Expose latin1 methods in WebString

Source/core/platform/chromium/support/WebString.cpp

 /*
  * Copyright (C) 2009 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 78 matching lines @@
 WebString WebString::fromUTF8(const char* data, size_t length)
 {
     return WTF::String::fromUTF8(data, length);
 }
 
 WebString WebString::fromUTF8(const char* data)
 {
     return WTF::String::fromUTF8(data);
 }
 
+std::string WebString::latin1() const
+{
+    WebCString latin1 = WTF::String(m_private.get()).latin1();
+    return std::string(latin1.data(), latin1.length());

[abarth-chromium, 2013/08/31 18:04:18]

WTF::  <--- This prefix isn't needed.

Do you actually need to be able to convert a WebString to Latin-1?  If so, we
can make a more efficient implementation.  If not, we shouldn't add this
function.

We shouldn't need to make a CString copy.  We should be able to copy directly
from the String to the std::string.  See
<https://docs.google.com/document/d/1kOCUlJdh2WJMJGDf-WoEQhmnjKLaOYRbiHz5TiGJl...>
for more information about how String represents its data.

[davidben, 2013/09/03 15:04:56]

Removed. I was just going by the UTF-8 ones which have the prefix. Should I
remove those while I'm here?
For getting XMLHttpRequest#setRequestHeader right I need to be able to go the
other direction too. That one we're not the only browser to encode as UTF-8
right now, unlike going the other direction (see pile of test results in first
comment), but a future Firefox will do Latin-1 and Safari already does. Also it
seems really weird to interpret response headers as Latin-1 but encode request
headers as UTF-8.
Made a new version. It still does a CString copy if we have a UChar instead of
LChar string, but otherwise it doesn't. StringUTF8Adaptor is similar. (Or would
you prefer reimplementing String::latin1 altogether? Hopefully that's by far the
less common case since characters outside of latin1 can't go in the result
anyway.)

+}
+
+WebString WebString::fromLatin1(const char* data, size_t length)
+{
+    return WTF::String(data, length);

[abarth-chromium, 2013/08/31 18:04:18]

WTF::  <---- No need for this prefix.

[davidben, 2013/09/03 15:04:56]

Done.

+}
+
+WebString WebString::fromLatin1(const char* data)
+{
+    return WTF::String(data);
+}

[abarth-chromium, 2013/08/31 18:04:18]

Do you actually need this function?  If not, we shouldn't added it.

[davidben, 2013/09/03 15:04:56]

Removed.

+
 bool WebString::equals(const WebString& s) const
 {
     return equal(m_private.get(), s.m_private.get());
 }
 
 WebString::WebString(const WTF::String& s)
 {
     m_private = s.impl();
 }
 
@@ skipping 23 matching lines @@
 {
     return WTF::AtomicString(m_private.get());
 }
 
 void WebString::assign(WTF::StringImpl* p)
 {
     m_private = p;
 }
 
 } // namespace WebKit