Parse precursor charge specified in PEPMASS (override CHARGE)

CHANGELOG

@@ -1,3 +1,10 @@
+4.7.3a2
+-------
+
+ - Add compatibility with NumPy 2.0.
+ - Fix `#153 <https://github.com/levitsky/pyteomics/issues/153>`_. MGF parser now recognizes precursor charge specified
+   on the PEPMASS line. If CHARGE is also specified, it is ignored.
+
 4.7.2
 -----
 

pyteomics/mgf.py

@@ -149,6 +149,25 @@ def __setstate__(self, state):
     def parse_precursor_charge(charge_text, list_only=False):
         return aux._parse_charge(charge_text, list_only=list_only)
 
+    @staticmethod
+    def parse_pepmass_charge(pepmass_str):
+        split = pepmass_str.split()
+        if len(split) > 3:
+            raise aux.PyteomicsError('MGF format error: cannot parse '
+                    'PEPMASS = {}'.format(pepmass_str))
+        elif len(split) == 3:
+            charge = split[2]
+            try:
+                pepmass = tuple(map(float, split[:2]))
+            except ValueError:
+                raise aux.PyteomicsError('MGF format error: cannot parse '
+                    'PEPMASS = {}'.format(pepmass_str))
+        else:
+            pepmass = tuple(map(float, split[:2]))
+            pepmass = pepmass + (None,) * (2-len(pepmass))
+            charge = None
+        return pepmass, charge
+
     @staticmethod
     def parse_peak_charge(charge_text, list_only=False):
         return aux._parse_charge(charge_text, list_only=False)
@@ -203,13 +222,9 @@ def _read_spectrum_lines(self, lines):
                 pass
             elif sline == 'END IONS':
                 if 'pepmass' in params:
-                    try:
-                        pepmass = tuple(map(float, params['pepmass'].split()))
-                    except ValueError:
-                        raise aux.PyteomicsError('MGF format error: cannot parse '
-                                'PEPMASS = {}'.format(params['pepmass']))
-                    else:
-                        params['pepmass'] = pepmass + (None,) * (2-len(pepmass))
+                    params['pepmass'], charge = self.parse_pepmass_charge(params['pepmass'])
+                    if charge is not None:
+                        params['charge'] = charge
                 if isinstance(params.get('charge'), aux.basestring):
                     params['charge'] = self.parse_precursor_charge(params['charge'], True)
                 if 'rtinseconds' in params:

pyteomics/version.py

@@ -19,7 +19,7 @@
 
 """
 
-__version__ = '4.7.3a1'
+__version__ = '4.7.3a2'
 
 from collections import namedtuple
 import re

tests/test_mgf.py

@@ -264,6 +264,12 @@ def test_charge_repr_multiple(self):
         self.assertEqual(mgf._charge_repr('charge', '2+, 3+'), 'CHARGE=2+ and 3+')
         self.assertEqual(mgf._charge_repr('charge', np.array([2, 3])), 'CHARGE=2+ and 3+')
 
+    def test_pepmass_parsing(self):
+        with mgf.MGF('test_pepmass.mgf') as f:
+            spectra = list(f)
+            self.assertEqual(len(spectra), 3)
+            self.assertEqual(spectra[0]['params'], spectra[1]['params'])
+            self.assertEqual(spectra[0]['params'], spectra[2]['params'])
 
 if __name__ == "__main__":
     unittest.main()