MRG: expose memory information

src/bbhash.pyx

@@ -45,6 +45,7 @@ cdef extern from "BooPHF.h" namespace "boomphf":
     cdef cppclass mphf[T,U]:
         mphf(unsigned long long, vector[T], int, float, bool, bool) except +
         uint64_t lookup(uint64_t)
+        uint64_t totalBitSize()
         void save(ofstream)
         void load(ifstream)
 
@@ -82,6 +83,9 @@ cdef class PyMPHF:
 
         return mp_hashes
 
+    def get_mem(self):
+        return deref(self.c_mphf).totalBitSize()
+
     def save(self, str filename):
         cdef ofstream* outputter
         outputter = new ofstream(filename.encode(), binary)

src/bbhash_table.pyx

@@ -57,6 +57,12 @@ class BBHashTable(object):
 
         self.mphf_to_value[mp_hash] = value
 
+    def get_mem(self):
+        "Get memory usage."
+        return self.mphf.get_mem() + \
+               self.mphf_to_hash.nbytes + \
+               self.mphf_to_value.nbytes
+
     def get_unique_values(self, hashes, require_exist=False):
         "Retrieve unique values for item."
         values = defaultdict(int)