Fix clang builds

.clangd

@@ -0,0 +1,43 @@
+If:
+  PathMatch: .*\.in
+
+CompileFlags:
+  Remove:
+    - -Wall
+    - -Wextra
+    - -Wpedantic
+
+Diagnostics:
+  Suppress:
+    - "*"
+
+---
+
+CompileFlags:
+  Add:
+    - -Wall
+    - -Wextra
+    - -Wpedantic
+
+    # The empty compile_flags.txt in the root directory causes clangd to treat
+    # these paths as relative to the root of the project when compile_commands.json
+    # is not present.
+    - -Ibuild/.dep-cache/utpp-repo-src/include/.
+    - -Iinclude/.
+    - -Itests/.
+
+    - -xc++  # Force .h files to be treated as C++
+    # - -std=gnu++17
+    - -std=c++20
+    - --target=x86_64-pc-windows-msvc
+    - -DUNICODE
+    - -D_UNICODE
+    - -D_DEBUG
+    - -D_DLL
+    - -D_MT
+    - -Xclang
+    - --dependent-lib=msvcrtd
+    - -g
+    - -Xclang
+    - -gcodeview
+    - -fno-char8_t

.cspell.jsonc

@@ -0,0 +1,50 @@
+{
+  "version": "0.2",
+  "ignorePaths": [
+    ".cspell.jsonc",
+    ".git/**",
+    "*.code-workspace",
+    ".local-gitignore",
+    "data/UnicodeData.txt"
+  ],
+  "dictionaryDefinitions": [
+    {
+      "name": "project-words",
+      "path": ".vscode/ltex.dictionary.en-US.txt",
+      "addWords": true
+    }
+  ],
+  "dictionaries": [
+    "bash",
+    "win32",
+    "scientific-terms-us",
+    "project-words"
+  ],
+  "ignoreWords": [
+    "ăâățî",
+    "ĂÂȚÎ",
+    "BCDEFGH",
+    "MIRCEANEACȘUĂÂȚÎ",
+    "ȚEPUȘ",
+    "αλφάβητο",
+    "αξία",
+    "αρχείο",
+    "ελληνικό",
+    "ΚΛΕΙΔΙ",
+    "Հայերեն",
+    "पंजाबी",
+    "ᓀᐦᐃᔭᐍᐏᐣ",
+    "Ελληνικός",
+    "العربي",
+    "اللغة",
+    "ܐܪܡܝܐ"
+  ],
+  "ignoreRegExpList": [
+    "/-W.*/",
+    "/-D.*/",
+    "/-X.*/",
+    "/-I.*/"
+  ],
+  "allowCompoundWords": true,
+  "words": []
+}

.github/workflows/build.yml

@@ -11,25 +11,42 @@ permissions:
 jobs:
   build:
     runs-on: windows-latest
-
-    steps:      
+
+    steps:
+      - name: Checkout code
+        uses: actions/checkout@v4
+
       - name: Get CPM
         uses: neacsum/configurator@v0.0.11
         with:
           name: cpm.exe
           url: https://github.com/neacsum/cpm/releases/latest/download/cpm.exe
-
-      - name: Build libraries
-        run: cpm -v --proto https -u https://github.com/neacsum/utf8.git -r $HOME utf8
-
+
+      - name: Create dependencies directory
+        shell: cmd
+        run: |
+          if not exist "build\.dep-cache" mkdir "build\.dep-cache"
+
+      - name: Download dependencies
+        run: cpm -v -f --proto https -r build\.dep-cache
+
+      - name: Copy utpp directory
+        shell: cmd
+        run: |
+          xcopy /E /I build\.dep-cache\utpp build\.dep-cache\utpp-repo-src
+
+      - name: Build library (local version, not from CPM)
+        shell: cmd
+        run: |
+          build.bat lib
+
       - name: Build and run tests
         shell: cmd
         run: |
-          %USERPROFILE%\utf8\build.bat tests
+          build.bat tests
 
       - name: Save tests result
         uses: actions/upload-artifact@v4
         with:
           name: test_results
-          path: ~/utf8/build/exe/x64/debug/utf8_tests.xml
-
+          path: build/exe/x64/debug/utf8_tests.xml

.github/workflows/doxygen.yml

@@ -26,11 +26,11 @@ jobs:
 
       - name: Generate Doxygen documentation
         run: doxygen tools/doxygen/Doxyfile
-      
+
       - name: Upload pages
         uses: actions/upload-pages-artifact@v3
         with:
           path: docs
-      
+
       - name: Deploy pages
         uses: actions/deploy-pages@v4

.github/workflows/purge.yml

@@ -5,11 +5,11 @@ on:
       days:
         description: 'Number of days.'
         required: true
-        default: 30
+        default: '30'
       minimum_runs:
         description: 'The minimum runs to keep for each workflow.'
         required: true
-        default: 6
+        default: '6'
       delete_workflow_pattern:
         description: 'The name or filename of the workflow. if not set then it will target all workflows.'
         required: false

