diff --git a/.gitignore b/.gitignore
index cd3cbc0..8eb07c9 100644
--- a/.gitignore
+++ b/.gitignore
@@ -14,3 +14,6 @@ junk/
 # linux binaries
 bin/freq*
 bin/hack*
+
+# osx trash
+*.dSYM
diff --git a/build/hopscotch/.gitignore b/build/hopscotch/.gitignore
new file mode 100644
index 0000000..28bcef5
--- /dev/null
+++ b/build/hopscotch/.gitignore
@@ -0,0 +1,12 @@
+CMakeLists.txt.user
+CMakeCache.txt
+CMakeFiles
+CMakeScripts
+Testing
+Makefile
+cmake_install.cmake
+install_manifest.txt
+compile_commands.json
+CTestTestfile.cmake
+_deps
+freq
\ No newline at end of file
diff --git a/build/hopscotch/CMakeLists.txt b/build/hopscotch/CMakeLists.txt
new file mode 100644
index 0000000..8be7b30
--- /dev/null
+++ b/build/hopscotch/CMakeLists.txt
@@ -0,0 +1,13 @@
+cmake_minimum_required(VERSION 3.2)
+
+project(freq)
+
+set(CMAKE_CXX_STANDARD 11)
+set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -std=c++17 -O3")
+
+set(CMAKE_INSTALL_PREFIX ${PROJECT_SOURCE_DIR})
+
+add_subdirectory(hopscotch-map)
+
+add_executable(freq src/main.cpp)
+target_link_libraries(freq PRIVATE tsl::hopscotch_map)
diff --git a/build/hopscotch/hopscotch-map b/build/hopscotch/hopscotch-map
new file mode 160000
index 0000000..29030f5
--- /dev/null
+++ b/build/hopscotch/hopscotch-map
@@ -0,0 +1 @@
+Subproject commit 29030f55ca518bb1be5113ab0a8e134772024a9d
diff --git a/build/hopscotch/src/main.cpp b/build/hopscotch/src/main.cpp
new file mode 100644
index 0000000..d64ea52
--- /dev/null
+++ b/build/hopscotch/src/main.cpp
@@ -0,0 +1,165 @@
+/*
+
+  clang++ -std=c++ -O3 src/freq04.cpp -o freq
+
+*/
+
+#include <cstdint>
+#include <algorithm>
+#include <string>
+#include <functional>
+#include <tsl/hopscotch_map.h>
+#include <tsl/hopscotch_set.h>
+
+#include <iostream>
+#include <fcntl.h> // open
+
+#if defined(_WIN32) || defined(WIN32)
+// windows
+#include "ext/windows-mmap.h"
+#include <io.h>
+#define lseek64 _lseeki64
+#pragma warning(disable : 4996)
+#else
+// unix
+#include <sys/mman.h> // mmap
+#include <unistd.h> // lseek
+#endif
+
+#if __APPLE__
+#define lseek64 lseek
+#endif
+
+#if __linux__
+#include <linux/version.h>
+#if LINUX_VERSION_CODE > KERNEL_VERSION(2,6,22)
+#define _MAP_POPULATE_AVAILABLE
+#endif
+#endif
+
+#ifdef _MAP_POPULATE_AVAILABLE
+#define MMAP_FLAGS (MAP_PRIVATE | MAP_POPULATE)
+#else
+#define MMAP_FLAGS MAP_PRIVATE
+#endif
+
+typedef unsigned char byte;
+typedef unsigned long long ullong;
+
+class Word {
+  public:
+  Word(const std::string &s, int i) : text(s), occurances(i) {}
+
+  std::string text;
+  int occurances;
+};
+
+int main(int argc, char **argv) {
+  if (argc != 3) {
+    exit(1);
+  }
+
+  int fd = open(argv[1], O_RDONLY);
+  if (fd == -1) {
+    std::cerr << "Can't open file" << std::endl;
+    exit(1);
+  }
+
+  FILE *out = fopen(argv[2], "wb+");
+  if (!out) {
+    std::cerr << "failed to open output file" << std::endl;
+    exit(1);
+  }
+
+  const size_t fsz = lseek(fd, 0, SEEK_END);
+  if (fsz == -1) {
+    std::cerr << "Failed to calculate file size" << std::endl;
+    exit(1);
+  }
+
+
+  const uint8_t *begin = reinterpret_cast<const uint8_t *>(
+    mmap(NULL, fsz, PROT_READ, MMAP_FLAGS, fd, 0));
+
+  if (begin == nullptr || begin == MAP_FAILED) {
+    std::cerr << "Unable to mmap" << std::endl;
+    exit(1);
+  }
+
+  std::cout << "file mmap'ed" << std::endl;
+
+  char letters[256];
+  memset(letters, 0, sizeof(letters));
+  for (size_t c = 0; c < 256; ++c) {
+    if (c >= 'a' and c <= 'z') {
+      letters[c] = c;
+    }
+    if (c >= 'A' and c <= 'Z') {
+      letters[c] = c - 'A' + 'a';
+    }
+  }
+
+  std::cout << "letters: " <<  std::string(letters, 256) << std::endl;
+
+  const char *start = nullptr;
+  size_t len = 0;
+
+  /*
+   * Calculating the hash and comparing two std::string may be slow.
+   * We can store the hash of each std::string in the hash map to make
+   * the inserts and lookups faster by setting StoreHash to true.
+   */
+  tsl::hopscotch_map<std::string, int, std::hash<std::string>,
+                     std::equal_to<std::string>,
+                     std::allocator<std::pair<std::string, int>>,
+                     30, true> map;
+
+  auto end = begin + fsz;
+  auto word = std::string(start, len);
+
+  for (auto s = begin; s != end; ++s) {
+    const auto ch = letters[*s];
+
+    if (ch) {
+      if (not start) {
+        start = reinterpret_cast<const char*>(s);
+      }
+      ++len;
+      continue;
+    } else if (len == 0) {
+      continue;
+    } else {
+      word.assign(start, len);
+
+      for (size_t i = 0; i< len; ++i) {
+        word[i] = letters[word[i]];
+      }
+
+      if (map.find(word) != map.end()) {
+        map[word]++;
+      } else {
+        map[word] = 1;
+      }
+    }
+
+    len = 0;
+    start = nullptr;
+  }
+
+  std::vector<Word> words;
+  words.reserve(map.size());
+
+  for(const auto& key_value : map) {
+    // TODO string copied here?..
+    words.emplace_back(Word(key_value.first, key_value.second));
+  }
+
+
+  std::sort(words.begin(), words.end(), [](const Word& a, const Word& b) -> bool {
+    return a.occurances > b.occurances;
+  });
+
+  for (const auto& word : words) {
+    fprintf(out, "%d %s\n", word.occurances, word.text.c_str());
+  }
+}
diff --git a/src/freq01.cpp b/src/freq01.cpp
index 3b301f0..7ab99f5 100644
--- a/src/freq01.cpp
+++ b/src/freq01.cpp
@@ -22,6 +22,23 @@
 #include <unistd.h>
 #endif
 
