Added a benchmark to the C++ code.

cxxmph/Makefile.am

@@ -1,10 +1,10 @@
-check_PROGRAMS = cmph_hash_map_test mphtable_test trigraph_test bm_urls
-noinst_PROGRAMS = bm_urls
+check_PROGRAMS = cmph_hash_map_test mphtable_test trigraph_test
+noinst_PROGRAMS = bm_numbers bm_urls
 bin_PROGRAMS = cxxmph
 lib_LTLIBRARIES = libcxxmph.la
 include_HEADERS = cmph_hash_map.h mphtable.h MurmurHash2.h trigraph.h cxxmph_hash.h stringpiece.h
 
-libcxxmph_la_SOURCES = MurmurHash2.h trigragh.h trigraph.cc mphtable.h mphtable.cc cxxmph_hash.h stringpiece.h
+libcxxmph_la_SOURCES = MurmurHash2.h trigragh.h trigraph.cc mphtable.h mphtable.cc cxxmph_hash.h stringpiece.h benchmark.h benchmark.cc
 libcxxmph_la_LDFLAGS = -version-info 0:0:0
 
 cmph_hash_map_test_LDADD   = libcxxmph.la
@@ -16,6 +16,9 @@ mphtable_test_SOURCES = mphtable_test.cc
 trigraph_test_LDADD   = libcxxmph.la
 trigraph_test_SOURCES = trigraph_test.cc
 
+bm_numbers_LDADD = libcxxmph.la
+bm_numbers_SOURCES = bm_numbers.cc
+
 bm_urls_LDADD = libcxxmph.la
 bm_urls_SOURCES = bm_urls.cc
 

cxxmph/MurmurHash2.h

@@ -17,6 +17,7 @@
 
 namespace cxxmph {
 
+inline  // not measured, for making compilation easier only
 unsigned int MurmurHash2 ( const void * key, int len, unsigned int seed )
 {
 	// 'm' and 'r' are mixing constants generated offline.

cxxmph/benchmark.cc

@@ -0,0 +1,104 @@
+#include "benchmark.h"
+
+#include <cstring>
+#include <cstdio>
+#include <sys/resource.h>
+
+#include <iostream>
+#include <vector>
+
+using std::cerr;
+using std::endl;
+using std::string;
+using std::vector;
+
+namespace {
+
+/* Subtract the `struct timeval' values X and Y,
+   storing the result in RESULT.
+   Return 1 if the difference is negative, otherwise 0.  */
+int timeval_subtract ( 
+    struct timeval *result, struct timeval *x, struct timeval* y) {
+  /* Perform the carry for the later subtraction by updating y. */
+  if (x->tv_usec < y->tv_usec) {
+    int nsec = (y->tv_usec - x->tv_usec) / 1000000 + 1;
+    y->tv_usec -= 1000000 * nsec;
+    y->tv_sec += nsec;
+  }
+  if (x->tv_usec - y->tv_usec > 1000000) {
+    int nsec = (x->tv_usec - y->tv_usec) / 1000000;
+    y->tv_usec += 1000000 * nsec;
+    y->tv_sec -= nsec;
+  }
+
+  /* Compute the time remaining to wait.
+     tv_usec is certainly positive. */
+  result->tv_sec = x->tv_sec - y->tv_sec;
+  result->tv_usec = x->tv_usec - y->tv_usec;
+
+  /* Return 1 if result is negative. */
+  return x->tv_sec < y->tv_sec;
+}
+
+struct rusage getrusage_or_die() {
+  struct rusage rs;
+  int ret = getrusage(RUSAGE_SELF, &rs);
+  if (ret != 0) {
+    cerr << "rusage failed: " << strerror(errno) << endl;
+    exit(-1);
+  }
+  return rs;
+}
+
+#ifdef HAVE_CXA_DEMANGLE
+string demangle(const string& name) {
+  char buf[1024];
+  unsigned int size = 1024;
+  int status;
+  char* res = abi::__cxa_demangle(
+     name.c_str(), buf, &size, &status);
+  return res;
+}
+#else
+string demangle(const string& name) { return name; }
+#endif
+ 
+
+static vector<cxxmph::Benchmark*> g_benchmarks;
+
+}  // anonymous namespace
+
+namespace cxxmph {
+
+/* static */ void Benchmark::Register(Benchmark* bm) {
+  if (bm->name().empty()) {
+    string name = demangle(typeid(*bm).name());
+     bm->set_name(name);
+  }
+  g_benchmarks.push_back(bm);
+}
+
+/* static */ void Benchmark::RunAll() {
+  for (auto it = g_benchmarks.begin(); it != g_benchmarks.end(); ++it) {
+    (*it)->MeasureRun();
+    delete *it;
+  }
+}
+
+void Benchmark::MeasureRun() {
+  struct rusage begin = getrusage_or_die();
+  Run(iters_);
+  struct rusage end = getrusage_or_die();
+
+  struct timeval utime;
+  timeval_subtract(&utime, &end.ru_utime, &begin.ru_utime);
+  struct timeval stime;
+  timeval_subtract(&stime, &end.ru_stime, &begin.ru_stime);
+
+  printf("Benchmark: %s\n", name().c_str());
+  printf("User time used  : %ld.%06ld\n", utime.tv_sec, utime.tv_usec);
+  printf("System time used: %ld.%06ld\n", stime.tv_sec, stime.tv_usec);
+  printf("\n");
+}
+
+}  // namespace cxxmph

