Compile time checks for size of UTF-16 code units

Use ARCH_INDEPENDENT with write_basic_package_version_file

.github/workflows/cmake-multi-platform.yml

@@ -45,7 +45,7 @@ jobs:
             c_compiler: cl
 
     steps:
-    - uses: actions/checkout@v3
+    - uses: actions/checkout@v4
 
     - name: Set reusable strings
       # Turn repeated input strings (such as the build output directory) into step outputs. These step outputs can be used throughout the workflow file.

CMakeLists.txt

@@ -1,6 +1,9 @@
-cmake_minimum_required (VERSION 3.5...3.27)
+# This file is deprecated and will be removed in a future release
+# Please see the instructions for installation in README.md file
+
+cmake_minimum_required (VERSION 3.14...3.27)
 project (utf8cpp 
-         VERSION 4.0.3
+         VERSION 4.0.6
          LANGUAGES CXX
          DESCRIPTION "C++ portable library for working with utf-8 encoding")
 
@@ -18,6 +21,7 @@ write_basic_package_version_file(
     "${PROJECT_BINARY_DIR}/${PROJECT_NAME}ConfigVersion.cmake"
     VERSION ${PROJECT_VERSION}
     COMPATIBILITY SameMajorVersion
+    ARCH_INDEPENDENT
 )
 
 install(TARGETS ${PROJECT_NAME}

README.md

@@ -112,8 +112,13 @@ I came up with a small, C++98 compatible generic library in order to handle UTF-
 <!-- TOC --><a name="installation"></a>
 ## Installation
 
-The recommended way to use the library is to download an official release and copy the content of source directory into location of your project's header files.
-If you use CMake for your builds, I still recommend just copying the files into your project, but if you want you can use the CMakeList.txt file included in the project.
+This is a header-only library and the supported way of deploying it is:
+- Download a release from https://github.com/nemtrif/utfcpp/releases into a temporary directory
+- Unzip the release
+- Copy the content of utfcpp/source file into the directory where you keep include files for your project
+
+
+The CMakeList.txt file was originally made for testing purposes only, but unfortunately over time I accepted contributions that added install target. *This is not a supported way of installing the utfcpp library* and I am considering removing the CMakeList.txt in a future release.
 
 <!-- TOC --><a name="examples-of-use"></a>
 ## Examples of use
@@ -834,7 +839,7 @@ assert (utf16result[2] == 0xd834);
 assert (utf16result[3] == 0xdd1e);
 ```
 
-In case of an invalid UTF-8 seqence, a `utf8::invalid_utf8` exception is thrown.
+In case of an invalid UTF-8 sequence, a `utf8::invalid_utf8` exception is thrown.
 
 
 <!-- TOC --><a name="stdu16string-utf8to16stdu8string_view-s"></a>
@@ -862,7 +867,7 @@ assert (utf16result[2] == 0xd834);
 assert (utf16result[3] == 0xdd1e);
 ```
 
-In case of an invalid UTF-8 seqence, a `utf8::invalid_utf8` exception is thrown.
+In case of an invalid UTF-8 sequence, a `utf8::invalid_utf8` exception is thrown.
 
 <!-- TOC --><a name="utf8utf32to8"></a>
 #### utf8::utf32to8
@@ -1078,7 +1083,7 @@ u32string utf32result = utf8to32(twochars);
 assert (utf32result.size() == 2);
 ```
 
-In case of an invalid UTF-8 seqence, a `utf8::invalid_utf8` exception is thrown.
+In case of an invalid UTF-8 sequence, a `utf8::invalid_utf8` exception is thrown.
 
 
 <!-- TOC --><a name="stdu32string-utf8to32const-stdu8string_view-s"></a>
@@ -1104,7 +1109,7 @@ u32string utf32result = utf8to32(stringview);
 assert (utf32result.size() == 2);
 ```
 
-In case of an invalid UTF-8 seqence, a `utf8::invalid_utf8` exception is thrown.
+In case of an invalid UTF-8 sequence, a `utf8::invalid_utf8` exception is thrown.
 
 
 <!-- TOC --><a name="stdu32string-utf8to32const-stdstring-s"></a>
@@ -1639,7 +1644,7 @@ class iterator;
 <!-- TOC --><a name="member-functions"></a>
 ##### Member functions
 
-`iterator();` the deafult constructor; the underlying octet_iterator is constructed with its default constructor.
+`iterator();` the default constructor; the underlying octet_iterator is constructed with its default constructor.
 
 `explicit iterator (const octet_iterator& octet_it, const octet_iterator& range_start, const octet_iterator& range_end);` a constructor that initializes the underlying octet_iterator with octet_it and sets the range in which the iterator is considered valid.
 
@@ -2081,7 +2086,7 @@ class iterator;
 <!-- TOC --><a name="member-functions-1"></a>
 ##### Member functions
 
-`iterator();` the deafult constructor; the underlying octet_iterator is constructed with its default constructor.
+`iterator();` the default constructor; the underlying octet_iterator is constructed with its default constructor.
 
