Merge pull request #107 from FAIRmat-NFDI/processing-order

Process yaml and json files last

.pre-commit-config.yaml

@@ -1,7 +1,7 @@
 repos:
   - repo: https://github.com/astral-sh/ruff-pre-commit
     # Ruff version.
-    rev: v0.8.0
+    rev: v0.9.1
     hooks:
       # Run the linter.
       - id: ruff

pyproject.toml

@@ -32,7 +32,7 @@ dependencies = [
 [project.optional-dependencies]
 dev = [
     "mypy",
-    "ruff>=0.6.0",
+    "ruff>=0.9.0",
     "pytest",
     "pytest-cov",
     "pytest-timeout",

src/pynxtools_xps/reader.py

@@ -157,7 +157,7 @@ class XPSReader(MultiFormatReader):
     )
 
     __vndr_err_msg__ = (
-        "Need an XPS data file from one of the following vendors: " f"{__vendors__}"
+        f"Need an XPS data file from one of the following vendors: {__vendors__}"
     )
 
     def __init__(self, *args, **kwargs):
@@ -176,6 +176,12 @@ def __init__(self, *args, **kwargs):
         for ext in XPSReader.__prmt_file_ext__:
             self.extensions[ext] = self.handle_data_file
 
+        self.processing_order = XPSReader.__prmt_file_ext__ + [
+            ".yml",
+            ".yaml",
+            ".json",
+        ]
+
     def set_config_file(self, file_path: str) -> Dict[str, Any]:
         if self.config_file is not None:
             logger.info(

src/pynxtools_xps/reader_utils.py

@@ -414,12 +414,12 @@ def construct_data_key(spectrum: Dict[str, Any]) -> str:
 
     """
     if "loop_no" in spectrum:
-        cycle_key = f'cycle{spectrum["loop_no"]}'
+        cycle_key = f"cycle{spectrum['loop_no']}"
     else:
         cycle_key = "cycle0"
 
     if "scan_no" in spectrum:
-        scan_key = f'scan{spectrum["scan_no"]}'
+        scan_key = f"scan{spectrum['scan_no']}"
     else:
         scan_key = "scan0"
 

src/pynxtools_xps/specs/sle/sle_specs.py

@@ -266,8 +266,8 @@ def _update_xps_dict_with_spectrum(
                 if units is not None:
                     self._xps_dict[f"{root}/{mpes_key}/@units"] = units
 
-        self._xps_dict[f'{path_map["electronanalyser"]}/name'] = spectrum["devices"][0]
-        self._xps_dict[f'{path_map["source"]}/name'] = spectrum["devices"][1]
+        self._xps_dict[f"{path_map['electronanalyser']}/name"] = spectrum["devices"][0]
+        self._xps_dict[f"{path_map['source']}/name"] = spectrum["devices"][1]
 
         # Create keys for writing to data
         scan_key = construct_data_key(spectrum)
@@ -298,7 +298,7 @@ def _update_xps_dict_with_spectrum(
             # TODO: fix this hotfix so that all data can be written
 
             # Add energy axis to energy_calibration
-            calib_energy_key = f'{path_map["process/energy_calibration"]}/energy'
+            calib_energy_key = f"{path_map['process/energy_calibration']}/energy"
             self._xps_dict[calib_energy_key] = energy
 
             self._xps_dict["data"][entry][scan_key.split("_")[0]] = xr.DataArray(

src/pynxtools_xps/specs/xml/xml_specs.py

@@ -400,9 +400,7 @@ def struct_fc_name_sc_value(self, element_, first_child, parent_path, skip_child
         for unit in units:
             if f"_[{unit}]" in section_nm_reslvr:
                 section_nm_reslvr, _ = section_nm_reslvr.split("_")
-                self.metadata_dict[f"{parent_path}/" f"{section_nm_reslvr}/@units"] = (
-                    unit
-                )
+                self.metadata_dict[f"{parent_path}/{section_nm_reslvr}/@units"] = unit
 
         parent_path, self.tail_part_frm_struct = self.check_last_part_repetition(
             parent_path, self.tail_part_frm_struct, section_nm_reslvr
@@ -513,7 +511,7 @@ def last_element_parser(self, element_: EmtT.Element, parent_path: str) -> None:
 
         if child_num == 0:
             if "name" in elmt_attr.keys():
-                section_nm_reslvr = f'{elmt_attr["name"]}'
+                section_nm_reslvr = f"{elmt_attr['name']}"
                 value = self.restructure_value(element_.text, element_.tag)
 
                 parent_path, self.tail_part_frm_othr = self.check_last_part_repetition(

tests/test_helpers.py

@@ -80,16 +80,16 @@ def test_extract_unit(
 ):
     result_value, result_unit = extract_unit(key, value, unit_missing)
 
-    assert isinstance(
-        result_value, type(expected_value)
-    ), f"Expected type {type(expected_value)} but got type {type(result_value)}"
+    assert isinstance(result_value, type(expected_value)), (
+        f"Expected type {type(expected_value)} but got type {type(result_value)}"
+    )
 
-    assert (
-        result_value == expected_value
-    ), f"Expected {expected_value} but got {result_value}"
-    assert (
-        result_unit == expected_unit
-    ), f"Expected {expected_unit} but got {result_unit}"
+    assert result_value == expected_value, (
+        f"Expected {expected_value} but got {result_value}"
+    )
+    assert result_unit == expected_unit, (
+        f"Expected {expected_unit} but got {result_unit}"
+    )
 
 
 @pytest.mark.parametrize(