Tkakar/CAT-1100-add-container-for-reading-image-metadata (#145)

* Added container

* Added workflow

* Fixed container name in readme

containers/ome-tiff-metadata/Dockerfile

@@ -0,0 +1,26 @@
+# We just need to use --file to point at it, instead of assuming it is in context.
+
+# Using Conda because pyarrow did not install easily on python base images.
+FROM --platform=linux/amd64 continuumio/miniconda3:24.5.0-0
+
+# For tiff packages
+RUN apt-get --allow-releaseinfo-change update &&\
+      apt-get install -y gcc python3-dev libhdf5-dev pkg-config python3-numcodecs
+
+RUN pip install --upgrade pip setuptools
+
+COPY requirements-freeze.txt .
+RUN pip install -r ./requirements-freeze.txt
+
+# In development, you may want to pin a single dependency in requirements.txt,
+# without throwing away the entire cache layer from requirements-freeze.txt.
+# (But once it works, you should check in an updated freeze!)
+
+COPY requirements.txt .
+RUN pip install -r ./requirements.txt
+
+COPY . .
+
+CMD [ "python", "main.py", \
+      "--input_dir", "/input", \
+      "--output_dir", "/output" ]

containers/ome-tiff-metadata/README.md

@@ -0,0 +1,25 @@
+# ome-tiff-metadata
+
+This docker container creates a JSON object for metadata (physical sizes and units) extracted from each TIFF within an input directory. This is needed particularly for laying segmentation masks over base-images when there is a misalignment between the physical sizes fo both images. Note that, even with same pixel sizes, if physical sizes are different, misalignment can occur. This metadata is useful to add scaling to the segmentation mask when visualized using Vitessce.
+
+## Input
+
+The input to the container is one or more ome-tiff image files.
+
+## Output
+
+The output is a json file that includes an object with following structure for every input ome-tiff image.
+
+```
+{"PhysicalSizeX": 0.5, "PhysicalSizeY": 0.5, "PhysicalSizeUnitX": "\u00b5m", "PhysicalSizeUnitY": "\u00b5m"}
+
+```
+
+## Normalization
+
+None
+
+## Example
+
+Example of a hubmap dataset using this container for metadata generation for Vitessce (visualization) would be
+`TODO`.

containers/ome-tiff-metadata/VERSION

@@ -0,0 +1 @@
+0.0.1

containers/ome-tiff-metadata/context/main.py

@@ -0,0 +1,72 @@
+import argparse
+from glob import glob
+from pathlib import Path
+from os import makedirs
+from itertools import chain
+import json
+from ome_types import from_tiff
+
+
+def get_metadata(tiff_file_path):
+    extracted_metadata = {}
+    try:
+        ome_data = from_tiff(tiff_file_path)
+        image = ome_data.images[0]
+        pixels = image.pixels
+
+        physical_size_x = pixels.physical_size_x
+        physical_size_y = pixels.physical_size_y
+
+        physical_size_unit_x = pixels.physical_size_x_unit  
+        physical_size_unit_y = pixels.physical_size_y_unit  
+
+
+        physical_size_unit_x = physical_size_unit_x.value if hasattr(physical_size_unit_x, 'value') else physical_size_unit_x
+        physical_size_unit_y = physical_size_unit_y.value if hasattr(physical_size_unit_y, 'value') else physical_size_unit_y
+        extracted_metadata = {
+            "PhysicalSizeX": physical_size_x,
+            "PhysicalSizeY": physical_size_y,
+            "PhysicalSizeUnitX": physical_size_unit_x, 
+            "PhysicalSizeUnitY": physical_size_unit_y, 
+        }
+        print(f"Extracted metadata: {extracted_metadata}")
+    except FileNotFoundError:
+        print(f"Error: The file {tiff_file_path} does not exist.")
+    except IndexError:
+        print("Error: The TIFF file does not contain any images.")
+    except Exception as e:
+        print(f"An unexpected error occurred: {e}")
+
+    return extracted_metadata
+
+
+def main(input_dir, output_dir):
+    makedirs(output_dir, exist_ok=True)
+
+    # Find all OME.TIFFs in the input directory.
+    tiffs = list(chain(input_dir.glob('**/*.ome.tif'), input_dir.glob('**/*.ome.tiff')))
+    if not tiffs:
+        raise Exception(f'No OME TIFFs found in {input_dir}')
+    for input_path in tiffs:
+        metadata = get_metadata(str(input_path))
+
+        # Create output path for each OME.TIFF:
+        new_output_dir = (output_dir / input_path.relative_to(input_dir)).parent
+        new_output_dir.mkdir(parents=True, exist_ok=True)
+
+        # Set output filename for JSON file and dump to disk:
+        output_path = str(output_dir / input_path.relative_to(input_dir).with_suffix('').with_suffix(''))+'.metadata.json'
+        with open(output_path, 'w') as f:
+            f.write(json.dumps(metadata))
+
+if __name__ == '__main__':
+    parser = argparse.ArgumentParser(
+        description='Create a json file for OME-TIFF offsets')
+    parser.add_argument(
+        '--input_dir', required=True, type=Path,
+        help='Directory containing ome-tiff files to read')
+    parser.add_argument(
+        '--output_dir', required=True, type=Path,
+        help='Directory where ome-tiff offsets should be written')
+    args = parser.parse_args()
+    main(args.input_dir, args.output_dir)

