add test

.gitignore

@@ -3,6 +3,4 @@
 /cpm_modules
 .DS_Store
 .idea
-cmake-*
-
-test/data
+cmake-*

library/CMakeLists.txt → include/CMakeLists.txt

@@ -4,19 +4,19 @@ cmake_minimum_required(VERSION 3.14...3.22)
 
 # Note: update this to your new project's name and version
 project(
-  boss
-  VERSION 1.0.0
-  LANGUAGES CXX
+        boss
+        VERSION 1.0.0
+        LANGUAGES CXX
 )
 
 # ---- Include guards ----
 
-if(PROJECT_SOURCE_DIR STREQUAL PROJECT_BINARY_DIR)
-  message(
-    FATAL_ERROR
-      "In-source builds not allowed. Please make a new directory (called a build directory) and run CMake from there."
-  )
-endif()
+if (PROJECT_SOURCE_DIR STREQUAL PROJECT_BINARY_DIR)
+    message(
+            FATAL_ERROR
+            "In-source builds not allowed. Please make a new directory (called a build directory) and run CMake from there."
+    )
+endif ()
 
 option(BOSS_HAS_BOOST "Do not download Boost" OFF)
 
@@ -33,31 +33,39 @@ include(../cmake/CPM.cmake)
 # PackageProject.cmake will be used to make our target installable
 CPMAddPackage("gh:TheLartians/[email protected]")
 
+CPMAddPackage(
+        GITHUB_REPOSITORY ylab-hi/BINARY
+        VERSION 1.1.0
+        OPTIONS "BINARY_BUILD_STANDALONE NO" "BINARY_BUILD_TESTS NO"
+)
+
+
 find_package(Boost REQUIRED COMPONENTS system iostreams)
 
 # ---- Add source files ----
 
 # Note: globbing sources is considered bad practice as CMake's generators may not detect new files
 # automatically. Keep that in mind when changing files, or explicitly mention them here.
-file(GLOB_RECURSE headers CONFIGURE_DEPENDS "${CMAKE_CURRENT_SOURCE_DIR}/include/*.hpp")
-file(GLOB_RECURSE sources CONFIGURE_DEPENDS "${CMAKE_CURRENT_SOURCE_DIR}/source/*.cpp")
+file(GLOB_RECURSE headers CONFIGURE_DEPENDS "${CMAKE_CURRENT_SOURCE_DIR}/boss/*.hpp")
+file(GLOB_RECURSE sources CONFIGURE_DEPENDS "${CMAKE_CURRENT_SOURCE_DIR}/../source/*.cpp")
 
 # ---- Create library ----
 
 # Note: for header-only libraries change all PUBLIC flags to INTERFACE and create an interface
 # target: add_library(${PROJECT_NAME} INTERFACE)
 add_library(${PROJECT_NAME} ${headers} ${sources})
+
 set_target_properties(${PROJECT_NAME} PROPERTIES CXX_STANDARD 20)
 
 # being a cross-platform target, we enforce standards conformance on MSVC
 target_compile_options(${PROJECT_NAME} PUBLIC "$<$<COMPILE_LANG_AND_ID:CXX,MSVC>:/permissive->")
 
 # Link dependencies
-target_link_libraries(${PROJECT_NAME} PUBLIC spdlog::spdlog Boost::boost Boost::iostreams)
+target_link_libraries(${PROJECT_NAME} PUBLIC spdlog::spdlog Boost::boost Boost::iostreams binary::binary)
 
 target_include_directories(
-  ${PROJECT_NAME} PUBLIC $<BUILD_INTERFACE:${PROJECT_SOURCE_DIR}/include>
-                         $<INSTALL_INTERFACE:include/${PROJECT_NAME}-${PROJECT_VERSION}>
+        ${PROJECT_NAME} PUBLIC $<BUILD_INTERFACE:${PROJECT_SOURCE_DIR}/../include>
+        $<INSTALL_INTERFACE:include/${PROJECT_NAME}-${PROJECT_VERSION}>
 )
 
