init datamate

runtime/ops/formatter/slide_formatter/__init__.py

@@ -0,0 +1,6 @@
+# -*- coding: utf-8 -*-
+
+from datamate.core.base_op import OPERATORS
+
+OPERATORS.register_module(module_name='SlideFormatter',
+                          module_path="ops.formatter.slide_formatter.process")

runtime/ops/formatter/slide_formatter/metadata.yml

@@ -0,0 +1,16 @@
+name: '病理图片内容抽取'
+name_en: 'Pathology Image Content Extraction'
+description: '解析病理图片。'
+description_en: 'Analyze pathological images.'
+language: 'python'
+vendor: 'huawei'
+raw_id: 'SlideFormatter'
+version: '1.0.0'
+types:
+  - 'collect'
+modal: 'image'
+effect:
+  before: ''
+  after: ''
+inputs: 'image'
+outputs: 'image'

runtime/ops/formatter/slide_formatter/process.py

@@ -0,0 +1,36 @@
+# -- encoding: utf-8 --
+
+"""
+Description: 医疗图片解析载入
+Create: 2025/02/08 11:00
+"""
+
+import time
+from typing import Dict, Any
+
+from loguru import logger
+
+from datamate.core.base_op import Mapper
+
+
+class SlideFormatter(Mapper):
+
+    def __init__(self, *args, **kwargs):
+        super(SlideFormatter, self).__init__(*args, **kwargs)
+
+    def execute(self, sample: Dict[str, Any]) -> Dict[str, Any]:
+        '''
+        Read medical image and corresponding mask file, each as Image type in COntent value. Return Content.
+        '''
+        start = time.time()
+        file_type = sample[self.filetype_key]
+        types_openslide = ['svs', 'tif', 'dcm', 'vms', 'vmu',
+                           'ndpi', 'scn', 'mrxs', 'tiff', 'svslide',
+                           'bif', 'czi', 'sdpc']
+        if file_type not in types_openslide:
+            raise TypeError(f"Format not supported: {file_type}. Supported formats are: {', '.join(types_openslide)}.")
+
+        file_name = sample[self.filename_key]
+        logger.info(f"fileName: {file_name}, method: SlideFormatter costs {(time.time() - start):6f} s")
+        # Not really loading the slide, instead, use path as lazy loading.
+        return sample