containers/ome-tiff-metadata/context/requirements-freeze.txt

@@ -0,0 +1,42 @@
+anaconda-anon-usage==0.4.4
+annotated-types==0.7.0
+archspec==0.2.3
+boltons==23.0.0
+Brotli==1.0.9
+certifi==2024.6.2
+cffi==1.16.0
+charset-normalizer==2.0.4
+conda==24.5.0
+conda-content-trust==0.2.0
+conda-libmamba-solver==24.1.0
+conda-package-handling==2.3.0
+conda_package_streaming==0.10.0
+cryptography==42.0.5
+distro==1.9.0
+frozendict==2.4.2
+idna==3.7
+jsonpatch==1.33
+jsonpointer==2.1
+libmambapy==1.5.8
+menuinst==2.1.1
+ome-types==0.5.2
+packaging==23.2
+pip==24.2
+platformdirs==3.10.0
+pluggy==1.0.0
+pycosat==0.6.6
+pycparser==2.21
+pydantic==2.10.5
+pydantic-compat==0.1.2
+pydantic_core==2.27.2
+PySocks==1.7.1
+requests==2.32.2
+ruamel.yaml==0.17.21
+setuptools==75.1.0
+tqdm==4.66.4
+truststore==0.8.0
+typing_extensions==4.12.2
+urllib3==2.2.2
+wheel==0.43.0
+xsdata==24.3.1
+zstandard==0.22.0

containers/ome-tiff-metadata/context/requirements.txt

@@ -0,0 +1 @@
+ome-types==0.5.2

containers/ome-tiff-metadata/test-output-expected/test_image.metadata.json

@@ -0,0 +1 @@
+{"PhysicalSizeX": 0.5, "PhysicalSizeY": 0.5, "PhysicalSizeUnitX": "\u00b5m", "PhysicalSizeUnitY": "\u00b5m"}

ome-tiff-metadata-manifest.json

@@ -0,0 +1,7 @@
+[
+  {
+    "pattern": "image_metadata/(.*)\\.metadata\\.json",
+    "description": "JSON object containing metadata about image physical size extracted from TIFF file to help in scaling for visualization.",
+    "edam_ontology_term": "EDAM_1.24.format_3464"
+  }
+]

ome-tiff-metadata.cwl

@@ -0,0 +1,19 @@
+#!/usr/bin/env cwl-runner
+
+cwlVersion: v1.2
+class: CommandLineTool
+# TODO: Make main.py executable?
+baseCommand: ['python', '/main.py', '--output_dir', 'image_metadata', '--input_dir']
+hints:
+  DockerRequirement:
+    dockerPull: hubmap/portal-container-ome-tiff-metadata:0.0.1
+inputs:
+  input_directory:
+    type: Directory
+    inputBinding:
+        position: 6
+outputs:
+  json:
+    type: Directory
+    outputBinding:
+      glob: image_metadata

workflows/ome-tiff-metadata/README.md

@@ -0,0 +1 @@
+See [containers/ome-tiff-metadata](https://github.com/hubmapconsortium/portal-containers/blob/master/containers/ome-tiff-metadata/README.md).

workflows/ome-tiff-metadata/test-job.yml

@@ -0,0 +1,3 @@
+input_directory:
+  class: Directory
+  path: test-input

workflows/ome-tiff-metadata/test-output-expected/image_metadata/test_image.metadata.json

@@ -0,0 +1 @@
+{"PhysicalSizeX": 0.5, "PhysicalSizeY": 0.5, "PhysicalSizeUnitX": "\u00b5m", "PhysicalSizeUnitY": "\u00b5m"}