feat: iteration benchmarks

CMakeLists.txt

@@ -67,6 +67,7 @@ add_executable(bench
   # "${PROJECT_SOURCE_DIR}/bench/node_16.cpp"
   # "${PROJECT_SOURCE_DIR}/bench/node_48.cpp"
   # "${PROJECT_SOURCE_DIR}/bench/node_256.cpp"
+  # "${PROJECT_SOURCE_DIR}/bench/query_iteration.cpp"
   "${PROJECT_SOURCE_DIR}/bench/query_sparse_uniform.cpp"
   "${PROJECT_SOURCE_DIR}/bench/query_sparse_uniform_64.cpp"
   "${PROJECT_SOURCE_DIR}/bench/query_sparse_zipf.cpp"

bench/query_iteration.cpp

@@ -0,0 +1,54 @@
+/**
+ * @file query iteration microbenchmarks
+ * @author Rafael Kallis <rk@rafaelkallis.com>
+ */
+
+#include "art.hpp"
+#include "picobench/picobench.hpp"
+#include "zipf.hpp"
+#include <functional>
+#include <map>
+#include <string>
+#include <unordered_map>
+
+using picobench::state;
+using std::string;
+using std::to_string;
+using std::hash;
+using std::mt19937_64;
+using std::map;
+using std::unordered_map;
+
+PICOBENCH_SUITE("query iteration");
+
+static void full_scan_zipf(state &s) {
+  art::art<int> m;
+  hash<uint32_t> h;
+  int v = 1;
+  int *v_ptr = &v;
+  art::tree_it<int> it, it_end;
+  fast_zipf rng(1000000);
+  for (int i = 0; i < 100000; ++i) {
+    m.set(to_string(h(rng())).c_str(), v_ptr);
+  }
+  for (auto i __attribute__((unused)) : s) {
+    for (it = m.begin(), it_end = m.end(); it != it_end; ++it) {}
+  }
+}
+PICOBENCH(full_scan_zipf).iterations({1000});
+
+static void full_scan_uniform(state &s) {
+  art::art<int> m;
+  hash<uint32_t> h;
+  int v = 1;
+  int *v_ptr = &v;
+  art::tree_it<int> it, it_end;
+  mt19937_64 rng(0);
+  for (int i = 0; i < 100000; ++i) {
+    m.set(to_string(h(rng())).c_str(), v_ptr);
+  }
+  for (auto i __attribute__((unused)) : s) {
+    for (it = m.begin(), it_end = m.end(); it != it_end; ++it) {}
+  }
+}
+PICOBENCH(full_scan_uniform).iterations({1000});