datamint 1.7.2__tar.gz → 1.7.4__tar.gz

{datamint-1.7.2 → datamint-1.7.4}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.3
 Name: datamint
-Version: 1.7.2
+Version: 1.7.4
 Summary: A library for interacting with the Datamint API, designed for efficient data management, processing and Deep Learning workflows.
 Requires-Python: >=3.10
 Classifier: Programming Language :: Python :: 3
@@ -19,7 +19,7 @@ Requires-Dist: humanize (>=4.0.0,<5.0.0)
 Requires-Dist: lazy-loader (>=0.3.0)
 Requires-Dist: lightning
 Requires-Dist: matplotlib
-Requires-Dist: medimgkit
+Requires-Dist: medimgkit (>=0.2.0)
 Requires-Dist: nest-asyncio (>=1.0.0,<2.0.0)
 Requires-Dist: nibabel (>=4.0.0)
 Requires-Dist: numpy

{datamint-1.7.2 → datamint-1.7.4}/datamint/apihandler/base_api_handler.py

@@ -138,7 +138,7 @@ class BaseAPIHandler:
             if isinstance(data, aiohttp.FormData):  # Check if it's aiohttp.FormData
                 # Handle FormData by extracting fields
                 form_parts = []
-                for options,headers,value in data._fields:
+                for options, headers, value in data._fields:
                     # get the name from options
                     name = options.get('name', 'file')
                     if hasattr(value, 'read'):  # File-like object
@@ -161,7 +161,7 @@ class BaseAPIHandler:
         if session is None:
             async with aiohttp.ClientSession() as s:
                 return await self._run_request_async(request_args, s, data_to_get)
-            
+
         async with self.semaphore:
             try:
                 _LOGGER.debug(f"Running request to {request_args['url']}")
@@ -222,7 +222,7 @@ class BaseAPIHandler:
 
     def _run_request(self,
                      request_args: dict,
-                     session: Session = None):
+                     session: Session | None = None):
         if session is None:
             with Session() as s:
                 return self._run_request(request_args, s)
@@ -281,24 +281,25 @@ class BaseAPIHandler:
     @staticmethod
     def convert_format(bytes_array: bytes,
                        mimetype: str,
-                       file_path: str = None
+                       file_path: str | None = None
                        ) -> pydicom.dataset.Dataset | Image.Image | cv2.VideoCapture | bytes | nib_FileBasedImage:
+        """ Convert the bytes array to the appropriate format based on the mimetype."""
         content_io = BytesIO(bytes_array)
-        if mimetype == 'application/dicom':
+        if mimetype.endswith('/dicom'):
             return pydicom.dcmread(content_io)
-        elif mimetype in ('image/jpeg', 'image/png', 'image/tiff'):
+        elif mimetype.startswith('image/'):
             return Image.open(content_io)
-        elif mimetype == 'video/mp4':
+        elif mimetype.startswith('video/'):
             if file_path is None:
-                raise NotImplementedError("file_path=None is not implemented yet for video/mp4.")
+                raise NotImplementedError("file_path=None is not implemented yet for video/* mimetypes.")
             return cv2.VideoCapture(file_path)
         elif mimetype == 'application/json':
             return json.loads(bytes_array)
         elif mimetype == 'application/octet-stream':
             return bytes_array
-        elif mimetype == 'application/nifti':
+        elif mimetype.endswith('nifti'):
             if file_path is None:
-                raise NotImplementedError("file_path=None is not implemented yet for application/nifti.")
+                raise NotImplementedError(f"file_path=None is not implemented yet for {mimetype}.")
             return nib.load(file_path)
 
         raise ValueError(f"Unsupported mimetype: {mimetype}")

