UnicodeClientCommunication < Main

Tags: view all tags
---+ Unicode Client Communications

_Client/server data flow has be reworked to support unicode data._

%TOC%
---++ Overview

Initially, Only the Java Client will communicate using unicode data. The Win32 and Character clients will not support unicode data at this time.

   * Client and Server exchange all alpha data as UTF-8
   * Client transcodes received data from UTF-8 to UTF-32
   * Client transcodes sent data from UTF-32 to UTF-8
   * Server transcodes received data from UTF-8 to the single-byte encoding (8859-15) for RAW fields or UTF-32 as appropriate
   * Server transcodes sent data from the internal single-byte encoding (8859-15) for RAW fields or UTF-32 to UTF-8.
---++ Test Plan:
---++ Bugs:

   1 <span style="color: red;" class="WYSIWYG_COLOR">Some characters do not display properly. In the range U+0400 to U+0402, the top part of the character is missing, which makes the first two characters look the same (Mac client) or the U+0400 character does not display at all (Windows client). The Windows probelm could be because the font DialogInput does not have anythhing to display for U+0400.</span>
   1 <span style="color: red;" class="WYSIWYG_COLOR">A field that contains '\n', chr(10) or '\r' and is set as a Label widget does not display correctly. Those characters should create a new line, but instead display as a space, or treat them as literals (as the non unicode client does).</span>

-- Main.PeteBrower - 2011-03-04
Topic revision: r5 - 2011-11-04 - JeanNeron