feat: Use image color when add layer to napari (#1200)

Use color associated with the channel in napari reader plugin.

<!-- Generated by sourcery-ai[bot]: start summary -->

## Summary by Sourcery

Add functionality to use channel-associated colors in napari layers by
adjusting hex color codes. Enhance the `_hex_to_rgb` function to support
additional hex code formats and update tests accordingly.

New Features:
- Integrate color adjustment for image layers in the napari reader
plugin, using the color associated with the channel.

Enhancements:
- Improve the `_hex_to_rgb` function to handle 4 and 8 character hex
codes, ensuring proper conversion and validation.

Tests:
- Expand tests for the `_hex_to_rgb` function to include cases for 4 and
8 character hex codes, ensuring correct functionality and error
handling.

<!-- Generated by sourcery-ai[bot]: end summary -->

<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit

- **New Features**
- Introduced `adjust_color` function to modify color strings based on
length, enhancing color handling.
- Updated `_hex_to_rgb` function to support additional hexadecimal color
formats.

- **Bug Fixes**
- Improved validation for hex color codes, ensuring robust error
handling for various input lengths.

- **Tests**
- Added new test cases for `_hex_to_rgb` to validate behavior with hex
codes including alpha channels.
- Enhanced tests for layer creation and colormap validation in image
processing.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->

package/PartSeg/common_gui/advanced_tabs.py

@@ -275,4 +275,8 @@ def __init__(self, settings: PartSettings, parent=None):
 
     def update_metadata(self):
         self._dict_viewer.set_data(self.settings.image.metadata)
-        self.channel_info.setText(f"Channels: {self.settings.image.channel_names}")
+        text = ", ".join(
+            f"{name}: {color}"
+            for name, color in zip(self.settings.image.channel_names, self.settings.image.get_colors())
+        )
+        self.channel_info.setText(f"Channels with colors: {text}")

package/PartSegCore/napari_plugins/loader.py

@@ -1,10 +1,54 @@
 import typing
+from importlib.metadata import version
 
 import numpy as np
+from packaging.version import parse as parse_version
 
 from PartSegCore.analysis import ProjectTuple
 from PartSegCore.io_utils import LoadBase, WrongFileTypeException
 from PartSegCore.mask.io_functions import MaskProjectTuple
+from PartSegImage import Image
+
+
+@typing.overload
+def adjust_color(color: str) -> str: ...
+
+
+@typing.overload
+def adjust_color(color: typing.List[int]) -> typing.List[float]: ...
+
+
+def adjust_color(color: typing.Union[str, typing.List[int]]) -> typing.Union[str, typing.List[float]]:
+    # as napari ignore alpha channel in color, and adding it to
+    # color cause that napari fails to detect that such colormap is already present
+    # in this function I remove alpha channel if it is present
+    if isinstance(color, str) and color.startswith("#"):
+        if len(color) == 9:
+            # case when color is in format #RRGGBBAA
+            return color[:7]
+        if len(color) == 5:
+            # case when color is in format #RGBA
+            return color[:4]
+    elif isinstance(color, list):
+        return [color[i] / 255 for i in range(3)]
+    # If not fit to an earlier case, return as is.
+    # Maybe napari will handle it
+    return color
+
+
+if parse_version(version("napari")) >= parse_version("0.4.19a1"):
+
+    def add_color(image: Image, idx: int) -> dict:
+        return {
+            "colormap": adjust_color(image.get_colors()[idx]),
+        }
+
+else:
+
+    def add_color(image: Image, idx: int) -> dict:  # noqa: ARG001
+        # Do nothing, as napari is not able to pass hex color to image
+        # the image and idx are present to keep the same signature
+        return {}
 
 
 def _image_to_layers(project_info, scale, translate):
@@ -27,6 +71,7 @@ def _image_to_layers(project_info, scale, translate):
                     "blending": "additive",
                     "translate": translate,
                     "metadata": project_info.image.metadata,
+                    **add_color(project_info.image, i),
                 },
                 "image",
             )

package/PartSegImage/image.py

@@ -393,6 +393,7 @@ def merge(self, image: Image, axis: str) -> Image:
                 self._channel_arrays + [self.reorder_axes(x, image.array_axis_order) for x in image._channel_arrays]
             )
             channel_names = self._merge_channel_names(self.channel_names, image.channel_names)
+            color_map = self.default_coloring + image.default_coloring
         else:
             index = self.array_axis_order.index(axis)
             data = self._image_data_normalize(
@@ -402,8 +403,11 @@ def merge(self, image: Image, axis: str) -> Image:
                 ]
             )
             channel_names = self.channel_names
+            color_map = self.default_coloring
 