{datamint-1.7.2 → datamint-1.7.4}/datamint/apihandler/root_api_handler.py

@@ -8,10 +8,11 @@ import asyncio
 import aiohttp
 from medimgkit.dicom_utils import anonymize_dicom, to_bytesio, is_dicom
 from medimgkit import dicom_utils
+from medimgkit.io_utils import is_io_object
+from medimgkit.format_detection import guess_type, guess_extension
 import pydicom
 from pathlib import Path
 from datetime import date
-import mimetypes
 from PIL import Image
 import cv2
 from nibabel.filebasedimages import FileBasedImage as nib_FileBasedImage
@@ -26,13 +27,6 @@ _LOGGER = logging.getLogger(__name__)
 _USER_LOGGER = logging.getLogger('user_logger')
 
 
-def _is_io_object(obj):
-    """
-    Check if an object is a file-like object.
-    """
-    return callable(getattr(obj, "read", None))
-
-
 def _infinite_gen(x):
     while True:
         yield x
@@ -65,7 +59,7 @@ class RootAPIHandler(BaseAPIHandler):
                                             publish: bool = False,
                                             metadata_file: Optional[str | dict] = None,
                                             ) -> str:
-        if _is_io_object(file_path):
+        if is_io_object(file_path):
             name = file_path.name
         else:
             name = file_path
@@ -91,15 +85,12 @@ class RootAPIHandler(BaseAPIHandler):
             name = new_file_path
             _LOGGER.debug(f"New file path: {name}")
 
-        if mimetype is None:
-            mimetype = mimetypes.guess_type(name)[0]
         is_a_dicom_file = None
         if mimetype is None:
-            is_a_dicom_file = is_dicom(name) or is_dicom(file_path)
-            if is_a_dicom_file:
-                mimetype = 'application/dicom'
-            elif name.endswith('.nii') or name.endswith('.nii.gz'):
-                mimetype = 'application/x-nifti'
+            mimetype, _ = guess_type(file_path, use_magic=True)
+            if mimetype == 'application/gzip' and name.lower().endswith('nii.gz'):
+                # Special case for gzipped NIfTI files
+                mimetype = 'image/x.nifti'
 
         filename = os.path.basename(name)
         _LOGGER.debug(f"File name '{filename}' mimetype: {mimetype}")
@@ -239,14 +230,14 @@ class RootAPIHandler(BaseAPIHandler):
                         names = _infinite_gen(names)
                     frame_indices = segfiles.get('frame_index', _infinite_gen(None))
                     for f, name, frame_index in tqdm(zip(fpaths, names, frame_indices),
-                                                        desc=f"Uploading segmentations for {file_path}",
-                                                        total=len(fpaths)):
+                                                     desc=f"Uploading segmentations for {file_path}",
+                                                     total=len(fpaths)):
                         if f is not None:
                             await self._upload_segmentations_async(rid,
-                                                                    file_path=f,
-                                                                    name=name,
-                                                                    frame_index=frame_index,
-                                                                    transpose_segmentation=transpose_segmentation)
+                                                                   file_path=f,
+                                                                   name=name,
+                                                                   frame_index=frame_index,
+                                                                   transpose_segmentation=transpose_segmentation)
                 return rid
 
             tasks = [__upload_single_resource(f, segfiles, metadata_file)
@@ -368,25 +359,26 @@ class RootAPIHandler(BaseAPIHandler):
     def _is_dicom_report(file_path: str | IO) -> bool:
         """
         Check if a DICOM file is a report (e.g., Structured Report).
-        
+
         Args:
             file_path: Path to the DICOM file or file-like object.
-            
+
         Returns:
             bool: True if the DICOM file is a report, False otherwise.
         """
         try:
             if not is_dicom(file_path):
                 return False
-            
+
             ds = pydicom.dcmread(file_path, stop_before_pixels=True)
             if hasattr(file_path, 'seek'):
                 file_path.seek(0)
             modality = getattr(ds, 'Modality', None)
-            
+
             # Common report modalities
-            report_modalities = {'SR', 'DOC', 'KO', 'PR', 'ESR'}  # SR=Structured Report, DOC=Document, KO=Key Object, PR=Presentation State
-            
+            # SR=Structured Report, DOC=Document, KO=Key Object, PR=Presentation State
+            report_modalities = {'SR', 'DOC', 'KO', 'PR', 'ESR'}
+
             return modality in report_modalities
         except Exception as e:
             _LOGGER.warning(f"Error checking if DICOM is a report: {e}")
@@ -445,21 +437,18 @@ class RootAPIHandler(BaseAPIHandler):
             list[str | Exception]: A list of resource IDs or errors.
         """
 
