scripts and typed

mgf_filterer/exceptions.py

@@ -0,0 +1,7 @@
+class MGFFileNotFoundError(FileNotFoundError):
+    """Exception raised when an MGF file cannot be found."""
+
+    def __init__(self, filepath: str) -> None:
+        message = f"No such file or directory: '{filepath}'"
+        super().__init__(message)
+        self.filepath = filepath

mgf_filterer/mgf_processing.py

@@ -0,0 +1,33 @@
+from pathlib import Path
+from typing import List
+
+from matchms.importing import load_from_mgf
+
+from mgf_filterer.exceptions import MGFFileNotFoundError
+
+
+def load_mgf_files(filepath: str) -> List:
+    """
+    Load spectra from a .mgf file.
+
+    Parameters:
+    filepath (str): The path to the .mgf file to be loaded.
+
+    Returns:
+    List[Spectrum]: A list of Spectrum objects loaded from the file.
+
+    Raises:
+    MGFFileNotFoundError: If the .mgf file cannot be found.
+    """
+    if not Path(filepath).exists():
+        raise MGFFileNotFoundError(filepath)
+
+    spectra = list(load_from_mgf(filepath))
+    return spectra
+
+
+# Example usage:
+if __name__ == "__main__":
+    # Replace 'path_to_your_file.mgf' with the actual file path
+    spectra = load_mgf_files("path_to_your_file.mgf")
+    print(f"Loaded {len(spectra)} spectra from the file.")

pyproject.toml

@@ -11,7 +11,8 @@ packages = [
 ]
 
 [tool.poetry.dependencies]
-python = ">=3.8,<4.0"
+python = ">=3.8,<3.13"
+matchms = "^0.24.3"
 
 [tool.poetry.group.dev.dependencies]
 pytest = "^7.2.0"
@@ -20,6 +21,7 @@ deptry = "^0.12.0"
 mypy = "^1.5.1"
 pre-commit = "^3.4.0"
 tox = "^4.11.1"
+ipykernel = "^6.29.4"
 
 [tool.poetry.group.docs.dependencies]
 mkdocs = "^1.4.2"
@@ -40,6 +42,10 @@ warn_return_any = "True"
 warn_unused_ignores = "True"
 show_error_codes = "True"
 
+[[tool.mypy.overrides]]
+module = "matchms.*"
+ignore_missing_imports = "True"
+
 [tool.pytest.ini_options]
 testpaths = ["tests"]
 

pytest.ini

@@ -0,0 +1,3 @@
+[pytest]
+filterwarnings =
+    ignore::DeprecationWarning:sparsestack.*

tests/test_mgf_processing.py

@@ -0,0 +1,26 @@
+from pathlib import Path
+
+import pytest
+
+from mgf_filterer.mgf_processing import load_mgf_files
+
+
+@pytest.fixture
+def mgf_file_path():
+    # Assuming the test data is in a subdirectory of the tests directory
+    return Path(__file__).parent / "testdata" / "erythroxylum_coca.mgf"
+
+
+def test_load_mgf_files_valid(mgf_file_path):
+    """Test loading a valid .mgf file."""
+    spectra = list(load_mgf_files(str(Path(__file__).parent / "testdata" / "erythroxylum_coca.mgf")))
+    assert len(spectra) == 1098, "No spectra loaded from a valid file."
+
+
+def test_load_mgf_files_nonexistent():
+    """Test loading from a nonexistent file path."""
+    with pytest.raises(FileNotFoundError):
+        load_mgf_files("nonexistent_file.mgf")
+
+
+# More tests can be added for different scenarios

tests/testdata/cocaine.mgf

@@ -0,0 +1,51 @@
+BEGIN IONS
+PEPMASS=304.151201163587
+CHARGE=1
+MSLEVEL=2
+51.02326202392578 2706680.25
+57.05759048461914 3669400.5
+58.06536102294922 2941336.25
+58.89093017578125 2128817.0
+62.50955581665039 2552632.0
+65.03856658935547 7410005.5
+68.04956817626953 5514949.5
+70.06505584716797 8248730.0
+74.17985534667969 2559632.25
+77.0383071899414 4350874.0
+79.0539321899414 6849389.5
+81.06940460205078 3283495.25
+82.06478118896484 3.54302624E8
+83.06818389892578 5227826.5
+83.07249450683594 1.8868914E7
+91.05368041992188 5.1169252E7
+91.764404296875 2450964.75
+93.03289031982422 2.487278E7
+93.06930541992188 2.4062002E7
+94.06431579589844 1.9296378E7
+94.44011688232422 2585272.75
+95.04850769042969 9126912.0
+96.08019256591797 1.6715993E7
+97.0640869140625 6438870.0
+105.0325698852539 2.17488176E8
+105.06919860839844 3719801.0
+107.04859161376953 5381365.5
+108.07981872558594 8.94284E7
+108.087158203125 4666916.5
+114.09021759033203 2537875.0
+118.04029846191406 1.0775417E7
+119.04790496826172 7.8783096E7
+122.09513854980469 5.6339332E7
+125.05862426757812 8981573.0
+132.07960510253906 5560816.0
+133.06361389160156 3509932.25
+135.06613159179688 3572607.5
+150.08990478515625 1.45036352E8
+151.0742950439453 1.4068616E7
+154.08468627929688 1.9900514E7
+182.11578369140625 1.685805568E9
+183.11920166015625 3.2352878E7
+194.062744140625 2746359.0
+272.1258239746094 1.6657989E7
+304.1513671875 1.19192128E9
+305.1551208496094 4.1532744E7
+END IONS