<html>
  <head>
    <meta content="text/html; charset=UTF-8" http-equiv="Content-Type">
  </head>
  <body bgcolor="#FFFFFF" text="#000000">
    <div class="moz-cite-prefix">Libedit internally uses wchar_t to
      handle wide characters.  For the types of things libedit does, I
      think a wchar_t is better suited than an array of utf8 coded
      bytes.  The translations in question are for getting data in and
      out of libedit.  <br>
      <br>
      This means that support for extended characters in the command
      line history will be dependent on having <codecvt> support
      and a libedit built with wchar-t support, which, AFAIK, is only
      OSX.<br>
      <br>
      Currently, I am reworking the patch, so it works with either
      libedit's char or wchar_t functions.  This is a compile time
      decision.<br>
      <br>
      If we want wide character support on other platforms down the
      road, it would make sense to bring the libedit functions into
      lldb.  We can add a custom wchar to utf8 translations if  gcc
      still does not support <codecvt>. <br>
      <br>
      On 10/29/2014 11:38 AM, Zachary Turner wrote:<br>
    </div>
    <blockquote
cite="mid:CAAErz9i2rWAfoHny3gJSqqdppkYAvne3b3mTbF6jWjLwSBr_oQ@mail.gmail.com"
      type="cite">If you're storing UTF8 anyway, why not just use
      regular character strings?  Doesn't it defeat the purpose of using
      UTF8 if you're combining it with a character type that isn't 1
      byte?<br>
      <br>
      <div class="gmail_quote">On Wed Oct 29 2014 at 11:27:29 AM Kate
        Stone <<a moz-do-not-send="true"
          href="mailto:katherine_stone@apple.com">katherine_stone@apple.com</a>>
        wrote:<br>
        <blockquote class="gmail_quote" style="margin:0 0 0
          .8ex;border-left:1px #ccc solid;padding-left:1ex">
          <div dir="auto">
            <div>On Oct 28, 2014, at 1:55 PM, Zachary Turner <<a
                moz-do-not-send="true" href="mailto:zturner@google.com"
                target="_blank">zturner@google.com</a>> wrote:<br>
              <br>
            </div>
            <blockquote type="cite">
              <div>
                <div class="gmail_quote">On Tue Oct 28 2014 at 1:46:26
                  PM Vince Harron <<a moz-do-not-send="true"
                    href="mailto:vharron@google.com" target="_blank">vharron@google.com</a>>
                  wrote:<br>
                  <blockquote class="gmail_quote" style="margin:0 0 0
                    .8ex;border-left:1px #ccc solid;padding-left:1ex">
                    <div dir="ltr"><br>
                    </div>
                    <div dir="ltr">
                      <div><font face="arial, sans-serif">> </font><span
style="font-family:arial,sans-serif;font-size:13px">- rework the
                          Editline rewrite, so it either uses standard 8
                          bit chars, or wchar_t/utf8 depending on the
                          platform.  This would be conditionally built
                          depending on the platform.</span></div>
                      <div><span
                          style="font-family:arial,sans-serif;font-size:13px"><br>
                        </span></div>
                    </div>
                    <div dir="ltr">
                      <div><span
                          style="font-family:arial,sans-serif;font-size:13px">This
                          would be my favorite option if possible.
                           wchar_t never really took roots in Linux
                          AFAIK.</span></div>
                    </div>
                  </blockquote>
                  <div><br>
                  </div>
                  <div>Also probably the best option for Windows,
                    although it's worth pointing out that at least for
                    now, most other stuff in LLDB doesn't really use
                    wide character strings either, so char would be the
                    path of least resistance for Windows right now.</div>
                </div>
              </div>
            </blockquote>
            <div><br>
            </div>
          </div>
          <div dir="auto">
            <div>With the Editline rewrite I made the explicit decision
              to insulate the rest of LLDB from wide characters and
              strings by encoding everything as UTF8.  I agree that
              reverting to char-only input is a perfectly reasonable
              solution for platforms that don't yet include wchar-aware
              libedit implementations.</div>
            <div><br>
            </div>
            <div style="word-wrap:break-word"><span
                style="background-color:rgba(255,255,255,0)">Kate Stone <a
                  moz-do-not-send="true" href="mailto:k8stone@apple.com"
                  target="_blank">k8stone@apple.com</a></span></div>
            <div style="word-wrap:break-word"><span
                style="background-color:rgba(255,255,255,0)"> Xcode Runtime
                Analysis Tools</span></div>
          </div>
        </blockquote>
      </div>
    </blockquote>
    <br>
  </body>
</html>