move and clean helper function

Author: florian-huber

ms2query/data_processing/__init__.py

@@ -1,6 +1,7 @@
 from .chemistry_utils import compute_morgan_fingerprints, inchikey14_from_full
 from .fingerprint_computation import compute_fingerprints_from_smiles
-from .merging_utils import cluster_block, get_merged_spectra, normalize_spectrum_sum
+from .merging_utils import cluster_block, get_merged_spectra
+from .spectra_processing import normalize_spectrum_sum
 
 
 __all__ = [

ms2query/data_processing/merging_utils.py

@@ -3,6 +3,7 @@
 from scipy.sparse import csr_matrix
 from scipy.sparse.csgraph import connected_components
 from ms2query.metrics import compute_cosine_greedy
+from .spectra_processing import normalize_spectrum_sum
 
 
 METADATA_FIELDS_FROM_FIRST = [
@@ -12,19 +13,6 @@
 
 
 # --------------------- Helper functions ---------------------
-def normalize_spectrum_sum(s):
-    """Return a spectrum with intensities normalized to sum=1 (if possible)."""
-    mz = np.asarray(s.peaks.mz, dtype=float)
-    intens = np.asarray(s.peaks.intensities, dtype=float)
-    tot = intens.sum()
-    if tot > 0:
-        intens = intens / tot
-
-    # Build a shallow copy with normalized peaks but same metadata
-    md = dict(s.metadata) if hasattr(s, "metadata") else {}
-    return Spectrum(mz=mz, intensities=intens, metadata=md)
-
-
 def _merge_cluster_to_consensus(cluster_spectra, mz_tol=0.01, min_frac=0.25):
     """
     Build a consensus spectrum from a list of matchms Spectra.

ms2query/data_processing/spectra_processing.py

@@ -0,0 +1,14 @@
+from matchms import Spectrum
+
+
+def normalize_spectrum_sum(spectrum: Spectrum) -> Spectrum:
+    """Return a spectrum with intensities normalized to sum=1 (if possible)."""
+    mz = spectrum.peaks.mz
+    intens = spectrum.peaks.intensities
+    tot = intens.sum()
+    if tot > 0:
+        intens = intens / tot
+
+    # Build a shallow copy with normalized peaks but same metadata
+    md = dict(spectrum.metadata) if hasattr(spectrum, "metadata") else {}
+    return Spectrum(mz=mz, intensities=intens, metadata=md)