Remove serialem code and revise context determination (#632)

The SerialEM code has never been used, and if we want to support SerialEM in future it would be better to redo it. This removes all code relating to it.

Also improves and tests the determination of the context in the analyser.

There are some potential difficulties with distinguishing data and metadata. The biggest problem is mdoc files which appear with the same name in both folders.

---------

Co-authored-by: Eu Pin Tien <[email protected]>

Author: stephen-riggs

src/murfey/client/analyser.py

@@ -118,9 +118,6 @@ def _find_context(self, file_path: Path) -> bool:
         in the Context classes themselves.
         """
         logger.debug(f"Finding context using file {str(file_path)!r}")
-        if "atlas" in file_path.parts:
-            self._context = SPAMetadataContext("epu", self._basepath)
-            return True
 
         # CLEM workflow checks
         # Look for LIF and XLIF files
@@ -129,75 +126,60 @@ def _find_context(self, file_path: Path) -> bool:
             return True
         # Look for TIFF files associated with CLEM workflow
         # Leica's autosave mode seems to name the TIFFs in the format
-        # PostionXX--ZXX-CXX.tif
-        if (
-            "--" in file_path.name
-            and file_path.suffix in (".tiff", ".tif")
-            and self._environment
-        ):
-            created_directories = set(
-                get_machine_config_client(
-                    str(self._environment.url.geturl()),
-                    instrument_name=self._environment.instrument_name,
-                    demo=self._environment.demo,
-                ).get("analyse_created_directories", [])
-            )
-            if created_directories.intersection(set(file_path.parts)):
-                self._context = CLEMContext("leica", self._basepath)
-                return True
+        # PostionXX--ZXX--CXX.tif
+        if all(
+            pattern in file_path.name for pattern in ("--Z", "--C")
+        ) and file_path.suffix in (".tiff", ".tif"):
+            self._context = CLEMContext("leica", self._basepath)
+            return True
 
         # Tomography and SPA workflow checks
-        split_file_name = file_path.name.split("_")
-        if split_file_name:
-            # Skip context for gain files
-            if "gain" in split_file_name[-1]:
+        if "atlas" in file_path.parts:
+            self._context = SPAMetadataContext("epu", self._basepath)
+            return True
+
+        if "Metadata" in file_path.parts or file_path.name == "EpuSession.dm":
+            self._context = SPAMetadataContext("epu", self._basepath)
+            return True
+        elif (
+            "Batch" in file_path.parts
+            or "SearchMaps" in file_path.parts
+            or "Thumbnails" in file_path.parts
+            or file_path.name == "Session.dm"
+        ):
+            self._context = TomographyMetadataContext("tomo", self._basepath)
+            return True
+
+        split_file_stem = file_path.stem.split("_")
+        if split_file_stem:
+            if split_file_stem[-1] == "gain":
                 return False
 
             # Files starting with "FoilHole" belong to the SPA workflow
-            if split_file_name[0].startswith("FoilHole"):
+            if split_file_stem[0].startswith("FoilHole") and split_file_stem[-1] in [
+                "Fractions",
+                "fractions",
+                "EER",
+            ]:
                 if not self._context:
                     logger.info("Acquisition software: EPU")
                     self._context = SPAModularContext("epu", self._basepath)
                 self.parameters_model = ProcessingParametersSPA
                 return True
 
             # Files starting with "Position" belong to the standard tomography workflow
+            # NOTE: not completely reliable, mdocs can be in tomography metadata as well
             if (
-                split_file_name[0] == "Position"
+                split_file_stem[0] == "Position"
                 or "[" in file_path.name
-                or "Fractions" in split_file_name[-1]
-                or "fractions" in split_file_name[-1]
-                or "EER" in split_file_name[-1]
+                or split_file_stem[-1] in ["Fractions", "fractions", "EER"]
+                or file_path.suffix == ".mdoc"
             ):
                 if not self._context:
                     logger.info("Acquisition software: tomo")
                     self._context = TomographyContext("tomo", self._basepath)
                     self.parameters_model = ProcessingParametersTomo
                 return True
-
-            # Files with these suffixes belong to the serial EM tomography workflow
-            if file_path.suffix in (".mrc", ".tiff", ".tif", ".eer"):
-                # Ignore batch files and search maps
-                if any(p in file_path.parts for p in ("Batch", "SearchMaps")):
-                    return False
-                # Ignore JPG files
-                if file_path.with_suffix(".jpg").is_file():
-                    return False
-                # Ignore the averaged movies written out by the Falcon
-                if (
-                    len(
-                        list(
-                            file_path.parent.glob(
-                                f"{file_path.name}*{file_path.suffix}"
-                            )
-                        )
-                    )
-                    > 1
-                ):
-                    return False
-                self._context = TomographyContext("serialem", self._basepath)
-                self.parameters_model = ProcessingParametersTomo
-                return True
         return False
 