.gitignore

@@ -10,8 +10,15 @@
 
 /build
 /.vs*
+!/.vscode/
+/.vscode/*
+!/.vscode/ltex.dictionary.en-US.txt
+!/.vscode/ltex.hiddenFalsePositives.en-US.txt
 /lib
 /.editorconfig
 /utf8.cppcheck
 /docs
 /.DS_Store
+.cache/
+/compile_commands.json
+.dep-cache/

.vscode/ltex.dictionary.en-US.txt

@@ -0,0 +1,64 @@
+asmx
+basecvt
+casecvt
+clangd
+codept
+cppcheck
+endl
+exetest
+fdat
+gclef
+gcodeview
+icompare
+ifndef
+ifstream
+ipch
+IWYU
+ized
+Kaspersky
+keyxx
+lfont
+ltex
+Mattraks
+mfcribbon
+msbuild
+msvc
+msvcrtd
+NDEBUG
+Neacsu
+neacșu
+neacsum
+ofstream
+opensdf
+resx
+sectionxx
+stdcpp
+SYMED
+usebackq
+utpp
+vect
+vsmsbuildcmd
+VSWHERE
+wcmd
+wdat
+wdata
+wdir
+wdrive
+wemoji
+wfile
+wfname
+wfull
+winutf
+wnam
+wnew
+wpath
+wpfx
+wptr
+wrel
+wrhs
+wsmiley
+wsubkey
+wval
+wvalue
+wvar
+xcopy

.vscode/ltex.hiddenFalsePositives.en-US.txt

@@ -0,0 +1,2 @@
+{"rule":"ADD_AN_ADDITIONAL","sentence":"^\\QThe C++20 standard has added an additional type \\E(?:Dummy|Ina|Jimmy-|Dummy-|Maniquí-|Maniquíes-)[0-9]+\\Q, designed to keep UTF-8 encoded characters, and a string type \\E(?:Dummy|Ina|Jimmy-|Dummy-|Maniquí-|Maniquíes-)[0-9]+\\Q.\\E$"}
+{"rule":"UPPERCASE_SENTENCE_START","sentence":"^\\Qwith the implied assumption that all \\E(?:Dummy|Ina|Jimmy-|Dummy-|Maniquí-|Maniquíes-)[0-9]+\\Q strings are UTF-8 encoded character strings.\\E$"}

BUILD.bat

@@ -23,7 +23,7 @@ echo Visual studio installation folder is: %VSInstallDir%
 call "%VSInstallDir%\common7\tools\vsmsbuildcmd.bat"
 
 rem
-rem Build tragets. Valid targets are "lib" and "tests"
+rem Build targets. Valid targets are "lib" and "tests"
 rem Default is to build all
 rem
 if "%~1"=="" (msbuild "%~dp0build.proj") else (msbuild -target:%1 "%~dp0build.proj")

CMakeLists.txt

@@ -1,4 +1,4 @@
-cmake_minimum_required(VERSION 3.13)
+cmake_minimum_required(VERSION 4.0)
 project(utf8)
 
 if (WIN32)
@@ -7,17 +7,42 @@ else()
 set(pfx "")
 endif()
 
-add_compile_options("$<$<C_COMPILER_ID:MSVC>:/utf-8>")
-add_compile_options("$<$<CXX_COMPILER_ID:MSVC>:/utf-8>")
-add_definitions(-DUNICODE -D_UNICODE)
+add_compile_definitions(_UNICODE UNICODE)
+if(MSVC)
+  add_compile_options(/utf-8)
+endif()
+
+# https://wg21.link/p2513
+if(MSVC)
+  add_compile_options(/Zc:char8_t-)
+else()
+  add_compile_options(-fno-char8_t)
+endif()
+
+## configure CMake module search paths that depend on the project
+## proj src dir is location of this file
+list(APPEND CMAKE_MODULE_PATH ${PROJECT_SOURCE_DIR}/cmake)
+include(ConfigFetchContent)
+include(CompileCommands)
+# include dependencies
+include(utpp)
 
 add_subdirectory(tools/gen_casetab)
+
+add_library(${PROJECT_NAME})
+set_target_properties(${PROJECT_NAME} PROPERTIES
+	ARCHIVE_OUTPUT_DIRECTORY ${PROJECT_SOURCE_DIR}/lib/${pfx}/$<CONFIG>
+  CXX_STANDARD 17
+)
 add_subdirectory(src)
+add_subdirectory(include)
 
+set(BUILD_TESTS FALSE CACHE BOOL "Build tests")
 if (BUILD_TESTS)
   add_subdirectory(tests)
 endif ()
 
+set(BUILD_EXAMPLES FALSE CACHE BOOL "Build examples")
 if (BUILD_EXAMPLES)
   add_subdirectory(examples)
 endif ()