ENH: Add PDB reader benchmark to track parsing performance

benchmarks/benchmarks/traj_reader.py

@@ -28,6 +28,12 @@
 except ImportError:
     pass
 
+try:
+    import MDAnalysis as mda
+    from MDAnalysisTests.datafiles import PDB_multiframe
+except ImportError:
+    pass
+
 traj_dict = {
     "XTC": [XTC, XTCReader],
     "TRR": [TRR, TRRReader],
@@ -71,3 +77,20 @@ def time_strides(self, traj_format):
         """
         for ts in self.reader_object:
             pass
+
+
+class PDBReaderBench(object):
+    """Benchmarks for PDB file format reading and parsing"""
+
+    units = "ms"
+    timeout = 60.0
+
+    def setup(self):
+        self.u = mda.Universe(PDB_multiframe)
+
+    def time_read(self):
+        mda.Universe(PDB_multiframe)
+
+    def time_iterate(self):
+        for ts in self.u.trajectory:
+            pass

package/CHANGELOG

@@ -22,6 +22,7 @@ The rules for this file:
  * 2.11.0
 
 Fixes
+ * Added ASV benchmark for PDB trajectory reading (PR #1234)
  * `MDAnalysis.analysis.nucleicacids.WatsonCrickDist`, `MinorPairDist`,
    and `MajorPairDist` now match residue names against the full resname
    instead of only the first character, fixing incorrect behaviour with