Merge pull request #267 from catalystneuro/improve_bruker_tiff

Improve Bruker xml parsing by using iterparse instead of parsing the complete xml file

CHANGELOG.md

@@ -1,5 +1,8 @@
 # Upcoming
 
+### Improvements
+* Improved xml parsing with Bruker [PR #267](https://github.com/catalystneuro/roiextractors/pull/267)
+
 
 # v0.5.5
 

src/roiextractors/extractors/tiffimagingextractors/brukertiffimagingextractor.py

@@ -65,9 +65,15 @@ def _determine_imaging_is_volumetric(folder_path: PathType) -> bool:
     is_volumetric: bool
         True if the imaging is volumetric (multiplane), False otherwise (single plane).
     """
-    xml_root = _parse_xml(folder_path=folder_path)
-    z_device_element = xml_root.find(".//PVStateValue[@key='zDevice']")
-    is_volumetric = bool(int(z_device_element.attrib["value"]))
+    folder_path = Path(folder_path)
+    xml_file_path = folder_path / f"{folder_path.name}.xml"
+    assert xml_file_path.is_file(), f"The XML configuration file is not found at '{xml_file_path}'."
+
+    is_volumetric = False
+    for event, elem in ElementTree.iterparse(xml_file_path, events=("start",)):
+        if elem.tag == "PVStateValue" and elem.attrib.get("key") == "zDevice":
+            is_volumetric = bool(int(elem.attrib["value"]))
+            break  # Stop parsing as we've found the required element
 
     return is_volumetric
 
@@ -107,25 +113,52 @@ def get_streams(cls, folder_path: PathType) -> dict:
         """
         natsort = get_package(package_name="natsort", installation_instructions="pip install natsort")
 
-        xml_root = _parse_xml(folder_path=folder_path)
-        channel_names = [file.attrib["channelName"] for file in xml_root.findall(".//File")]
-        unique_channel_names = natsort.natsorted(set(channel_names))
+        folder_path = Path(folder_path)
+        xml_file_path = folder_path / f"{folder_path.name}.xml"
+        assert xml_file_path.is_file(), f"The XML configuration file is not found at '{folder_path}'."
+
+        channel_names = set()
+        channel_ids = set()
+        file_names = []
+
+        # Parse the XML file iteratively to find the first Sequence element
+        first_sequence_element = None
+        for _, elem in ElementTree.iterparse(xml_file_path, events=("end",)):
+            if elem.tag == "Sequence":
+                first_sequence_element = elem
+                break
+
+        if first_sequence_element is None:
+            raise ValueError("No Sequence element found in the XML configuration file. Can't get streams")
+
+        # Then in the first Sequence we find all the Frame elements
+        if first_sequence_element is not None:
+            # Iterate over all Frame elements within the first Sequence
+            frame_elements = first_sequence_element.findall(".//Frame")
+            for frame_elemenet in frame_elements:
+                # Iterate over all File elements within each Frame
+                for file_elem in frame_elemenet.findall("File"):
+                    channel_names.add(file_elem.attrib["channelName"])
+                    channel_ids.add(file_elem.attrib["channel"])
+                    file_names.append(file_elem.attrib["filename"])
+
+        unique_channel_names = natsort.natsorted(channel_names)
+        unique_channel_ids = natsort.natsorted(channel_ids)
+
         streams = dict(channel_streams=unique_channel_names)
         streams["plane_streams"] = dict()
+
         if not _determine_imaging_is_volumetric(folder_path=folder_path):
             return streams
-        # The "channelName" can be any name that the experimenter sets (e.g. 'Ch1', 'Ch2', 'Green', 'Red')
-        # Use the identifier of a channel "channel" (e.g. 1, 2) to match it to the file name
-        channel_ids = [file.attrib["channel"] for file in xml_root.findall(".//File")]
-        unique_channel_ids = natsort.natsorted(set(channel_ids))
+
         for channel_id, channel_name in zip(unique_channel_ids, unique_channel_names):
             plane_naming_pattern = rf"(?P<stream_name>Ch{channel_id}_\d+)"
-            plane_stream_names = [
-                re.search(plane_naming_pattern, file.attrib["filename"])["stream_name"]
-                for file in xml_root.findall(f".//File")
-            ]
+            regular_expression_matches = [re.search(plane_naming_pattern, filename) for filename in file_names]
+            plane_stream_names = [matches["stream_name"] for matches in regular_expression_matches if matches]
+
             unique_plane_stream_names = natsort.natsorted(set(plane_stream_names))
             streams["plane_streams"][channel_name] = unique_plane_stream_names
+
         return streams
 
     def __init__(

tests/test_brukertiffimagingextactor.py

@@ -133,9 +133,9 @@ def setUpClass(cls):
         cls.test_video[..., 1] = second_plane_video
 
     def test_stream_names(self):
-        self.assertEqual(
-            BrukerTiffMultiPlaneImagingExtractor.get_streams(folder_path=self.folder_path), self.available_streams
-        )
+        found_streams = BrukerTiffMultiPlaneImagingExtractor.get_streams(folder_path=self.folder_path)
+        expected_streams = self.available_streams
+        self.assertEqual(found_streams, expected_streams)
 
     def test_brukertiffextractor_image_size(self):
         self.assertEqual(self.extractor.get_image_size(), (512, 512, 2))