     def post_transfer(self, transferred_file: Path):
@@ -288,12 +270,7 @@ def _analyse(self):
                             if not dc_metadata:
                                 try:
                                     dc_metadata = self._context.gather_metadata(
-                                        (
-                                            transferred_file.with_suffix(".mdoc")
-                                            if self._context._acquisition_software
-                                            == "serialem"
-                                            else self._xml_file(transferred_file)
-                                        ),
+                                        self._xml_file(transferred_file),
                                         environment=self._environment,
                                     )
                                 except NotImplementedError:

src/murfey/client/contexts/tomo.py

@@ -400,33 +400,6 @@ def _add_tomo_tilt(
             required_strings=required_strings,
         )
 
-    def _add_serialem_tilt(
-        self, file_path: Path, environment: MurfeyInstanceEnvironment | None = None
-    ) -> List[str]:
-        delimiters = ("_", "-")
-        for d in delimiters:
-            if file_path.name.count(d) > 1:
-                delimiter = d
-                break
-        else:
-            delimiter = delimiters[0]
-
-        def _extract_tilt_series(p: Path) -> str:
-            split = p.name.split(delimiter)
-            for s in split:
-                if s.isdigit():
-                    return s
-            raise ValueError(
-                f"No digits found in {p.name} after splitting on {delimiter}"
-            )
-
-        return self._add_tilt(
-            file_path,
-            lambda x: ".".join(x.name.split(delimiter)[-1].split(".")[:-1]),
-            environment=environment,
-            required_strings=[],
-        )
-
     def post_transfer(
         self,
         transferred_file: Path,
@@ -464,10 +437,8 @@ def post_transfer(
                         required_strings=kwargs.get("required_strings")
                         or required_strings,
                     )
-                elif self._acquisition_software == "serialem":
-                    completed_tilts = self._add_serialem_tilt(
-                        transferred_file, environment=environment
-                    )
+                else:
+                    logger.warning(f"Unknown data file {transferred_file}")
             if transferred_file.suffix == ".mdoc":
                 with open(transferred_file, "r") as md:
                     tilt_series = transferred_file.stem

src/murfey/client/multigrid_control.py

@@ -23,7 +23,7 @@
 from murfey.util.api import url_path_for
 from murfey.util.client import capture_delete, capture_post, get_machine_config_client
 
-log = logging.getLogger("murfey.client.mutligrid_control")
+log = logging.getLogger("murfey.client.multigrid_control")
 
 
 @dataclass

src/murfey/util/dummy_setup.py

@@ -24,7 +24,7 @@ def initialise(dummy_location: Path) -> Path:
         yaml.dump(
             {
                 "m12": {
-                    "acquisition_software": ["epu", "tomo", "serialem"],
+                    "acquisition_software": ["epu", "tomo"],
                     "data_directories": [str(detector_dir)],
                     "rsync_basepath": str(dummy_location),
                     "calibrations": {"dummy": 0},

