Ok, I'm quite baffled about this one. And it's not complex, just very
very very strange. Bear with me.
Let me first give a quick description of the setup: I'm using Hessian
3.0.13 to communicate between an applet and a class running on Tomcat
4.1.31. Both the appplet and Tomcat run on Java 1.6.0_03. The applet
calls a method with a GregorianCalendar as a parameter. The parameter is
then cloned twice and one is moved to the beginning of the week, and the
other to the end of the week. Then some more processing is done, but the
problem is in the "beginning of the week part".
For example, if the passed date is "2006-01-08" then the first day in
the week will be "2006-01-02". This is done by some fairly simple math
using the Calendar's getFirstDayOfWeek() and get(Calendar.DAY_OF_WEEK)
contained in the static gotoFirstDateOfWeek(c) utility method .
The fun part comes now.
If the gotoFirstDateOfWeek() method is called in the applet BEFORE the
Calendar is sent using Hessian, this is the debug output:
/gotoFirstDateOfWeek in=2006-01-08 19:49:38
!!!java.util.GregorianCalendar[time=1136746178453,areFieldsSet=true,areAllFieldsSet=true,lenient=true,zone=sun.util.calendar.ZoneInfo[id="Europe/Berlin",offset=3600000,dstSavings=3600000,useDaylight=true,transitions=143,lastRule=java.util.SimpleTimeZone[id=Europe/Berlin,offset=3600000,dstSavings=3600000,useDaylight=true,startYear=0,startMode=2,startMonth=2,startDay=-1,startDayOfWeek=1,startTime=3600000,startTimeMode=2,endMode=2,endMonth=9,endDay=-1,endDayOfWeek=1,endTime=3600000,endTimeMode=2]],firstDayOfWeek=2,minimalDaysInFirstWeek=4,ERA=1,YEAR=2006,MONTH=0,WEEK_OF_YEAR=1,WEEK_OF_MONTH=1,DAY_OF_MONTH=8,DAY_OF_YEAR=8,DAY_OF_WEEK=1,DAY_OF_WEEK_IN_MONTH=2,AM_PM=1,HOUR=7,HOUR_OF_DAY=19,MINUTE=49,SECOND=38,MILLISECOND=453,ZONE_OFFSET=3600000,DST_OFFSET=0]
!!!1.6.0_03
lFirstDayOfWeek=2
lCurrentDayOfWeek=1
delta=-6
gotoFirstDateOfWeek out=2006-01-02 19:49:38
//!!!java.util.GregorianCalendar[time=1136230848015,areFieldsSet=true,areAllFieldsSet=true,lenient=true,zone=sun.util.calendar.ZoneInfo[id="Europe/Berlin",offset=3600000,dstSavings=3600000,useDaylight=true,transitions=143,lastRule=java.util.SimpleTimeZone[id=Europe/Berlin,offset=3600000,dstSavings=3600000,useDaylight=true,startYear=0,startMode=2,startMonth=2,startDay=-1,startDayOfWeek=1,startTime=3600000,startTimeMode=2,endMode=2,endMonth=9,endDay=-1,endDayOfWeek=1,endTime=3600000,endTimeMode=2]],firstDayOfWeek=2,minimalDaysInFirstWeek=4,ERA=1,YEAR=2006,MONTH=0,WEEK_OF_YEAR=1,WEEK_OF_MONTH=1,DAY_OF_MONTH=2,DAY_OF_YEAR=2,DAY_OF_WEEK=2,DAY_OF_WEEK_IN_MONTH=1,AM_PM=1,HOUR=8,HOUR_OF_DAY=20,MINUTE=40,SECOND=48,MILLISECOND=15,ZONE_OFFSET=3600000,DST_OFFSET=0]
/
Perfect! You can see the incoming date, the fact that 6 days need to be
substracted and the resulting date.
