Rui/add windows ci

.github/workflows/wheels.yml

@@ -41,6 +41,10 @@ jobs:
             macos_version: "14"
             platform_id: "macosx_arm64"
 
+          - name: "windows-latest"
+            platform: "windows"
+            platform_id: "win_amd64"      
+
     steps:
       - name: Checkout
         uses: actions/checkout@v4
@@ -50,13 +54,20 @@ jobs:
         with:
           python-version: ${{ matrix.python-version }}
 
-      - name: Install poetry
-        run: pipx install poetry==1.8.3 --python $(which python3)
-        shell: bash
-
-      - name: Test poetry
+      - name: Install Poetry
+        run: python -m pip install poetry==1.8.3
+
+      - name: Set up custom PATH and set py version to cpXYZ [windows]
+        if: ${{matrix.os.platform_id == 'win_amd64'}}
+        shell: pwsh
         run: |
-          poetry run python --version
+          echo "C:\msys64\usr\bin" | Out-File -FilePath $env:GITHUB_PATH -Encoding utf8 -Append
+          echo "C:\msys64\mingw64\bin" | Out-File -FilePath $env:GITHUB_PATH -Encoding utf8 -Append
+          echo "C:\Users\runneradmin\.local\bin" | Out-File -FilePath $env:GITHUB_PATH -Encoding utf8 -Append
+          echo "C:\nasm" | Out-File -FilePath $env:GITHUB_PATH -Encoding utf8 -Append
+          $version = "${{ matrix.python-version }}"
+          $cp_version = "cp$($version -replace '\.', '')"
+          Add-Content -Path $env:GITHUB_ENV -Value "python_cp_version=$cp_version"
 
       - name: Setup pypi for poetry [for releases only]
         if: ${{ startsWith(github.ref, 'refs/tags/') }}
@@ -65,6 +76,7 @@ jobs:
           poetry config pypi-token.pypi ${{ secrets.PYPI_TOKEN }}
 
       - name: Convert python version to cpXYZ
+        if: ${{matrix.os.platform_id != 'win_amd64'}}
         run: |
           version=${{ matrix.python-version }}
           cp_version="cp${version//.}"
@@ -147,20 +159,6 @@ jobs:
           mkdir ./dist
           cp wheelhouse/*.whl ./dist/
 
-      - name: publish wheels (dry run) [macos]
-        if: matrix.os.platform == 'macos'
-        run: |
-          ls -l ./
-          ls -l ./dist
-          poetry publish --dry-run --no-interaction -vvv
-
-      - name: publish wheels (on publishing) [macos]
-        if: ${{ matrix.os.platform == 'macos' && startsWith(github.ref, 'refs/tags/') }}
-        run: |
-          ls -l ./
-          ls -l ./dist
-          poetry publish --no-interaction -vvv
-
       - name: Set up QEMU [linux]
         if: matrix.os.name == 'ubuntu-latest'
         uses: docker/setup-qemu-action@v3
@@ -191,17 +189,69 @@ jobs:
           mkdir ./dist
           cp wheelhouse/*.whl ./dist/
 
