From 59d56923f5678e9bb9881f4a91265d99347c202f Mon Sep 17 00:00:00 2001
From: greg7mdp <greg7mdp@gmail.com>
Date: Sun, 9 Apr 2023 18:48:26 -0400
Subject: [PATCH] Add example mt_word_counter

---
 CMakeLists.txt                    |  1 +
 examples/mt_word_counter.cc       | 83 +++++++++++++++++++++++++++++++
 parallel_hashmap/phmap_fwd_decl.h | 32 ++++++++++++
 3 files changed, 116 insertions(+)
 create mode 100644 examples/mt_word_counter.cc

diff --git a/CMakeLists.txt b/CMakeLists.txt
index b40eb4a..7b0a6ea 100644
--- a/CMakeLists.txt
+++ b/CMakeLists.txt
@@ -195,6 +195,7 @@ if (PHMAP_BUILD_EXAMPLES)
     add_executable(ex_dump_load examples/dump_load.cc phmap.natvis)
     add_executable(ex_btree examples/btree.cc phmap.natvis)
     add_executable(ex_matt examples/matt.cc phmap.natvis)
+    add_executable(ex_mt_word_counter examples/mt_word_counter.cc phmap.natvis)
 
     target_link_libraries(ex_knucleotide Threads::Threads)
     target_link_libraries(ex_bench Threads::Threads)
diff --git a/examples/mt_word_counter.cc b/examples/mt_word_counter.cc
new file mode 100644
index 0000000..aad7124
--- /dev/null
+++ b/examples/mt_word_counter.cc
@@ -0,0 +1,83 @@
+#include <iostream>
+#include <fstream>
+#include <sstream>
+#include <parallel_hashmap/phmap.h>
+#include <parallel_hashmap/btree.h>
+#include <thread>
+#include <array>
+#include <vector>
+#include <algorithm>
+#include <cstdlib>
+
+
+/*
+ * count the number of occurrences of each word in a large text file using multiple threads
+ */
+
+int main() {
+    // download Jane Austin "Pride and Prejudice"
+    // ------------------------------------------
+    if (system("wget https://www.gutenberg.org/files/1342/1342-0.txt") != 0) {
+        std::cout << "Error: could not retrieve test file https://www.gutenberg.org/files/1342/1342-0.txt\n";
+        return 1;
+    }
+    
+    const std::string filename = "1342-0.txt";
+    
+    constexpr int num_threads = 4;
+    std::vector<std::thread> threads;
+    std::array<std::vector<std::string>, num_threads> lines_array;
+
+    {
+        // populate 4 vectors with lines from the book
+        std::ifstream file(filename);
+        if (!file.is_open()) {
+            std::cout << "Error: could not open file " << filename << std::endl;
+            return 1;
+        }
+        int line_idx = 0;
+        std::string line;
+        while (std::getline(file, line)) {
+            lines_array[line_idx % num_threads].push_back(std::move(line));
+            ++line_idx;
+        }
+    }
+
+    using Map = phmap::parallel_flat_hash_map_m<std::string, int>; // parallel_flat_hash_map_m has default internal mutex
+    Map word_counts;
+
+    // run 4 threads, each thread processing lines from one of the vectors
+    // -------------------------------------------------------------------
+    for (int i = 0; i < num_threads; ++i) {
+        threads.emplace_back(
+            [&word_counts](std::vector<std::string>&& lines) {
+                for (auto& line : lines) {
+                    std::replace_if(line.begin(), line.end(), [](char c) -> bool { return !std::isalnum(c); }, ' ');
+                    std::istringstream iss(line);
+                    std::string word;
+                    while (iss >> word) {
+                        // use lazy_emplace to modify the map while the mutex is locked
+                        word_counts.lazy_emplace_l(word,
+                                                   [&](Map::value_type& p) { ++p.second; },  // called only when key was already present
+                                                   [&](const Map::constructor& ctor) // construct value_type in place when key not present
+                                                   { ctor(std::move(word), 1); } );
+                    }
+                }
+            },
+            std::move(lines_array[i]));
+    }
+    
+    for (auto& thread : threads)
+        thread.join();
+
+    // print one word used at each frequency
+    // -------------------------------------
+    phmap::btree_map<int, std::string> result;
+    for (const auto& pair : word_counts)
+        result[pair.second] = pair.first;
+
+    for (const auto& p : result) 
+        std::cout << p.first << ": " << p.second << std::endl;
+
+    return 0;
+}
diff --git a/parallel_hashmap/phmap_fwd_decl.h b/parallel_hashmap/phmap_fwd_decl.h
index a7719c4..c625be1 100644
--- a/parallel_hashmap/phmap_fwd_decl.h
+++ b/parallel_hashmap/phmap_fwd_decl.h
@@ -20,6 +20,7 @@
 
 #include <memory>
 #include <utility>
+#include <mutex>
 
 #if defined(PHMAP_USE_ABSL_HASH) && !defined(ABSL_HASH_HASH_H_)
     namespace absl { template <class T> struct Hash; };
@@ -127,6 +128,37 @@ namespace phmap {
               class Mutex = phmap::NullMutex>   // use std::mutex to enable internal locks
         class parallel_node_hash_map;
 
+    // -----------------------------------------------------------------------------
+    // phmap::parallel_*_hash_* using std::mutex by default
+    // -----------------------------------------------------------------------------
+    template <class T,
+              class Hash  = phmap::priv::hash_default_hash<T>,
+              class Eq    = phmap::priv::hash_default_eq<T>,
+              class Alloc = phmap::priv::Allocator<T>,
+              size_t N    = 4>
+    using parallel_flat_hash_set_m = parallel_flat_hash_set<T, Hash, Eq, Alloc, N, std::mutex>;
+
+    template <class K, class V,
+              class Hash  = phmap::priv::hash_default_hash<K>,
+              class Eq    = phmap::priv::hash_default_eq<K>,
+              class Alloc = phmap::priv::Allocator<phmap::priv::Pair<const K, V>>,
+              size_t N    = 4>
+    using parallel_flat_hash_map_m = parallel_flat_hash_map<K, V, Hash, Eq, Alloc, N, std::mutex>;
+
+    template <class T,
+              class Hash  = phmap::priv::hash_default_hash<T>,
+              class Eq    = phmap::priv::hash_default_eq<T>,
+              class Alloc = phmap::priv::Allocator<T>,
+              size_t N    = 4>
+    using parallel_node_hash_set_m = parallel_node_hash_set<T, Hash, Eq, Alloc, N, std::mutex>;
+
+    template <class K, class V,
+              class Hash  = phmap::priv::hash_default_hash<K>,
+              class Eq    = phmap::priv::hash_default_eq<K>,
+              class Alloc = phmap::priv::Allocator<phmap::priv::Pair<const K, V>>,
+              size_t N     = 4>
+    using parallel_node_hash_map_m = parallel_node_hash_map<K, V, Hash, Eq, Alloc, N, std::mutex>;
+
     // ------------- forward declarations for btree containers ----------------------------------
     template <typename Key, typename Compare = phmap::Less<Key>,
               typename Alloc = phmap::Allocator<Key>>