Moving ICU dependent function ReadUTFChar into the icu.cc file....

src/url_canon_internal.h

 // Copyright 2007, Google Inc.
 // All rights reserved.
 //
 // Redistribution and use in source and binary forms, with or without
 // modification, are permitted provided that the following conditions are
 // met:
 //
 //     * Redistributions of source code must retain the above copyright
 // notice, this list of conditions and the following disclaimer.
 //     * Redistributions in binary form must reproduce the above
@@ skipping 145 matching lines @@
 // UTF-8 functions ------------------------------------------------------------
 
 // Reads one character in UTF-8 starting at |*begin| in |str| and places
 // the decoded value into |*code_point|. If the character is valid, we will
 // return true. If invalid, we'll return false and put the
 // kUnicodeReplacementCharacter into |*code_point|.
 //
 // |*begin| will be updated to point to the last character consumed so it
 // can be incremented in a loop and will be ready for the next character.
 // (for a single-byte ASCII character, it will not be changed).
-inline bool ReadUTFChar(const char* str, int* begin, int length,
-                        unsigned* code_point_out) {
-  int code_point;  // Avoids warning when U8_NEXT writes -1 to it.
-  U8_NEXT(str, *begin, length, code_point);
-  *code_point_out = static_cast<unsigned>(code_point);
-
-  // The ICU macro above moves to the next char, we want to point to the last
-  // char consumed.
-  (*begin)--;
-
-  // Validate the decoded value.
-  if (U_IS_UNICODE_CHAR(code_point))
-    return true;
-  *code_point_out = kUnicodeReplacementCharacter;
-  return false;
-}
+//
+// Implementation is in url_canon_icu.cc.
+bool ReadUTFChar(const char* str, int* begin, int length,
+                 unsigned* code_point_out);
 
 // Generic To-UTF-8 converter. This will call the given append method for each
 // character that should be appended, with the given output method. Wrappers
 // are provided below for escaped and non-escaped versions of this.
 template<class Output, void Appender(unsigned char, Output*)>
 inline void DoAppendUTF8(unsigned char_value, Output* output) {
   if (char_value <= 0x7f) {
     Appender(static_cast<unsigned char>(char_value), output);
   } else if (char_value <= 0x7ff) {
     // 110xxxxx 10xxxxxx
@@ skipping 61 matching lines @@
 // UTF-16 functions -----------------------------------------------------------
 
 // Reads one character in UTF-16 starting at |*begin| in |str| and places
 // the decoded value into |*code_point|. If the character is valid, we will
 // return true. If invalid, we'll return false and put the
 // kUnicodeReplacementCharacter into |*code_point|.
 //
 // |*begin| will be updated to point to the last character consumed so it
 // can be incremented in a loop and will be ready for the next character.
 // (for a single-16-bit-word character, it will not be changed).
-inline bool ReadUTFChar(const char16* str, int* begin, int length,
-                        unsigned* code_point) {
-  if (U16_IS_SURROGATE(str[*begin])) {
-    if (!U16_IS_SURROGATE_LEAD(str[*begin]) || *begin + 1 >= length ||
-        !U16_IS_TRAIL(str[*begin + 1])) {
-      // Invalid surrogate pair.
-      *code_point = kUnicodeReplacementCharacter;
-      return false;
-    } else {
-      // Valid surrogate pair.
-      *code_point = U16_GET_SUPPLEMENTARY(str[*begin], str[*begin + 1]);
-      (*begin)++;
-    }
-  } else {
-    // Not a surrogate, just one 16-bit word.
-    *code_point = str[*begin];
-  }
-
-  if (U_IS_UNICODE_CHAR(*code_point))
-    return true;
-
-  // Invalid code point.
-  *code_point = kUnicodeReplacementCharacter;
-  return false;
-}
+//
+// Implementation is in url_canon_icu.cc.
+bool ReadUTFChar(const char16* str, int* begin, int length,
+                 unsigned* code_point);
 
 // Equivalent to U16_APPEND_UNSAFE in ICU but uses our output method.
 inline void AppendUTF16Value(unsigned code_point,
                              CanonOutputT<char16>* output) {
   if (code_point > 0xffff) {
     output->push_back(static_cast<char16>((code_point >> 10) + 0xd7c0));
     output->push_back(static_cast<char16>((code_point & 0x3ff) | 0xdc00));
   } else {
     output->push_back(static_cast<char16>(code_point));
   }
@@ skipping 184 matching lines @@
 inline unsigned long long _strtoui64(const char* nptr,
                                      char** endptr, int base) {
   return strtoull(nptr, endptr, base);
 }
 
 #endif  // WIN32
 
 }  // namespace url_canon
 
 #endif  // GOOGLEURL_SRC_URL_CANON_INTERNAL_H__