Implement lru_cache for get_variables_by_attributes

src/netCDF4/_netCDF4.pyx

@@ -1226,6 +1226,7 @@ from cpython.bytes cimport PyBytes_FromStringAndSize
 from .utils import (_StartCountStride, _quantize, _find_dim, _walk_grps,
                     _out_array_shape, _sortbylist, _tostr, _safecast, _is_int)
 import sys
+import functools
 
 __version__ = "1.6.3"
 
@@ -2617,7 +2618,9 @@ Is the Dataset open or closed?
         return bool(self._isopen)
 
     def __dealloc__(self):
-        # close file when there are no references to object left
+        # close file when there are no references to object left and clear the cache.
+        if self.get_variables_by_attributes:
+            self.get_variables_by_attributes.cache_clear()
         if self._isopen:
            self._close(False)
 
@@ -3333,9 +3336,10 @@ of existing (sub-) groups and their variables.
             for group in groups:
                 group.set_ncstring_attrs(value) # recurse into subgroups...
 
+    @functools.lru_cache(maxsize=128)
     def get_variables_by_attributes(self, **kwargs):
         """
-**`get_variables_by_attribute(self, **kwargs)`**
+**`get_variables_by_attributes(self, **kwargs)`**
 
 Returns a list of variables that match specific conditions.
 

test/run_all.py

@@ -57,6 +57,11 @@
     test_files.remove('tst_cdl.py');
     sys.stdout.write('not running tst_cdl.py ...\n')
 
+# Don't run computationally intensive test
+if os.getenv('MEMORY_LEAK_TEST'):
+    test_files.remove('tst_multiple_open_close.py');
+    sys.stdout.write('not running tst_multiple_open_close.py ...\n')
+
 # Build the test suite from the tests found in the test files.
 testsuite = unittest.TestSuite()
 for f in test_files:

test/tst_multiple_open_close.py

@@ -0,0 +1,50 @@
+import os
+import tracemalloc
+import unittest
+
+import netCDF4
+
+class MultipleVariablesByAttributesCallsTests(unittest.TestCase):
+
+
+    def test_multiple_calls(self):
+        netcdf_file = os.path.join(os.path.dirname(__file__), "netcdf_dummy_file.nc")
+        tracemalloc.start()
+        snapshot = tracemalloc.take_snapshot()
+
+        k_times = 10
+        for _k in range(k_times):
+            nc = netCDF4.Dataset(netcdf_file)
+
+            vs = nc.get_variables_by_attributes(axis='Z')
+            self.assertEqual(len(vs), 1)
+
+            vs = nc.get_variables_by_attributes(units='m/s')
+            self.assertEqual(len(vs), 4)
+
+            vs = nc.get_variables_by_attributes(axis='Z', units='m')
+            self.assertEqual(len(vs), 1)
+
+            vs = nc.get_variables_by_attributes(axis=lambda v: v in ['X', 'Y', 'Z', 'T'])
+            self.assertEqual(len(vs), 1)
+
+            vs = nc.get_variables_by_attributes(grid_mapping=lambda v: v is not None)
+            self.assertEqual(len(vs), 12)
+
+            vs = nc.get_variables_by_attributes(grid_mapping=lambda v: v is not None, long_name=lambda v: v is not None and 'Upward (w) velocity' in v)
+            self.assertEqual(len(vs), 1)
+
+            vs = nc.get_variables_by_attributes(units='m/s', grid_mapping=lambda v: v is not None)
+            self.assertEqual(len(vs), 4)
+
+            vs = nc.get_variables_by_attributes(grid_mapping=lambda v: v is not None, long_name='Upward (w) velocity')
+            self.assertEqual(len(vs), 1)
+            nc.close()
+        stats = tracemalloc.take_snapshot().compare_to(snapshot, 'filename')
+        tracemalloc.stop()
+        print("[ Top 10 differences ]")
+        for stat in stats[:10]:
+            print(stat)
+
+if __name__ == '__main__':    
+    unittest.main()