Now the calendar is sent over the line and the same method is called in
the class:
/gotoFirstDateOfWeek in=2006-01-08 19:49:38
!!!java.util.GregorianCalendar[time=1136746178453,areFieldsSet=true,areAllFieldsSet=true,lenient=true,zone=sun.util.calendar.ZoneInfo[id="Europe/Berlin",offset=3600000,dstSavings=3600000,useDaylight=true,transitions=143,lastRule=java.util.SimpleTimeZone[id=Europe/Berlin,offset=3600000,dstSavings=3600000,useDaylight=true,startYear=0,startMode=2,startMonth=2,startDay=-1,startDayOfWeek=1,startTime=3600000,startTimeMode=2,endMode=2,endMonth=9,endDay=-1,endDayOfWeek=1,endTime=3600000,endTimeMode=2]],firstDayOfWeek=2,minimalDaysInFirstWeek=4,ERA=1,YEAR=2006,MONTH=0,WEEK_OF_YEAR=1,WEEK_OF_MONTH=1,DAY_OF_MONTH=8,DAY_OF_YEAR=8,DAY_OF_WEEK=1,DAY_OF_WEEK_IN_MONTH=2,AM_PM=1,HOUR=7,HOUR_OF_DAY=19,MINUTE=49,SECOND=38,MILLISECOND=453,ZONE_OFFSET=3600000,DST_OFFSET=0]
lFirstDayOfWeek=2
lCurrentDayOfWeek=1
//delta=-6/
/gotoFirstDateOfWeek out=2008-01-04 19:49:38
!!!java.util.GregorianCalendar[time=1199475648015,areFieldsSet=true,areAllFieldsSet=true,lenient=true,zone=sun.util.calendar.ZoneInfo[id="Europe/Berlin",offset=3600000,dstSavings=3600000,useDaylight=true,transitions=143,lastRule=java.util.SimpleTimeZone[id=Europe/Berlin,offset=3600000,dstSavings=3600000,useDaylight=true,startYear=0,startMode=2,startMonth=2,startDay=-1,startDayOfWeek=1,startTime=3600000,startTimeMode=2,endMode=2,endMonth=9,endDay=-1,endDayOfWeek=1,endTime=3600000,endTimeMode=2]],firstDayOfWeek=2,minimalDaysInFirstWeek=4,ERA=1,YEAR=2008,MONTH=0,WEEK_OF_YEAR=1,WEEK_OF_MONTH=1,DAY_OF_MONTH=4,DAY_OF_YEAR=4,DAY_OF_WEEK=6,DAY_OF_WEEK_IN_MONTH=1,AM_PM=1,HOUR=8,HOUR_OF_DAY=20,MINUTE=40,SECOND=48,MILLISECOND=15,ZONE_OFFSET=3600000,DST_OFFSET=0]
/A totally different outcome! Look at the year! And it still only
substracted 6 days. To answer the cry for sourcecode:
static public Calendar gotoFirstDateOfWeek(Calendar c)
{
System.out.println( "gotoFirstDateOfWeek in=" +
quickFormatCalendar(c));
System.out.println( "!!!" + c);
System.out.println( "!!!" + System.getProperty("java.version"));
int lFirstDayOfWeek = c.getFirstDayOfWeek();
System.out.println( "lFirstDayOfWeek=" + lFirstDayOfWeek);
int lCurrentDayOfWeek = c.get(Calendar.DAY_OF_WEEK);
System.out.println( "lCurrentDayOfWeek=" + lCurrentDayOfWeek);
int lDelta = 0;
if (lFirstDayOfWeek <= lCurrentDayOfWeek)
{
lDelta = -lCurrentDayOfWeek + lFirstDayOfWeek;
}
else
{
lDelta = -lCurrentDayOfWeek - (7-lFirstDayOfWeek);
}
System.out.println( "delta=" + lDelta);
c.add(Calendar.DATE, lDelta);
System.out.println( "gotoFirstDateOfWeek out=" +
quickFormatCalendar(c));
System.out.println( "!!!" + c);
return c;
}
Now, I can actually work around this, by inserting the following code at
the beginning of the method:
static public Calendar gotoFirstDateOfWeek(Calendar c)
{
Calendar o = c;
c = new GregorianCalendar();
c.setFirstDayOfWeek( o.getFirstDayOfWeek() );
c.setMinimalDaysInFirstWeek( o.getMinimalDaysInFirstWeek() );
c.setTimeInMillis( o.getTimeInMillis() );
System.out.println( "gotoFirstDateOfWeek in=" +
quickFormatCalendar(c));
...
}
This basically creates a new Calendar and sets all relevant properties
so the behavior is correct again. Cloning does not help, so something
must get screwed when sending a Calendar via Hessian.
I tried the 3.1.3 release, but I get a bunch of HTTP 500 errors, so no
luck there so far.
Any suggestions?
Tom
_______________________________________________
hessian-interest mailing list
[email protected]
http://maillist.caucho.com/mailman/listinfo/hessian-interest
