[commit-cp] [task #4211] More effective reading of the native wide character strings.

[Audrius Meškauskas]

URL:
  <http://savannah.gnu.org/task/?func=detailitem&item_id=4211>

                 Summary: More effective reading of the native wide character
strings.
                 Project: classpath
            Submitted by: audriusa
            Submitted on: Wed 06/01/2005 at 17:11
                Category: CORBA
         Should Start On: Wed 06/01/2005 at 00:00
   Should be Finished on: Thu 06/01/2006 at 00:00
                Priority: 1 - Later
                  Status: None
                 Privacy: Public
        Percent Complete: 0%
             Assigned to: None
             Open/Closed: Open
                  Effort: 0.00

    _______________________________________________________

Details:

The method read_wstring_UTF_16()in gnu.CORBA.cdrInput is responsible for
reading wide character strings in java native UTF-16 encoding. The input can
be in either Big or Little endian, depending from the optional endian
marker.

This method is tested and works, but it seems not really optimized. It may be
faster to use classes from java.nio that convert between byte buffer and char
buffer and provide support for the variable byte order. The wide string is a
frequent element in a CORBA messages, performance may be important here. The
method is indirectly covered by Mauve tests, so the possibility to introduce
a bug is comparatively small. 






    _______________________________________________________

Reply to this item at:

  <http://savannah.gnu.org/task/?func=detailitem&item_id=4211>

_______________________________________________
  Message sent via/by Savannah
  http://savannah.gnu.org/