-        return self.substitute(data=data, ranges=self.ranges + image.ranges, channel_names=channel_names)
+        return self.substitute(
+            data=data, ranges=self.ranges + image.ranges, channel_names=channel_names, default_coloring=color_map
+        )
 
     @property
     def channel_names(self) -> list[str]:
@@ -494,7 +498,7 @@ def substitute(
         channel_names = self.channel_names if channel_names is None else channel_names
 
         channel_info = [
-            ChannelInfoFull(name=name, color_map=color, contrast_limits=contrast_limits)
+            ChannelInfo(name=name, color_map=color, contrast_limits=contrast_limits)
             for name, color, contrast_limits in zip_longest(channel_names, default_coloring, ranges)
         ]
 
@@ -973,9 +977,9 @@ def _hex_to_rgb(hex_code: str) -> tuple[int, int, int]:
     """
     hex_code = hex_code.lstrip("#")
 
-    if len(hex_code) == 3:
+    if len(hex_code) in {3, 4}:
         hex_code = "".join([c * 2 for c in hex_code])
-    elif len(hex_code) != 6:
+    elif len(hex_code) not in {6, 8}:
         raise ValueError(f"Invalid hex code format: {hex_code}")
 
     return int(hex_code[:2], 16), int(hex_code[2:4], 16), int(hex_code[4:6], 16)

package/tests/conftest.py

@@ -25,7 +25,7 @@
 from PartSegCore.roi_info import ROIInfo
 from PartSegCore.segmentation.restartable_segmentation_algorithms import BorderRim, LowerThresholdAlgorithm
 from PartSegCore.segmentation.segmentation_algorithm import ThresholdAlgorithm
-from PartSegImage import Image
+from PartSegImage import ChannelInfo, Image
 
 
 @pytest.fixture(scope="module")
@@ -69,20 +69,33 @@ def image2(image, tmp_path):
 def image2d(tmp_path):
     data = np.zeros([20, 20], dtype=np.uint8)
     data[10:-1, 1:-1] = 20
-    return Image(data, (10**-3, 10**-3), axes_order="YX", file_path=str(tmp_path / "test.tiff"))
+    return Image(data, spacing=(10**-3, 10**-3), axes_order="YX", file_path=str(tmp_path / "test.tiff"))
 
 
 @pytest.fixture
 def stack_image():
-    data = np.zeros([20, 40, 40, 2], dtype=np.uint8)
+    data = np.zeros([20, 40, 40, 3], dtype=np.uint8)
     for x, y in itertools.product([0, 20], repeat=2):
         data[1:-1, x + 2 : x + 18, y + 2 : y + 18] = 100
     for x, y in itertools.product([0, 20], repeat=2):
         data[3:-3, x + 4 : x + 16, y + 4 : y + 16] = 120
     for x, y in itertools.product([0, 20], repeat=2):
         data[5:-5, x + 6 : x + 14, y + 6 : y + 14] = 140
 
-    return MaskProjectTuple("test_path", Image(data, (2, 1, 1), axes_order="ZYXC", file_path="test_path"))
+    return MaskProjectTuple(
+        "test_path",
+        Image(
+            data,
+            spacing=(2, 1, 1),
+            axes_order="ZYXC",
+            file_path="test_path",
+            channel_info=[
+                ChannelInfo(name="channel 1", color_map="#00FF00FF"),
+                ChannelInfo(name="channel 2", color_map="#00FF"),
+                ChannelInfo(name="channel 3", color_map="#FF0000"),
+            ],
+        ),
+    )
 
