cache the results of obj_to_dict to speed up RMSWriter

after calling `obj_to_dict` with a species once, the result is saved in a lookup dictionary so that subsequent calls are 'instant'. This costs memory but for an expensive function like this that inherently ends up calling on the same input a lot, it should be worth it.

related comment: ReactionMechanismGenerator#2499 (comment)

rmgpy/yml.py

@@ -34,6 +34,7 @@
 
 import os
 import yaml
+from functools import cache
 
 from rmgpy.chemkin import load_chemkin_file
 from rmgpy.species import Species
@@ -104,7 +105,7 @@ def get_radicals(spc):
     else:
         return spc.molecule[0].multiplicity-1
 
-
+@cache
 def obj_to_dict(obj, spcs, names=None, label="solvent"):
     result_dict = dict()
     if isinstance(obj, Species):