-      - name: publish wheels (dry run) [linux]
-        if: matrix.os.platform == 'linux'
+      - name: Set up MSYS2 [windows]
+        if: ${{matrix.os.platform_id == 'win_amd64'}}
+        uses: msys2/setup-msys2@v2
+        with:
+          update: true
+          install: >
+            mingw-w64-x86_64-toolchain 
+            mingw-w64-i686-toolchain
+
+      - name: Set up QPDF external-libs [windows]
+        if: ${{matrix.os.platform_id == 'win_amd64'}}
+        shell: pwsh
+        run: |
+          New-Item -Path 'C:\windows-libs' -ItemType Directory -Force
+          Invoke-WebRequest -Uri 'https://github.com/qpdf/external-libs/releases/download/release-2024-06-07/qpdf-external-libs-bin.zip' -OutFile 'C:\windows-libs\qpdf-external-libs-bin.zip'
+          Expand-Archive -Path 'C:\windows-libs\qpdf-external-libs-bin.zip' -DestinationPath 'C:\windows-libs' -Force
+          Get-ChildItem -Path 'C:\windows-libs' -Recurse
+
+      - name: Set up NASM assembler [windows]
+        if: ${{matrix.os.platform_id == 'win_amd64'}}
+        shell: pwsh
+        run: |
+          New-Item -Path 'C:\nasm' -ItemType Directory -Force
+          Invoke-WebRequest -Uri 'https://fossies.org/windows/misc/nasm-2.16.03-win64.zip/nasm-2.16.03/nasm.exe' -OutFile 'C:\nasm\nasm.exe'
+          nasm -v
+
+      - name: Build wheels
+        if: ${{matrix.os.platform_id == 'win_amd64'}}
+        env:
+          CIBW_PLATFORM: windows
+          CIBW_BUILD_VERBOSITY: 3
+          CIBW_ARCHS: AMD64
+          CIBW_PROJECT_REQUIRES_PYTHON: "~=${{ matrix.python-version }}.0"
+          PKG_CONFIG_PATH: "C:/msys64/mingw64/lib/pkgconfig"
+          PKG_CONFIG_EXECUTABLE: "C:/msys64/usr/bin/pkg-config.exe"
+          CMAKE_PREFIX_PATH: "C:/msys64/mingw64;C:/windows-libs/external-libs"
+          CMAKE_LIBRARY_PATH: "C:/msys64/mingw64/lib;C:/windows-libs/external-libs/lib-mingw64"
+          CMAKE_INCLUDE_PATH: "C:/msys64/mingw64/include;C:/windows-libs/external-libs/include"
+          CMAKE_GENERATOR: "MSYS Makefiles"
+          BUILD_THREADS: 1
+          ASM_NASM: "C:/nasm/nasm.exe"
+        shell: pwsh
+        run: |
+          Remove-Item -Recurse -Force "C:\Strawberry\"
+          $env:CMAKE_ARGS = "-DZLIB_LIBRARY=C:/windows-libs/external-libs/lib-mingw64/libz.a -DZLIB_INCLUDE_DIR=C:/windows-libs/external-libs/include -DJPEG_LIBRARY=C:/windows-libs/external-libs/lib-mingw64/libjpeg.a -DJPEG_INCLUDE_DIR=C:/windows-libs/external-libs/include"
+          poetry install --no-interaction --no-root --only=build
+          poetry run python -m cibuildwheel --output-dir .\wheelhouse
+          Get-ChildItem -Path .\wheelhouse -Filter *.whl | Format-List
+          Get-ChildItem -Path .\wheelhouse -Filter *.whl | ForEach-Object {
+              $file = $_.FullName
+              Write-Output "Inspecting $file"
+              poetry run python -m zipfile --list "$file"
+          }
+          if (-not (Test-Path -Path .\dist)) {
+              New-Item -Path .\dist -ItemType Directory
+          }
+          Copy-Item -Path .\wheelhouse\*.whl -Destination .\dist\
+
+      - name: publish wheels (dry run)
         run: |
-          ls -l ./
-          ls -l ./dist
           poetry publish --dry-run --no-interaction -vvv
 
-      - name: publish wheels (on publishing) [linux]
-        if: ${{ matrix.os.platform == 'linux' && startsWith(github.ref, 'refs/tags/') }}
+      - name: publish wheels (on publishing) [for releases only]
+        if: ${{ startsWith(github.ref, 'refs/tags/') }}
         run: |
-          ls -l ./
-          ls -l ./dist
           poetry publish --no-interaction -vvv
-

CMakeLists.txt

@@ -69,7 +69,15 @@ set(CMAKE_CXX_STANDARD          20)
 set(CMAKE_CXX_STANDARD_REQUIRED ON)
 
 #set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -Wall -Wno-sign-compare -g3 -DROOT_PATH='\"${TOPLEVEL_PREFIX_PATH}\"' ${ENV_ARCHFLAGS}")
-set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -Wall -Wno-sign-compare -O3 -DROOT_PATH='\"${TOPLEVEL_PREFIX_PATH}\"' ${ENV_ARCHFLAGS}")
+#set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -Wall -Wno-sign-compare -O3 -DROOT_PATH='\"${TOPLEVEL_PREFIX_PATH}\"' ${ENV_ARCHFLAGS}")
+
+if (WIN32)
+    set(TEST_PATH "\\\"${TOPLEVEL_PREFIX_PATH}\\\"")
+    add_definitions(-DROOT_PATH="\\\"${TOPLEVEL_PREFIX_PATH}\\\"")
+    set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -Wall -Wno-sign-compare -O3 ${ENV_ARCHFLAGS}")
+else()
+    set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -Wall -Wno-sign-compare -O3 -DROOT_PATH='\"${TOPLEVEL_PREFIX_PATH}\"' ${ENV_ARCHFLAGS}")
+endif()
 