-        if discard_dicom_reports:
-            if isinstance(files_path, (str, Path)):
-                files_path = [files_path]
-            elif isinstance(files_path, pydicom.dataset.Dataset):
-                files_path = [files_path]
-
-            old_size = len(files_path)
-            files_path = [f for f in files_path if not RootAPIHandler._is_dicom_report(f)]
-            if old_size != len(files_path):
-                _LOGGER.info(f"Discarded {old_size - len(files_path)} DICOM report files from upload.")
-
         if on_error not in ['raise', 'skip']:
             raise ValueError("on_error must be either 'raise' or 'skip'")
 
         files_path, is_multiple_resources = RootAPIHandler.__process_files_parameter(files_path)
+
+        # Discard DICOM reports
+        if discard_dicom_reports:
+            files_path = [f for f in files_path if not RootAPIHandler._is_dicom_report(f)]
+            old_size = len(files_path)
+            if old_size is not None and old_size != len(files_path):
+                _LOGGER.info(f"Discarded {old_size - len(files_path)} DICOM report files from upload.")
+
         if isinstance(metadata, (str, dict)):
             _LOGGER.debug("Converting metadatas to a list")
             metadata = [metadata]
@@ -680,7 +669,7 @@ class RootAPIHandler(BaseAPIHandler):
                 is_list = False
                 new_file_path = [file_path]
         # Check if is an IO object
-        elif _is_io_object(file_path):
+        elif is_io_object(file_path):
             is_list = False
             new_file_path = [file_path]
         elif not hasattr(file_path, '__len__'):
@@ -730,8 +719,8 @@ class RootAPIHandler(BaseAPIHandler):
 
     def get_resources(self,
                       status: Optional[ResourceStatus] = None,
-                      from_date: Optional[date] = None,
-                      to_date: Optional[date] = None,
+                      from_date: date | str | None = None,
+                      to_date: date | str | None = None,
                       tags: Optional[Sequence[str]] = None,
                       modality: Optional[str] = None,
                       mimetype: Optional[str] = None,
@@ -749,8 +738,8 @@ class RootAPIHandler(BaseAPIHandler):
 
         Args:
             status (ResourceStatus): The resource status. Possible values: 'inbox', 'published', 'archived' or None. If None, it will return all resources.
-            from_date (Optional[date]): The start date.
-            to_date (Optional[date]): The end date.
+            from_date (date | str | None): The start date.
+            to_date (date | str | None): The end date.
             tags (Optional[list[str]]): The tags to filter the resources.
             modality (Optional[str]): The modality of the resources.
             mimetype (Optional[str]): The mimetype of the resources.
@@ -769,9 +758,15 @@ class RootAPIHandler(BaseAPIHandler):
         """
         # Convert datetime objects to ISO format
         if from_date:
-            from_date = from_date.isoformat()
+            if isinstance(from_date, str):
+                date.fromisoformat(from_date)
+            else:
+                from_date = from_date.isoformat()
         if to_date:
-            to_date = to_date.isoformat()
+            if isinstance(to_date, str):
+                date.fromisoformat(to_date)
+            else:
+                to_date = to_date.isoformat()
 
