Port back CLD to Linux and Mac and fix Trad Chinese detection

third_party/cld/base/string_util.h

Index: third_party/cld/base/string_util.h
===================================================================
--- third_party/cld/base/string_util.h	(revision 36372)
+++ third_party/cld/base/string_util.h	(working copy)
@@ -7,17 +7,47 @@
 #ifndef BASE_STRING_UTIL_H_
 #define BASE_STRING_UTIL_H_
 
+#include <string>
 #include <string.h>
 
 namespace base {
 
+#ifdef WIN32
 // Compare the two strings s1 and s2 without regard to case using
 // the current locale; returns 0 if they are equal, 1 if s1 > s2, and -1 if
 // s2 > s1 according to a lexicographic comparison.
 inline int strcasecmp(const char* s1, const char* s2) {
   return _stricmp(s1, s2);
 }
+#else
+inline int strcasecmp(const char* s1, const char* s2) {
+  return strcasecmp(s1, s2);
+}
+#endif
 
+// This is mpcomplete's pattern for saving a string copy when dealing with
+// a function that writes results into a wchar_t[] and wanting the result to
+// end up in a std::wstring.  It ensures that the std::wstring's internal
+// buffer has enough room to store the characters to be written into it, and
+// sets its .length() attribute to the right value.
+//
+// The reserve() call allocates the memory required to hold the string
+// plus a terminating null.  This is done because resize() isn't
+// guaranteed to reserve space for the null.  The resize() call is
+// simply the only way to change the string's 'length' member.
+//
+// XXX-performance: the call to wide.resize() takes linear time, since it fills
+// the string's buffer with nulls.  I call it to change the length of the
+// string (needed because writing directly to the buffer doesn't do this).
+// Perhaps there's a constant-time way to change the string's length.
+template <class string_type>
+inline typename string_type::value_type* WriteInto(string_type* str,
+                                                   size_t length_with_null) {
+  str->reserve(length_with_null);
+  str->resize(length_with_null - 1);
+  return &((*str)[0]);
 }
 
+}
+
 #endif  // BASE_STRING_UTIL_H_