 message(STATUS "cxx-compiler: " ${CMAKE_CXX_COMPILER})
 message(STATUS "cxx-flags   : " ${CMAKE_CXX_FLAGS})
@@ -147,7 +155,7 @@ include(FetchContent)
 FetchContent_Declare(pybind11
   FETCHCONTENT_BASE_DIR extlib_pybind11 
   GIT_REPOSITORY https://github.com/pybind/pybind11.git
-  GIT_TAG        v2.10.0
+  GIT_TAG        v2.13.5
   GIT_SHALLOW    TRUE
 )
 FetchContent_MakeAvailable(pybind11)
@@ -176,4 +184,3 @@ install(TARGETS docling_parse DESTINATION "${TOPLEVEL_PREFIX_PATH}/docling_parse
 
 # do a bunch of result based tests
 # do_test(Tutorial 4 "4 is 2")
-

cmake/extlib_pybind11.cmake

@@ -5,7 +5,7 @@ include(ExternalProject)
 include(CMakeParseArguments)
 
 set(PYBIND11_URL https://github.com/pybind/pybind11.git)
-set(PYBIND11_TAG v2.10.0)
+set(PYBIND11_TAG v2.13.5)
 
 ExternalProject_Add(extlib_pybind11
     PREFIX extlib_pybind11
@@ -25,4 +25,4 @@ ExternalProject_Add(extlib_pybind11
 
 add_library(pybind11 INTERFACE)
 add_custom_target(install_extlib_pybind11 DEPENDS extlib_pybind11)
-add_dependencies(pybind11 install_extlib_pybind11)
+add_dependencies(pybind11 install_extlib_pybind11)

cmake/os_opts.cmake

@@ -4,6 +4,8 @@ message(STATUS "entering in 'os_opts.cmake'")
 if(WIN32)
    message(STATUS "compiling on windows")
 
+   set(LIB_LINK qpdf jpeg utf8 z)
+
 elseif(APPLE)
    message(STATUS "compiling on mac-osx")
 
@@ -27,5 +29,3 @@ elseif(UNIX)
    list(APPEND LIB_LINK dl m pthread rt resolv)
 
 endif()
-
-

src/include_libs.h

@@ -6,3 +6,4 @@
 #include <assert.h>
 #include <filesystem>
 #include <iostream>
+#include <string>

src/proj_folders/pdf_interface/parser.h

@@ -308,7 +308,8 @@ namespace pdf_lib
 
     // FIXME
     //std::string font_data_dir = PDF_DATA_DIR;
-    std::string font_data_dir = resource_utils::get_resources_dir(true);
+    //std::string font_data_dir = resource_utils::get_resources_dir(true);
+    std::string font_data_dir = resource_utils::get_resources_dir(true).string();
 
     std::vector<std::string> font_data_key = {"font-data", "path"};
     if (input.has(font_data_key))

src/proj_folders/pdf_library/core/tools/writer.h

@@ -64,9 +64,11 @@ namespace pdf_lib
 		   std::pair<scalar_type, scalar_type> page_r0,
 		   std::pair<scalar_type, scalar_type> page_dim);      
 
-      template<typename scalar_type, long unsigned int N>
+      // template<typename scalar_type, long unsigned int N>
+      // std::vector<scalar_type> to_vector(std::array<scalar_type, N> arr);
+      template<typename scalar_type, std::size_t N>
       std::vector<scalar_type> to_vector(std::array<scalar_type, N> arr);
-      
+
       template<typename scalar_type>
       std::vector<scalar_type> get_bvec(std::array<scalar_type, 4> rhs);
 
@@ -308,14 +310,24 @@ namespace pdf_lib
       }
     }
 
-    template<typename scalar_type, long unsigned int N>
+  //   template<typename scalar_type, long unsigned int N>
+  //   std::vector<scalar_type> writer::to_vector(std::array<scalar_type, N> arr)
+  //   {
+  //     std::vector<scalar_type> result(N, 0);
+
+  //     for(int l=0; l<N; l++)
+	// result[l] = arr[l];
+
+  //     return result;
+  //   }
+    template<typename scalar_type, std::size_t N>
     std::vector<scalar_type> writer::to_vector(std::array<scalar_type, N> arr)
     {
-      std::vector<scalar_type> result(N, 0);
+      std::vector<scalar_type> result(arr.size(), 0);
+
+      for(std::size_t l = 0; l < arr.size(); ++l)
+        result[l] = arr[l];
 
