Make parse_1d_scan_coords() more robust (#607)

The `parse_1d_scan_coords()` crashed for a specific TS scan output file for two independent reasons. Both bugs were fixed now, and a test was added.
ReactionMechanismGenerator · Mar 10, 2023 · b7fc76f · b7fc76f
2 parents a9bc6ad + 1c2f580
commit b7fc76f
Show file tree

Hide file tree

Showing 3 changed files with 38,041 additions and 3 deletions.
diff --git a/arc/parser.py b/arc/parser.py
@@ -475,17 +475,21 @@ def parse_1d_scan_coords(path: str) -> List[Dict[str, tuple]]:
         if i >= len(lines) or 'Normal termination of Gaussian' in lines[i] or 'Error termination via' in lines[i]:
             done = True
         elif 'Optimization completed' in lines[i]:
-            while len(lines) and 'Input orientation:' not in lines[i]:
+            while i < len(lines) + 10 and 'Input orientation:' not in lines[i] or 'Forces (Hartrees/Bohr)' in lines [i + 7]:
                 i += 1
                 if 'Error termination via' in lines[i]:
                     return traj
             i += 5
-            xyz_str = ''
+            xyz_str, skip_traj = '', False
             while len(lines) and '--------------------------------------------' not in lines[i]:
+                if 'DIIS: error' in lines[i]:
+                    skip_traj = True
+                    break
                 splits = lines[i].split()
                 xyz_str += f'{qcel.periodictable.to_E(int(splits[1]))}  {splits[3]}  {splits[4]}  {splits[5]}\n'
                 i += 1
-            traj.append(str_to_xyz(xyz_str))
+            if not skip_traj:
+                traj.append(str_to_xyz(xyz_str))
         i += 1
     return traj
 

diff --git a/arc/parser_test.py b/arc/parser_test.py
@@ -421,6 +421,13 @@ def test_parse_1d_scan_coords(self):
         self.assertEqual(len(traj_3), 45)
         self.assertEqual(traj_3[0]['symbols'], ('C', 'C', 'O', 'H', 'H', 'H', 'H', 'H', 'H'))
 
+        path_4 = os.path.join(ARC_PATH, 'arc', 'testing', 'rotor_scans', 'TS_scan.out')
+        traj_4 = parser.parse_1d_scan_coords(path_4)
+        self.assertEqual(len(traj_4), 8)  # output file trimmed
+        self.assertEqual(traj_4[0]['symbols'], ('C', 'C', 'H', 'H', 'H', 'H', 'H', 'C', 'C', 'C', 'C', 'C', 'C', 'C',
+                                                'C', 'C', 'C', 'C', 'H', 'H', 'H', 'H', 'H', 'H', 'H', 'H', 'H', 'H',
+                                                'H', 'H', 'H', 'H', 'H', 'H', 'H', 'H', 'H', 'H', 'H', 'H', 'H', 'H'))
+
     def test_parse_t1(self):
         """Test T1 diagnostic parsing"""
         path = os.path.join(ARC_PATH, 'arc', 'testing', 'sp', 'mehylamine_CCSD(T).out')