cxxmph/benchmark.h

@@ -0,0 +1,31 @@
+#ifndef __CXXMPH_BENCHMARK_H__
+#define __CXXMPH_BENCHMARK_H__
+
+#include <string>
+#include <typeinfo>
+
+namespace cxxmph {
+
+class Benchmark {
+ public:
+  Benchmark(int iters = 1) : iters_(iters) { }
+  virtual void Run(int iters) = 0;
+  virtual ~Benchmark() { }
+  const std::string& name() { return name_; }
+  void set_name(const std::string& name) { name_ = name; }
+
+  static void Register(Benchmark* bm);
+  static void RunAll();
+
+ protected:
+  int iters() { return iters_; }
+
+ private:
+  int iters_;
+  std::string name_;
+  void MeasureRun();
+};
+
+}  // namespace cxxmph
+
+#endif

cxxmph/bm_numbers.cc

@@ -0,0 +1,52 @@
+#include <set>
+#include <vector>
+
+#include "benchmark.h"
+#include "mphtable.h"
+
+using std::set;
+using std::vector;
+
+namespace cxxmph {
+class BM_NumbersCreate : public Benchmark {
+ public:
+  BM_NumbersCreate(int iters = 1) : Benchmark(iters) {
+    set<int> unique;
+    while (unique.size() < 1000 * 1000) {
+      int v = random();
+      if (unique.find(v) == unique.end()) {
+        unique.insert(v);
+        random_unique_.push_back(v);
+      }
+    }
+  }
+ protected:
+  virtual void Run(int iters) {
+    SimpleMPHTable<int> table;
+    table.Reset(random_unique_.begin(), random_unique_.end());
+  }
+  std::vector<int> random_unique_;
+};
+
+class BM_NumbersFind : public BM_NumbersCreate {
+ public:
+  BM_NumbersFind(int iters) : BM_NumbersCreate(iters) { table_.Reset(random_unique_.begin(), random_unique_.end()); } 
+  virtual void Run(int iters) {
+    for (int i = 0; i < iters * 100; ++i) {
+      int pos = random() % random_unique_.size();;
+      int h = table_.index(pos);
+    }
+  }
+ private:
+   SimpleMPHTable<int> table_;
+};
+
+}  // namespace cxxmph
+
+using namespace cxxmph;
+
+int main(int argc, char** argv) {
+  Benchmark::Register(new BM_NumbersCreate());
+  Benchmark::Register(new BM_NumbersFind(1000 * 1000));
+  Benchmark::RunAll();
+}

cxxmph/bm_urls.cc

@@ -1,21 +1,70 @@
-#include <string>
-#include <iostream>
 #include <fstream>
+#include <iostream>
+#include <set>
+#include <string>
 #include <vector>
+#include <unordered_map>
 
-#include "mphtable.h"
+#include "benchmark.h"
+#include "cmph_hash_map.h"
 
 using std::ifstream;
+using std::set;
 using std::string;
 using std::vector;
-using cxxmph::SimpleMPHTable;
 
-int main(int argc, char** argv) {
+namespace cxxmph {
+
+class BM_UrlsCreate : public Benchmark {
+ public:
+  BM_UrlsCreate(int iters = 1) : Benchmark(iters) {
+    ReadUrls();
+  }
+ protected:
+  virtual void Run(int iters) {
+    BuildTable();
+  }
+  void BuildTable() {
+    for (auto it = urls_.begin(); it != urls_.end(); ++it) {
+      table_[*it] = it - urls_.begin();
+    }
+    table_.pack();
+  }
+  void ReadUrls() {
     vector<string> urls;
-  std::ifstream f("URLS1k");
+    std::ifstream f("URLS100k");
     string buffer;
     while(std::getline(f, buffer)) urls.push_back(buffer);
-
-  SimpleMPHTable<string> table;
-  table.Reset(urls.begin(), urls.end());
+    set<string> unique(urls.begin(), urls.end());
+    if (unique.size() != urls.size()) {
+      cerr << "Input file has repeated keys." << endl;
+      exit(-1);
+    }
+    urls_.swap(urls);
+  }
+  vector<string> urls_;
+  cxxmph::cmph_hash_map<string, int> table_;
+};
+   
+class BM_UrlsFind : public BM_UrlsCreate {
+ public:
+  BM_UrlsFind(int iters = 1) : BM_UrlsCreate(iters) { ReadUrls(); BuildTable(); }
+ protected:
+  virtual void Run(int iters) {
+    for (int i = 0; i < iters * 100; ++i) {
+      int pos = random() % urls_.size();;
+      int h = table_[urls_[pos]];
+      assert(h == pos);
+    }
+  }
+};
+
+}  // namespace cxxmph
+
+using namespace cxxmph;
+
+int main(int argc, char** argv) {
+  Benchmark::Register(new BM_UrlsCreate());
+  Benchmark::Register(new BM_UrlsFind(1000 * 1000));
+  Benchmark::RunAll();
 }

cxxmph/mphtable.h

@@ -85,13 +85,13 @@ bool MPHTable::Reset(ForwardIterator begin, ForwardIterator end) {
   n_ = 3*r_;
   k_ = 1U << b_;
 
-  cerr << "m " << m_ << " n " << n_ << " r " << r_ << endl;
+  // cerr << "m " << m_ << " n " << n_ << " r " << r_ << endl;
 
   int iterations = 10;
   std::vector<TriGraph::Edge> edges;
   std::vector<uint32_t> queue;
   while (1) {
-    cerr << "Iterations missing: " << iterations << endl;
+    // cerr << "Iterations missing: " << iterations << endl;
     for (int i = 0; i < 3; ++i) hash_seed_[i] = random() % m_;
     // for (int i = 0; i < 3; ++i) hash_seed_[i] = random() + i;
     if (Mapping<SeededHashFcn>(begin, end, &edges, &queue)) break;