 
 @pytest.fixture

package/tests/test_PartSeg/test_viewer.py

@@ -75,13 +75,13 @@ def test_base(self, image, analysis_segmentation2, tmp_path):
         assert len(viewer.layers) == 2
         settings.image = analysis_segmentation2.image
         viewer.create_initial_layers(True, True, True, True)
-        assert len(viewer.layers) == 2
+        assert len(viewer.layers) == 3
         settings.roi = analysis_segmentation2.roi_info.roi
         viewer.create_initial_layers(True, True, True, True)
-        assert len(viewer.layers) == 3
+        assert len(viewer.layers) == 4
         settings.mask = analysis_segmentation2.mask
         viewer.create_initial_layers(True, True, True, True)
-        assert len(viewer.layers) == 4
+        assert len(viewer.layers) == 5
         viewer.close()
 
     def test_points(self, image, tmp_path, qtbot):

package/tests/test_PartSegCore/test_napari_plugins.py

@@ -1,10 +1,12 @@
 # pylint: disable=no-self-use
 
 import os
+from importlib.metadata import version
 
 import numpy as np
 import pytest
 from napari.layers import Image, Labels, Layer
+from packaging.version import parse as parse_version
 
 from PartSegCore.analysis import ProjectTuple
 from PartSegCore.mask.io_functions import LoadROIFromTIFF
@@ -27,20 +29,39 @@
 def test_project_to_layers_analysis(analysis_segmentation):
     analysis_segmentation.roi_info.alternative["test"] = np.zeros(analysis_segmentation.image.shape, dtype=np.uint8)
     res = project_to_layers(analysis_segmentation)
-    assert len(res) == 4
+    assert len(res) == 5
     l1 = Layer.create(*res[0])
     assert isinstance(l1, Image)
     assert l1.name == "channel 1"
     assert np.allclose(l1.scale[1:] / 1e9, analysis_segmentation.image.spacing)
-    l2 = Layer.create(*res[2])
-    assert isinstance(l2, Labels)
-    assert l2.name == "ROI"
-    assert np.allclose(l2.scale[1:] / 1e9, analysis_segmentation.image.spacing)
     l3 = Layer.create(*res[3])
     assert isinstance(l3, Labels)
-    assert l3.name == "test"
+    assert l3.name == "ROI"
     assert np.allclose(l3.scale[1:] / 1e9, analysis_segmentation.image.spacing)
-    assert not l3.visible
+    l4 = Layer.create(*res[4])
+    assert isinstance(l4, Labels)
+    assert l4.name == "test"
+    assert np.allclose(l4.scale[1:] / 1e9, analysis_segmentation.image.spacing)
+    assert not l4.visible
+
+
+@pytest.mark.skipif(
+    parse_version(version("napari")) < parse_version("0.4.19a16"), reason="not supported by old napari versions"
+)
+def test_passing_colormap(analysis_segmentation):
+    res = project_to_layers(analysis_segmentation)
+    l1 = Layer.create(*res[0])
+    assert isinstance(l1, Image)
+    assert l1.name == "channel 1"
+    assert l1.colormap.name == "green"
+    l2 = Layer.create(*res[1])
+    assert isinstance(l2, Image)
+    assert l2.name == "channel 2"
+    assert l2.colormap.name == "blue"
+    l2 = Layer.create(*res[2])
+    assert isinstance(l2, Image)
+    assert l2.name == "channel 3"
+    assert l2.colormap.name == "red"
 
 
 def test_project_to_layers_roi():
@@ -55,7 +76,7 @@ def test_project_to_layers_roi():
 
 def test_project_to_layers_mask(stack_segmentation1):
     res = project_to_layers(stack_segmentation1)
-    assert len(res) == 3
+    assert len(res) == 4
     assert res[0][2] == "image"
 
 

package/tests/test_PartSegImage/test_image.py

@@ -700,11 +700,13 @@ def test_merge_channel_props_with_none(channel_name, default_coloring, ranges):
 
 def test_hex_to_rgb():
     assert _hex_to_rgb("#ff0000") == (255, 0, 0)
+    assert _hex_to_rgb("#ff0000ff") == (255, 0, 0)
     assert _hex_to_rgb("#00FF00") == (0, 255, 0)
     assert _hex_to_rgb("#b00") == (187, 0, 0)
+    assert _hex_to_rgb("#b00f") == (187, 0, 0)
     assert _hex_to_rgb("#B00") == (187, 0, 0)
     with pytest.raises(ValueError, match="Invalid hex code format"):
-        _hex_to_rgb("#b000")
+        _hex_to_rgb("#b0000")
 
 
 def test_name_to_rgb():