Fix for issue #92: Clang 10 Compiler warnings

Fix warnings when -Wconversion or -Wsign-conversion are enabled.
nemtrif · Dec 26, 2022 · e3e57c8 · e3e57c8
1 parent d736c29
commit e3e57c8
Show file tree

Hide file tree

Showing 5 changed files with 59 additions and 37 deletions.
diff --git a/source/utf8/checked.h b/source/utf8/checked.h
@@ -50,6 +50,7 @@ namespace utf8
         uint8_t u8;
     public:
         invalid_utf8 (uint8_t u) : u8(u) {}
+        invalid_utf8 (char c) : u8(static_cast<uint8_t>(c)) {}
         virtual const char* what() const UTF_CPP_NOEXCEPT UTF_CPP_OVERRIDE { return "Invalid UTF-8"; }
         uint8_t utf8_octet() const {return u8;}
     };
@@ -75,24 +76,7 @@ namespace utf8
         if (!utf8::internal::is_code_point_valid(cp))
             throw invalid_code_point(cp);
 
-        if (cp < 0x80)                        // one octet
-            *(result++) = static_cast<uint8_t>(cp);
-        else if (cp < 0x800) {                // two octets
-            *(result++) = static_cast<uint8_t>((cp >> 6)            | 0xc0);
-            *(result++) = static_cast<uint8_t>((cp & 0x3f)          | 0x80);
-        }
-        else if (cp < 0x10000) {              // three octets
-            *(result++) = static_cast<uint8_t>((cp >> 12)           | 0xe0);
-            *(result++) = static_cast<uint8_t>(((cp >> 6) & 0x3f)   | 0x80);
-            *(result++) = static_cast<uint8_t>((cp & 0x3f)          | 0x80);
-        }
-        else {                                // four octets
-            *(result++) = static_cast<uint8_t>((cp >> 18)           | 0xf0);
-            *(result++) = static_cast<uint8_t>(((cp >> 12) & 0x3f)  | 0x80);
-            *(result++) = static_cast<uint8_t>(((cp >> 6) & 0x3f)   | 0x80);
-            *(result++) = static_cast<uint8_t>((cp & 0x3f)          | 0x80);
-        }
-        return result;
+        return internal::append(cp, result);
     }
 
     template <typename octet_iterator, typename output_iterator>

diff --git a/source/utf8/core.h b/source/utf8/core.h
@@ -297,6 +297,55 @@ namespace internal
         return utf8::internal::validate_next(it, end, ignored);
     }
 
+    // Internal implementation of both checked and unchecked append() function
+    // This function will be invoked by the overloads below, as they will know
+    // the octet_type.
+    template <typename octet_iterator, typename octet_type>
+    octet_iterator append(uint32_t cp, octet_iterator result) {
+        if (cp < 0x80)                        // one octet
+            *(result++) = static_cast<octet_type>(cp);
+        else if (cp < 0x800) {                // two octets
+            *(result++) = static_cast<octet_type>((cp >> 6)          | 0xc0);
+            *(result++) = static_cast<octet_type>((cp & 0x3f)        | 0x80);
+        }
+        else if (cp < 0x10000) {              // three octets
+            *(result++) = static_cast<octet_type>((cp >> 12)         | 0xe0);
+            *(result++) = static_cast<octet_type>(((cp >> 6) & 0x3f) | 0x80);
+            *(result++) = static_cast<octet_type>((cp & 0x3f)        | 0x80);
+        }
+        else {                                // four octets
+            *(result++) = static_cast<octet_type>((cp >> 18)         | 0xf0);
+            *(result++) = static_cast<octet_type>(((cp >> 12) & 0x3f)| 0x80);
+            *(result++) = static_cast<octet_type>(((cp >> 6) & 0x3f) | 0x80);
+            *(result++) = static_cast<octet_type>((cp & 0x3f)        | 0x80);
+        }
+        return result;
+    }
+
+    // One of the following overloads will be invoked from the API calls
+
+    // A simple (but dangerous) case: the caller appends byte(s) to a char array
+    inline char* append(uint32_t cp, char* result) {
+        return append<char*, char>(cp, result);
+    }
+
+    // Hopefully, most common case: the caller uses back_inserter
+    // i.e. append(cp, std::back_inserter(str));
+    template<typename container_type>
+    std::back_insert_iterator<container_type> append
+            (uint32_t cp, std::back_insert_iterator<container_type> result) {
+        return append<std::back_insert_iterator<container_type>,
+            typename container_type::value_type>(cp, result);
+    }
+
+    // The caller uses some other kind of output operator - not covered above
+    // Note that in this case we are not able to determine octet_type
+    // so we assume it's uint_8; that can cause a conversion warning if we are wrong.
+    template <typename octet_iterator>
+    octet_iterator append(uint32_t cp, octet_iterator result) {
+        return append<octet_iterator, uint8_t>(cp, result);
+    }
+
 } // namespace internal
 