         # Prepare the payload
         payload = {
@@ -905,7 +900,7 @@ class RootAPIHandler(BaseAPIHandler):
                                     ) -> None:
         """
         Download multiple resources and save them to the specified paths.
-        
+
         Args:
             resource_ids (list[str]): A list of resource unique ids.
             save_path (list[str] | str): A list of paths to save the files or a directory path.
@@ -955,6 +950,9 @@ class RootAPIHandler(BaseAPIHandler):
             >>> api_handler.download_resource_file('resource_id', save_path='path/to/dicomfile.dcm')
                 saves the file in the specified path.
         """
+        if save_path is None and add_extension:
+            raise ValueError("If add_extension is True, save_path must be provided.")
+
         url = f"{self._get_endpoint_url(RootAPIHandler.ENDPOINT_RESOURCES)}/{resource_id}/file"
         request_params = {'method': 'GET',
                           'headers': {'accept': 'application/octet-stream'},
@@ -966,7 +964,8 @@ class RootAPIHandler(BaseAPIHandler):
             mimetype = None
             if auto_convert or add_extension:
                 resource_info = self.get_resources_by_ids(resource_id)
-                mimetype = resource_info['mimetype']
+                mimetype = resource_info.get('mimetype', guess_type(response.content)[0])
+                
 
             if auto_convert:
                 try:
@@ -987,15 +986,15 @@ class RootAPIHandler(BaseAPIHandler):
             raise e
 
         if save_path is not None:
-            if add_extension:
-                ext = mimetypes.guess_extension(mimetype)
+            if add_extension and mimetype is not None:
+                ext = guess_extension(mimetype)
                 if ext is not None and not save_path.endswith(ext):
                     save_path += ext
             with open(save_path, 'wb') as f:
                 f.write(response.content)
 
-        if add_extension:
-            return resource_file, save_path
+            if add_extension:
+                return resource_file, save_path
         return resource_file
 
     def download_resource_frame(self,

{datamint-1.7.2 → datamint-1.7.4}/datamint/dataset/base_dataset.py

@@ -283,10 +283,12 @@ class DatamintBaseDataset:
         """Post-process data after loading metadata."""
         self._check_integrity()
         self._calculate_dataset_length()
-        self._precompute_frame_data()
+        if self.return_frame_by_frame:
+            self._precompute_frame_data()
+        self.subset_indices = list(range(self.dataset_length))
         self._setup_labels()
 
-        if self.discard_without_annotations and self.return_frame_by_frame:
+        if self.discard_without_annotations:
             self._filter_unannotated()
 
     def _calculate_dataset_length(self) -> None:
@@ -301,9 +303,8 @@ class DatamintBaseDataset:
 
     def _precompute_frame_data(self) -> None:
         """Precompute frame-related data for efficient indexing."""
-        self.num_frames_per_resource = self.__compute_num_frames_per_resource()
-        self._cumulative_frames = np.cumsum([0] + self.num_frames_per_resource)
-        self.subset_indices = list(range(self.dataset_length))
+        num_frames_per_resource = self.__compute_num_frames_per_resource()
+        self._cumulative_frames = np.cumsum([0] + num_frames_per_resource)
 
     def _setup_labels(self) -> None:
         """Setup label sets and mappings."""
@@ -989,9 +990,11 @@ class DatamintBaseDataset:
             return Path(resource['file'])
         else:
             ext = guess_extension(resource['mimetype'], strict=False)
+            _LOGGER.debug(f"Guessed extension for resource {resource['id']}|{resource['mimetype']}: {ext}")
             if ext is None:
                 _LOGGER.warning(f"Could not guess extension for resource {resource['id']}.")
                 ext = ''
+                raise Exception
             return Path('images', f"{resource['id']}{ext}")
 
     def _get_annotation_file_path(self, annotation: dict | Annotation) -> Path | None:

{datamint-1.7.2 → datamint-1.7.4}/pyproject.toml

@@ -1,7 +1,7 @@
 [project]
 name = "datamint"
 description = "A library for interacting with the Datamint API, designed for efficient data management, processing and Deep Learning workflows."
-version = "1.7.2"
+version = "1.7.4"
 dynamic = ["dependencies"]
 requires-python = ">=3.10"
 readme = "README.md"
@@ -40,7 +40,7 @@ matplotlib = "*"
 lightning = "*"
 albumentations = ">=2.0.0"
 lazy-loader = ">=0.3.0"
-medimgkit = "*"
+medimgkit = ">=0.2.0"
 # For compatibility with the datamintapi package
 datamintapi = "0.0.*"
 # Extra dependencies for docs