch import ImageOpsCropStitch
|
|
|
|
image, result = self._crop()
|
|
patch = torch.ones_like(result[0])
|
|
|
|
with_payload = _unwrap(ImageOpsCropStitch().execute(original=image, crop=patch, crop_bbox=result[2]))
|
|
with_core_bbox = _unwrap(ImageOpsCropStitch().execute(original=image, crop=patch, bounding_box=result[3]))
|
|
|
|
assert torch.allclose(with_payload[0], with_core_bbox[0])
|
|
assert not torch.allclose(with_core_bbox[0], image)
|