     /// The library API - functions intended to be called by the users

diff --git a/source/utf8/unchecked.h b/source/utf8/unchecked.h
@@ -37,24 +37,7 @@ namespace utf8
         template <typename octet_iterator>
         octet_iterator append(uint32_t cp, octet_iterator result)
         {
-            if (cp < 0x80)                        // one octet
-                *(result++) = static_cast<uint8_t>(cp);
-            else if (cp < 0x800) {                // two octets
-                *(result++) = static_cast<uint8_t>((cp >> 6)          | 0xc0);
-                *(result++) = static_cast<uint8_t>((cp & 0x3f)        | 0x80);
-            }
-            else if (cp < 0x10000) {              // three octets
-                *(result++) = static_cast<uint8_t>((cp >> 12)         | 0xe0);
-                *(result++) = static_cast<uint8_t>(((cp >> 6) & 0x3f) | 0x80);
-                *(result++) = static_cast<uint8_t>((cp & 0x3f)        | 0x80);
-            }
-            else {                                // four octets
-                *(result++) = static_cast<uint8_t>((cp >> 18)         | 0xf0);
-                *(result++) = static_cast<uint8_t>(((cp >> 12) & 0x3f)| 0x80);
-                *(result++) = static_cast<uint8_t>(((cp >> 6) & 0x3f) | 0x80);
-                *(result++) = static_cast<uint8_t>((cp & 0x3f)        | 0x80);
-            }
-            return result;
+            return internal::append(cp, result);
         }
 
         template <typename octet_iterator, typename output_iterator>

diff --git a/tests/CMakeLists.txt b/tests/CMakeLists.txt
@@ -21,7 +21,7 @@ target_link_libraries(noexceptionstests PRIVATE utf8::cpp)
 
 target_compile_options(${PROJECT_NAME} INTERFACE
   $<$<CXX_COMPILER_ID:MSVC>:/W4>
-  $<$<NOT:$<CXX_COMPILER_ID:MSVC>>:-Wall -Wextra -Wpedantic>)
+  $<$<NOT:$<CXX_COMPILER_ID:MSVC>>:-Wall -Wextra -Wpedantic -Wconversion -Wsign-conversion>)
 
 target_compile_options(noexceptionstests PUBLIC -fno-exceptions)
 

diff --git a/tests/test_checked_api.cpp b/tests/test_checked_api.cpp
@@ -38,6 +38,12 @@ TEST(CheckedAPITests, test_append)
     EXPECT_EQ (u[2], 0x8d);
     EXPECT_EQ (u[3], 0x86);
     EXPECT_EQ (u[4], 0);
+
+    // Ensure no warnings with plain char
+    char c[2] = {0,0};
+    append('a', c);
+    EXPECT_EQ (u[0], 'a');
+    EXPECT_EQ (u[1], 0);
 }
 
 TEST(CheckedAPITests, test_next)