Break dependency on google::protobuf::UTF8FirstLetterNumBytes() (#814)

Remove dependency on the `google::protobuf::UTF8FirstLetterNumBytes()` to prep for the 22.x upgrade. The UTF8 utilities were removed from protobuf in 22.x. Jump to 22.x protobuf is needed for the C++20 transition. Since the needed code was very simple I have copied the functions from the protobuf to the validate.h file. Signed-off-by: Yan Avlasov <[email protected]> Co-authored-by: Elliot Jackson <[email protected]>
bufbuild · Mar 20, 2023 · bec8c04 · bec8c04
1 parent d153612
commit bec8c04
Showing 1 changed file with 16 additions and 3 deletions.
diff --git a/validate/validate.h b/validate/validate.h
@@ -28,7 +28,6 @@
 #endif
 
 #include "google/protobuf/message.h"
-#include "google/protobuf/stubs/strutil.h" // for UTF8Len
 
 namespace pgv {
 using std::string;
@@ -151,20 +150,34 @@ static inline bool IsHostname(const string& to_validate) {
   return true;
 }
 
-static inline size_t Utf8Len(const string& narrow_string) {
+namespace {
+
+inline int OneCharLen(const char* src) {
+  return "\1\1\1\1\1\1\1\1\1\1\1\1\2\2\3\4"[(*src & 0xFF) >> 4];
+}
+
+inline int UTF8FirstLetterNumBytes(const char *utf8_str, int str_len) {
+  if (str_len == 0)
+    return 0;
+  return OneCharLen(utf8_str);
+}
+
+inline size_t Utf8Len(const string& narrow_string) {
   const char* str_char = narrow_string.c_str();
   ptrdiff_t byte_len = narrow_string.length();
   size_t unicode_len = 0;
   int char_len = 1;
   while (byte_len > 0 && char_len > 0) {
-    char_len = google::protobuf::UTF8FirstLetterNumBytes(str_char, byte_len);
+    char_len = UTF8FirstLetterNumBytes(str_char, byte_len);
     str_char += char_len;
     byte_len -= char_len;
     ++unicode_len;
   }
   return unicode_len;
 }
 
+} // namespace
+
 } // namespace pgv
 
 #endif // _VALIDATE_H