 `explicit iterator (const octet_iterator& octet_it);` a constructor that initializes the underlying octet_iterator with `octet_it`.
 

source/utf8/core.h

@@ -43,9 +43,14 @@ DEALINGS IN THE SOFTWARE.
 #if UTF_CPP_CPLUSPLUS >= 201103L // C++ 11 or later
     #define UTF_CPP_OVERRIDE override
     #define UTF_CPP_NOEXCEPT noexcept
+    #define UTF_CPP_STATIC_ASSERT(condition) static_assert(condition, "UTFCPP static assert");
 #else // C++ 98/03
     #define UTF_CPP_OVERRIDE
     #define UTF_CPP_NOEXCEPT throw()
+    // Simulate static_assert:
+    template <bool Condition> struct StaticAssert {static void assert() {int static_assert_impl[(Condition ? 1 : -1)];} };
+    template <> struct StaticAssert<true> {static void assert() {}};
+    #define UTF_CPP_STATIC_ASSERT(condition) StaticAssert<condition>::assert();
 #endif // C++ 11 or later
 
 
@@ -215,7 +220,7 @@ namespace internal
 
         UTF8_CPP_INCREASE_AND_RETURN_ON_ERROR(it, end)
 
-        code_point += (*it) & 0x3f;
+        code_point = static_cast<utfchar32_t>(code_point + ((*it) & 0x3f));
 
         return UTF8_OK;
     }
@@ -234,11 +239,11 @@ namespace internal
 
         UTF8_CPP_INCREASE_AND_RETURN_ON_ERROR(it, end)
 
-        code_point += (utf8::internal::mask8(*it) << 6) & 0xfff;
+        code_point = static_cast<utfchar32_t>(code_point + ((utf8::internal::mask8(*it) << 6) & 0xfff));
 
         UTF8_CPP_INCREASE_AND_RETURN_ON_ERROR(it, end)
 
-        code_point += (*it) & 0x3f;
+        code_point = static_cast<utfchar32_t>(code_point + ((*it) & 0x3f));
 
         return UTF8_OK;
     }
@@ -308,6 +313,10 @@ namespace internal
     template <typename word_iterator>
     utf_error validate_next16(word_iterator& it, word_iterator end, utfchar32_t& code_point)
     {
+        // Make sure the iterator dereferences a large enough type
+        typedef typename std::iterator_traits<word_iterator>::value_type word_type;
+        UTF_CPP_STATIC_ASSERT(sizeof(word_type) >= sizeof(utfchar16_t));
+        // Check the edge case:
         if (it == end)
             return NOT_ENOUGH_ROOM;
         // Save the original value of it so we can go back in case of failure
@@ -327,7 +336,7 @@ namespace internal
             else if (is_lead_surrogate(first_word)) {
                 const utfchar16_t second_word = *it++;
                 if (is_trail_surrogate(second_word)) {
-                    code_point = (first_word << 10) + second_word + SURROGATE_OFFSET;
+                    code_point = static_cast<utfchar32_t>(first_word << 10) + second_word + SURROGATE_OFFSET;
                     return UTF8_OK;
                 } else 
                     err = INCOMPLETE_SEQUENCE; 
@@ -395,6 +404,7 @@ namespace internal
     // the word_type.
     template <typename word_iterator, typename word_type>
     word_iterator append16(utfchar32_t cp, word_iterator result) {
+        UTF_CPP_STATIC_ASSERT(sizeof(word_type) >= sizeof(utfchar16_t));
         if (is_in_bmp(cp))
             *(result++) = static_cast<word_type>(cp);
         else {

source/utf8/unchecked.h

@@ -115,15 +115,15 @@ namespace utf8
                     ++it; 
                     cp = ((cp << 12) & 0xffff) + ((utf8::internal::mask8(*it) << 6) & 0xfff);
                     ++it;
-                    cp += (*it) & 0x3f;
+                    cp = static_cast<utfchar32_t>(cp + ((*it) & 0x3f));
                     break;
                 case 4:
                     ++it;
                     cp = ((cp << 18) & 0x1fffff) + ((utf8::internal::mask8(*it) << 12) & 0x3ffff);                
                     ++it;
-                    cp += (utf8::internal::mask8(*it) << 6) & 0xfff;
+                    cp = static_cast<utfchar32_t>(cp + ((utf8::internal::mask8(*it) << 6) & 0xfff));
                     ++it;
-                    cp += (*it) & 0x3f; 
+                    cp = static_cast<utfchar32_t>(cp + ((*it) & 0x3f)); 
                     break;
             }
             ++it;

tests/CMakeLists.txt

@@ -6,6 +6,14 @@ add_library(${PROJECT_NAME} INTERFACE)
 
 include_directories("${PROJECT_SOURCE_DIR}/../source")
 
+if (MSVC)
+    # warning level 4
+    add_compile_options(/W4)
+else()
+    # additional warnings
+    add_compile_options(-Wall -Wextra -Wpedantic -Wconversion -Wsign-conversion)
+endif()
+
 add_executable(negative negative.cpp)
 add_executable(cpp11 test_cpp11.cpp)
 add_executable(cpp17 test_cpp17.cpp)
@@ -14,12 +22,6 @@ add_executable(apitests apitests.cpp)
 
 add_executable(noexceptionstests noexceptionstests.cpp)
 
-
-
-target_compile_options(${PROJECT_NAME} INTERFACE
-  $<$<CXX_COMPILER_ID:MSVC>:/W4>
-  $<$<NOT:$<CXX_COMPILER_ID:MSVC>>:-Wall -Wextra -Wpedantic -Wconversion>)
-
 target_compile_options(noexceptionstests PUBLIC -fno-exceptions)
 
 set_target_properties(negative apitests noexceptionstests