Add toStdString methods to CString and WTF::String

third_party/WebKit/Source/wtf/text/WTFString.h

 /*
  * (C) 1999 Lars Knoll (knoll@kde.org)
  * Copyright (C) 2004, 2005, 2006, 2007, 2008, 2009, 2010, 2012, 2013 Apple Inc. All rights reserved.
  *
  * This library is free software; you can redistribute it and/or
  * modify it under the terms of the GNU Library General Public
  * License as published by the Free Software Foundation; either
  * version 2 of the License, or (at your option) any later version.
  *
  * This library is distributed in the hope that it will be useful,
@@ skipping 161 matching lines @@
     {
         if (!m_impl)
             return 0;
         return m_impl->length() * (is8Bit() ? sizeof(LChar) : sizeof(UChar));
     }
 
     CString ascii() const;
     CString latin1() const;
     CString utf8(UTF8ConversionMode = LenientUTF8Conversion) const;
 
+    std::string toUTF8() const;

[Mikhail, 2015/09/30 14:02:01]

why not toStdString() ?

[Primiano Tucci (use gerrit), 2015/09/30 15:08:28]

AFAIK technically speaking std::string does not mandate a specific encoding [1],
so toStdString wouldn't explain what encoding the caller should expect.
More specifically, if you construct a WTF::String out of a UTF-16 string
[0x20AC, 0xD801], I could expect a toStdString() to just return a string
containing 4 bytes (0x20, 0xAC, 0xD8, 0x01), which is NOT the case here. What
you get is the equivalent UTF8 encoding of the same string (See the unittest)

[1] cplusplus.com says "If used to handle sequences of multi-byte or
variable-length characters", http://www.cplusplus.com/reference/string/string/

[Mikhail, 2015/09/30 15:58:39]

Agree, however having both 'CString utf8()' and 'std::string toUTF8()' at the
same time looks a bit awkward. Also 'to' prefix usually implies casting to a
type and here we just create a new 'std::string' instance. Maybe
'utf8StdString()' or smth. like that?

[Primiano Tucci (use gerrit), 2015/09/30 16:03:37]

Hmm ok for the StdString part, but I'd keep the to.
In my mind "to" implies a transformation, without to I expect it to be a trivial
operation / casting.
How about toUTF8StdString() ?

+
     UChar operator[](unsigned index) const
     {
         if (!m_impl || index >= m_impl->length())
             return 0;
         return (*m_impl)[index];
     }
 
     static String number(int);
     static String number(unsigned);
     static String number(long);
@@ skipping 506 matching lines @@
 using WTF::charactersToFloat;
 using WTF::equal;
 using WTF::equalIgnoringCase;
 using WTF::find;
 using WTF::isAllSpecialCharacters;
 using WTF::isSpaceOrNewline;
 using WTF::reverseFind;
 
 #include "wtf/text/AtomicString.h"
 #endif // WTFString_h