From 1be6d9d2b959a6fdc930bad5612cc60b9889d9e4 Mon Sep 17 00:00:00 2001
From: Zengtudor <zengtudor@outlook.com>
Date: Sun, 20 Oct 2024 12:48:05 +0800
Subject: [PATCH] update

---
 CMakeLists.txt |  7 +++++--
 src/main.cpp   | 34 ++++++++++++++++++++++++++--------
 2 files changed, 31 insertions(+), 10 deletions(-)

diff --git a/CMakeLists.txt b/CMakeLists.txt
index 1dae334..ccb0fbe 100644
--- a/CMakeLists.txt
+++ b/CMakeLists.txt
@@ -27,9 +27,12 @@ foreach(SRC_CPP IN LISTS SRC_CPPS)
 endforeach()
 
 # add_executable(${PROJECT_NAME}_bin ${SRC_HPPS} ${SRC_CPPS})
-
+set(TBB_TEST OFF)
 add_subdirectory(pybind11)
+# add_subdirectory(oneTBB)
 
 pybind11_add_module(${PROJECT_NAME} ${SRC_HPPS} ${SRC_CPPS})
 
-target_compile_definitions(${PROJECT_NAME} PRIVATE DNA_IS_PYBIND)
\ No newline at end of file
+target_compile_definitions(${PROJECT_NAME} PRIVATE DNA_IS_PYBIND)
+
+# target_link_libraries(${PROJECT_NAME} PRIVATE TBB::tbb)
\ No newline at end of file
diff --git a/src/main.cpp b/src/main.cpp
index 7af3ff9..6178a6f 100644
--- a/src/main.cpp
+++ b/src/main.cpp
@@ -10,10 +10,17 @@
 #include <pybind11/stl.h>
 #include <stdexcept>
 #include <string>
+// #define DNA_USE_TBB
+#ifdef DNA_USE_TBB
+    #include <tbb/parallel_for_each.h>
+#endif
 
 namespace py = pybind11;
 static constexpr size_t default_buffer_size{512*1024*1024};
 
+
+
+#define DNA_USE_BUFFER
 static const std::unordered_map<char, char> complement = { //这里使用查表的方式大大提高CPU速度，因为if分支CPU不容易命中缓存，需要使用查表加速
     {'A', 'T'}, {'a', 'T'},
     {'T', 'A'}, {'t', 'A'},
@@ -25,14 +32,22 @@ void reverseComplement(char *begin, char *end)
 {
     //注意end是开区间，不能访问end
     std::reverse(begin, end); //翻转DNA序列
-
-    for (ptrdiff_t i = 0; i < (end - begin); ++i) {
-        // static int _ = (zt::print(NAME_VALUE(omp_get_num_threads()),"\n"),0); // 打印线程数量
-        auto it = complement.find(begin[i]);
-        if (it != complement.end()) {
-            begin[i] = it->second;
+    #ifndef DNA_USE_TBB
+        for (ptrdiff_t i = 0; i < (end - begin); ++i) {
+            // static int _ = (zt::print(NAME_VALUE(omp_get_num_threads()),"\n"),0); // 打印线程数量
+            auto it = complement.find(begin[i]);
+            if (it != complement.end()) {
+                begin[i] = it->second;
+            }
         }
-    }
+    #else
+        tbb::parallel_for_each(begin,end,[](char &c){
+            const auto it = complement.find(c);
+            if (it != complement.end()) {
+                c = it->second;
+            }
+        });
+    #endif
 }
 
 void convert_from_file(
@@ -42,7 +57,7 @@ void convert_from_file(
 )
 {
     // std::iostream::sync_with_stdio(false);
-
+    
     const size_t max_size_pre_dna{(size_t)5e4+5};
     const size_t all_buf_size = {buffer_size};
     const size_t read_bufsize{all_buf_size/2},write_bufsize{all_buf_size/2}; 
@@ -57,9 +72,12 @@ void convert_from_file(
     std::ofstream ofs(destination_file);
     if(ofs.is_open()==false)throw std::runtime_error(fmt("Cannot open output file stream\nfilename: ",destination_file,'\n'));
     std::cout<<"Open file ok ,getting memory\n";
+
+    #ifdef DNA_USE_BUFFER
     std::vector<char> read_buf(read_bufsize), write_buf(write_bufsize);
     ifs.rdbuf()->pubsetbuf(read_buf.data(), read_buf.size());
     ofs.rdbuf()->pubsetbuf(write_buf.data(), write_buf.size());
+    #endif
 
     std::array<char, max_size_pre_dna> dna_buf;
     bool is_dna_line{false};