 # ---- Create an installable target ----
@@ -68,13 +76,13 @@ target_include_directories(
 string(TOLOWER ${PROJECT_NAME}/version.h VERSION_HEADER_LOCATION)
 
 packageProject(
-  NAME ${PROJECT_NAME}
-  VERSION ${PROJECT_VERSION}
-  NAMESPACE ${PROJECT_NAME}
-  BINARY_DIR ${PROJECT_BINARY_DIR}
-  INCLUDE_DIR include
-  INCLUDE_DESTINATION include/${PROJECT_NAME}-${PROJECT_VERSION}
-  VERSION_HEADER "${VERSION_HEADER_LOCATION}"
-  COMPATIBILITY SameMajorVersion
-  DEPENDENCIES "spdlog 1.10.0"
+        NAME ${PROJECT_NAME}
+        VERSION ${PROJECT_VERSION}
+        NAMESPACE ${PROJECT_NAME}
+        BINARY_DIR ${PROJECT_BINARY_DIR}
+        INCLUDE_DIR .
+        INCLUDE_DESTINATION ../include/${PROJECT_NAME}-${PROJECT_VERSION}
+        VERSION_HEADER "${VERSION_HEADER_LOCATION}"
+        COMPATIBILITY SameMajorVersion
+        DEPENDENCIES "spdlog 1.10.0"
 )

include/boss/fq_utils.hpp

@@ -0,0 +1,83 @@
+#ifndef FQ_UTILS_H
+#define FQ_UTILS_H
+
+#include <spdlog/spdlog.h>
+
+#include <boost/iostreams/categories.hpp>  // input_filter_tag
+#include <boost/iostreams/operations.hpp>  // get, put, WOULD_BLOCK
+#include <cstdio>                          // EOF.
+#include <filesystem>
+
+namespace boss::fqsp {
+
+  namespace fs = std::filesystem;
+  enum class FqDirection;
+  class fq_filter;
+
+  enum class FqDirection { Forward, Reverse };
+
+  //--------------Definitions of fq input filter for splitting ncbi fq files----//
+  class fq_filter {
+  public:
+    using char_type = char;
+    using category = boost::iostreams::input_filter_tag;
+
+    explicit constexpr fq_filter(FqDirection direction) : direction_(direction) {}
+
+    template <typename Source> int get(Source& src) {
+      int c;
+      while ((c = boost::iostreams::get(src)) != EOF && c != boost::iostreams::WOULD_BLOCK
+             && is_skip(c))
+        ;
+
+      return get_char(c);
+    }
+
+  private:
+    int get_char(int c) {
+      ++saved_line_count_;
+      return c;
+    }
+
+    [[nodiscard]] bool is_skip(int c) {
+      if (c == '\n') {
+        ++line_count_;  // count lines 0-based index
+        if(saved_line_count_ == 0) return true;
+      }
+
+      if (FqDirection::Forward == direction_) {
+        // forward reads : first 4 lines block start
+        return 1 == ((line_count_ >> 2) & 1);
+      } else {
+        // reverse reads : second 4 lines block start
+        return 0 == ((line_count_ >> 2) & 1);
+      }
+    }
+
+    FqDirection direction_;
+    // line_count_ starts at 1 b
+    // Max line_count_ is  2^64 - 1 = 18,446,744,073,709,551,615
+    uint64_t line_count_{0};
+    uint64_t saved_line_count_{0};
+  };
+
+  namespace details {
+
+    //----------------------------Define helper functions---------------------------//
+    [[nodiscard]] bool is_gzip(const fs::path& input);
+
+    void split_fq_impl_gz(const fs::path& input, FqDirection direction);
+
+    void split_fq_impl_txt(const fs::path& input, FqDirection direction);
+
+  }  // namespace details
+
+  //--------------Definitions of functions for splitting ncbi fq files----//
+
+  void split_fq(const fs::path& input);
+  void test_gz_read(const fs::path& path);
+  void test_gz_write(const fs::path& path);
+
+}  // namespace boss::fqsp
+
+#endif  // FQ_UTILS_H