tests/client/test_analyser.py

@@ -1,6 +1,122 @@
 from __future__ import annotations
 
+import pytest
+
 from murfey.client.analyser import Analyser
+from murfey.client.contexts.clem import CLEMContext
+from murfey.client.contexts.spa import SPAModularContext
+from murfey.client.contexts.spa_metadata import SPAMetadataContext
+from murfey.client.contexts.tomo import TomographyContext
+from murfey.client.contexts.tomo_metadata import TomographyMetadataContext
+from murfey.util.models import ProcessingParametersSPA, ProcessingParametersTomo
+
+example_files = [
+    # Tomography
+    ["visit/Position_1_001_0.0_20250715_012434_fractions.tiff", TomographyContext],
+    ["visit/Position_1_2_002_3.0_20250715_012434_Fractions.mrc", TomographyContext],
+    ["visit/Position_1_2_003_6.0_20250715_012434_EER.eer", TomographyContext],
+    ["visit/name1_004_9.0_20250715_012434_fractions.tiff", TomographyContext],
+    ["visit/Position_1_[30.0].tiff", TomographyContext],
+    ["visit/Position_1.mdoc", TomographyContext],
+    ["visit/name1_2.mdoc", TomographyContext],
+    # Tomography metadata
+    ["visit/Session.dm", TomographyMetadataContext],
+    ["visit/SearchMaps/SearchMap.xml", TomographyMetadataContext],
+    ["visit/Batch/BatchPositionsList.xml", TomographyMetadataContext],
+    ["visit/Thumbnails/file.mrc", TomographyMetadataContext],
+    # SPA
+    ["visit/FoilHole_01234_fractions.tiff", SPAModularContext],
+    ["visit/FoilHole_01234_EER.eer", SPAModularContext],
+    # SPA metadata
+    ["atlas/atlas.mrc", SPAMetadataContext],
+    ["visit/EpuSession.dm", SPAMetadataContext],
+    ["visit/Metadata/GridSquare.dm", SPAMetadataContext],
+    # CLEM LIF file
+    ["visit/images/test_file.lif", CLEMContext],
+    # CLEM TIFF files
+    [
+        "visit/images/2024_03_14_12_34_56--Project001/grid1/Position 12--Z02--C01.tif",
+        CLEMContext,
+    ],
+    [
+        "visit/images/2024_03_14_12_34_56--Project001/grid1/Position 12_Lng_LVCC--Z02--C01.tif",
+        CLEMContext,
+    ],
+    [
+        "visit/images/2024_03_14_12_34_56--Project001/grid1/Series001--Z00--C00.tif",
+        CLEMContext,
+    ],
+    [
+        "visit/images/2024_03_14_12_34_56--Project001/grid1/Series001_Lng_LVCC--Z00--C00.tif",
+        CLEMContext,
+    ],
+    # CLEM TIFF file accompanying metadata
+    [
+        "visit/images/2024_03_14_12_34_56--Project001/grid1/Metadata/Position 12.xlif",
+        CLEMContext,
+    ],
+    [
+        "visit/images/2024_03_14_12_34_56--Project001/grid1/Metadata/Position 12_Lng_LVCC.xlif",
+        CLEMContext,
+    ],
+    [
+        "visit/images/2024_03_14_12_34_56--Project001/grid1/Position 12/Metadata/Position 12_histo.xlif",
+        CLEMContext,
+    ],
+    [
+        "visit/images/2024_03_14_12_34_56--Project001/grid1/Position 12/Metadata/Position 12_Lng_LVCC_histo.xlif",
+        CLEMContext,
+    ],
+    [
+        "visit/images/2024_03_14_12_34_56--Project001/grid1/Metadata/Series001.xlif",
+        CLEMContext,
+    ],
+    [
+        "visit/images/2024_03_14_12_34_56--Project001/grid1/Metadata/Series001_Lng_LVCC.xlif",
+        CLEMContext,
+    ],
+]
+
+
+@pytest.mark.parametrize("file_and_context", example_files)
+def test_find_context(file_and_context, tmp_path):
+    # Unpack parametrised variables
+    file_name, context = file_and_context
+
+    # Pass the file to the Analyser; add environment as needed
+    analyser = Analyser(basepath_local=tmp_path)
+
+    # Check that the results are as expected
+    assert analyser._find_context(tmp_path / file_name)
+    assert isinstance(analyser._context, context)
+
+    # Checks for the specific workflow contexts
+    if isinstance(analyser._context, TomographyContext):
+        assert analyser.parameters_model == ProcessingParametersTomo
+    if isinstance(analyser._context, SPAModularContext):
+        assert analyser.parameters_model == ProcessingParametersSPA
+
+
+contextless_files = [
+    "visit/Position_1_gain.tiff",
+    "visit/FoilHole_01234_gain.tiff",
+    "visit/file_1.mrc",
+    "visit/FoilHole_01234.mrc",
+    "visit/FoilHole_01234.jpg",
+    "visit/FoilHole_01234.xml",
+    "visit/images/test_file.lifext",
+    "visit/images/2024_03_14_12_34_56--Project001/Project001.xlef",
+    "visit/images/2024_03_14_12_34_56--Project001/Project001.xlef.lock",
+    "visit/images/2024_03_14_12_34_56--Project001/grid1/Position 12/Position 12_histo.lof",
+    "visit/images/2024_03_14_12_34_56--Project001/grid1/Position 12/Series001_histo.lof",
+]
+
+
+@pytest.mark.parametrize("bad_file", contextless_files)
+def test_ignore_contextless_files(bad_file, tmp_path):
+    analyser = Analyser(tmp_path)
+    assert not analyser._find_context(tmp_path / bad_file)
+    assert not analyser._context
 
 
 def test_analyser_setup_and_stopping(tmp_path):
@@ -23,7 +139,7 @@ def test_analyser_tomo_determination(tmp_path):
 
 
 def test_analyser_epu_determination(tmp_path):
-    tomo_file = tmp_path / "FoilHole_12345_Data_6789.tiff"
+    tomo_file = tmp_path / "FoilHole_12345_Data_6789_Fractions.tiff"
     analyser = Analyser(tmp_path)
     analyser.start()
     analyser.queue.put(tomo_file)