+#if __APPLE__
+#define lseek64 lseek
+#endif
+
+#if __linux__
+#include <linux/version.h>
+#if LINUX_VERSION_CODE > KERNEL_VERSION(2,6,22)
+#define _MAP_POPULATE_AVAILABLE
+#endif
+#endif
+
+#ifdef _MAP_POPULATE_AVAILABLE
+#define MMAP_FLAGS (MAP_PRIVATE | MAP_POPULATE)
+#else
+#define MMAP_FLAGS MAP_PRIVATE
+#endif
+
 constexpr uint32_t H = 2166136261;
 
 int usage(char *process_name) {
@@ -88,8 +105,14 @@ int main(int argc, char **argv) {
 	}
 
 	const size_t fsz = lseek64(fd, 0, SEEK_END);
+
+	if (fsz == -1) {
+		std::cerr << "Failed to calculate file size" << std::endl;
+		exit(1);
+	}
+
 	const uint8_t *begin = reinterpret_cast<const uint8_t *>(
-		mmap(NULL, fsz, PROT_READ, MAP_PRIVATE | MAP_POPULATE, fd, 0));
+		mmap(NULL, fsz, PROT_READ, MMAP_FLAGS, fd, 0));
 
 	char letters[256];
 	for (size_t i = 0; i < 256; ++i) {
diff --git a/src/freq03.cpp b/src/freq03.cpp
index a3e58a3..a4507e8 100644
--- a/src/freq03.cpp
+++ b/src/freq03.cpp
@@ -16,6 +16,23 @@
 #define __forceinline __attribute__((always_inline)) inline
 #endif
 
+#if __APPLE__
+#define lseek64 lseek
+#endif
+
+#if __linux__
+#include <linux/version.h>
+#if LINUX_VERSION_CODE > KERNEL_VERSION(2,6,22)
+#define _MAP_POPULATE_AVAILABLE
+#endif
+#endif
+
+#ifdef _MAP_POPULATE_AVAILABLE
+#define MMAP_FLAGS (MAP_PRIVATE | MAP_POPULATE)
+#else
+#define MMAP_FLAGS MAP_PRIVATE
+#endif
+
 typedef unsigned char byte;
 typedef unsigned int uint;
 typedef unsigned long long ullong;
@@ -184,7 +201,7 @@ int main(int argc, char **argv) {
 
 	const ullong fsz = lseek64(fd, 0, SEEK_END);
 	const byte *fbegin =
-		(const byte *)mmap(NULL, fsz, PROT_READ, MAP_PRIVATE | MAP_POPULATE, fd, 0);
+		(const byte *)mmap(NULL, fsz, PROT_READ, MMAP_FLAGS, fd, 0);
 
 	static byte lc[256];
 	memset(lc, 0, sizeof(lc));