-      for(int l=0; l<N; l++)
-	result[l] = arr[l];
-
       return result;
     }
 
@@ -745,5 +757,4 @@ namespace pdf_lib
 
 }
 
-#endif
-
+#endif

src/proj_folders/pdf_library/qpdf/parser/cid_cmap.h

@@ -52,7 +52,8 @@ namespace pdf_lib
 
       // FIXME
       //directory = PDF_DATA_DIR;
-      directory = resource_utils::get_resources_dir(true);
+      //directory = resource_utils::get_resources_dir(true);
+	  directory = resource_utils::get_resources_dir(true).string();
 
       directory += directory.back()!='/'? "/":"";
 

src/proj_folders/pdf_library/qpdf/parser/cid_to_utf8.h

@@ -52,7 +52,8 @@ namespace pdf_lib
     {
       // FIXME
       //directory = PDF_DATA_DIR;
-      directory = resource_utils::get_resources_dir(true);
+      //directory = resource_utils::get_resources_dir(true);
+	  directory = resource_utils::get_resources_dir(true).string();
 
       directory += directory.back()!='/'? "/":"";
 

src/pybind/docling_parser.h

@@ -77,7 +77,8 @@ namespace docling
     docling_resources(),
     interface()
   {
-    std::string font_data_dir = resource_utils::get_resources_dir(true);
+    // std::string font_data_dir = resource_utils::get_resources_dir(true);
+    std::string font_data_dir = resource_utils::get_resources_dir(true).string();
 
     pdf_lib::core::object<pdf_lib::core::FONT>::initialize(font_data_dir);
   }

src/pybind/docling_resources.h

@@ -28,7 +28,8 @@ namespace docling
 
   std::string docling_resources::get_resources_path()
   {
-    return resource_utils::get_resources_dir(true);
+    // return resource_utils::get_resources_dir(true);
+    return resource_utils::get_resources_dir(true).string();
   }
 
   bool docling_resources::set_resources_path()

src/utils/logging_library/logger.h

@@ -186,27 +186,50 @@ namespace logging_lib {
     return Logger(domain);
   }
 
+  // std::function<void(std::string, logging_level_type)> Logger::to_file(std::string filename, bool append)
+  // {
+  //   static std::map<std::string, std::ofstream> streams;
+  //   static char buffer[32];
+
+  //   if(not streams.count(filename))
+  //     streams[filename].open(filename, append ? std::ios::app : std::ios::out);
+
+  //   std::ofstream & stream = streams[filename];
+  //   return [&](std::string string, logging_level_type type) { 
+
+  //     auto time = std::chrono::system_clock::now();
+  //     std::time_t t = std::chrono::system_clock::to_time_t(time);
+
+  //     std::tm timeinfo;
+  //     localtime_r(&t, &timeinfo);
+  //     strftime(buffer, 32, "%Y-%m-%d %H:%M:%S", &timeinfo);
+
+  //     stream << buffer << " [" << std::setw(7) << to_string(type) << "] " << string << std::endl;
+  //   };
+  // }
+
   std::function<void(std::string, logging_level_type)> Logger::to_file(std::string filename, bool append)
   {
     static std::map<std::string, std::ofstream> streams;
-    static char buffer[32];
-
-    if(not streams.count(filename))
-      streams[filename].open(filename, append ? std::ios::app : std::ios::out);
-
-    std::ofstream & stream = streams[filename];
     return [&](std::string string, logging_level_type type) { 
-
       auto time = std::chrono::system_clock::now();
       std::time_t t = std::chrono::system_clock::to_time_t(time);
-
-      struct tm * timeinfo;
-      timeinfo = localtime(&t);
-      strftime(buffer,32,"%F %T", timeinfo);
-
-      stream << buffer << " [" << std::setw(7) << to_string(type) << "] " << string << std::endl;
+
+  #ifdef _WIN32
+      struct tm timeinfo;
+      localtime_s(&timeinfo, &t);
+  #else
+      struct tm timeinfo;
+      localtime_r(&t, &timeinfo);
+  #endif
+
+      char buffer[80];
+      strftime(buffer, 80, "%d-%m-%Y %I:%M:%S", &timeinfo);
+      std::string str(buffer);
+
+      streams[filename] << buffer << " [" << std::setw(7) << to_string(type) << "] " << string;
     };
-  }
+  };
 
   void Logger::warn(std::string str, std::string domain)
   {