Hello everyone,
The fact that we are seeing differences in Linux versus Windows is an important
clue.
This *may* have something to do with the default locale of the JVM on the
Windows machine.
We occasionally see these sorts of problems especially for dates on
international machines.
I have attached a very simple Java program that may help us solve this.
Please run this command:
C:\Program Files\IDV_5.1u3\jre\bin\java LocaleTest
What is the result?
You can change the default locale on the Windows machine by editing the
C:\Program Files\IDV_5.1u3\runIDV.bat script.
At the bottom of the file, there should be something like this:
jre\bin\java -Xmx%idv_memory%m -XX:PermSize=%idv_maxpermgensize%m
-XX:MaxPermSize=%idv_maxpermgensize%m -XX:+DisableExplicitGC
-Didv.enableStereo=false -jar idv.jar %*
After is says "jre\bin\java", please add
-Duser.country=US -Duser.language=en
What happens?
Best,
Unidata IDV Support
> HI
>
> also, it show error like
> couldn't parse at 5
> couldn't parse at 2
> couldn't parse at 0...
>
> Seems when using GradsTimeDimension.class
>
> Line 172
> -------------------------------------------------------------------------------------------
> SimpleDateFormat sdf = new SimpleDateFormat(pattern);
>
> sdf.setTimeZone(TimeZone.getTimeZone("GMT"));
> ParsePosition p = new ParsePosition(0);
> Date d = sdf.parse(tstart, p);
> if (d == null)
> {
> System.out.println("couldn't parse at " + p.getErrorIndex());
> d = new Date(0L);
> }
> sdf.applyPattern("yyyy-MM-dd HH:mm:ss Z");
> setUnit("hours since " + sdf.format(d, new StringBuffer(), new
> FieldPosition(0)));
> -------------------------------------------------------------------------------------------
>
> it cannot get tstart, so it pass 0L to d which lead to 1970-01-01-00
>
> I am not familiar with JAVA.
> Hope this could help you guys to get a start.
>
> I also thought it could be a unicode difference between windows and linux.
> but after I use tool to swich unicode of the ctl file from unix to windows,
> nothing happenes.
>
> Best
> Jian
>
>
> > -----Original Messages-----
> > From: "Unidata IDV Support" <address@hidden>
> > Sent Time: 2015-05-22 01:09:07 (Friday)
> > To: address@hidden
> > Cc: address@hidden, address@hidden, address@hidden
> > Subject: [IDV #RMI-290185]: Problems of grads ctl file time?
> >
> > > Full Name: Dr. Jian Tang
> > > Email Address: address@hidden
> > > Organization: China Meteorological Administration
> > > Package Version: 5.0u2,5.1u2
> > > Operating System: win7 x64
> > > Hardware:
> > > Description of problem: hi there,
> > >
> > > I don't know how to comment on [IDV #RMI-290185]: Problems of grads ctl
> > > file time?
> > > So I just make a new post... Sorry for the trouble. I sent a email to idv
> > > user mail list, after second thought, probably i should not have done
> > > that...
> > >
> > > we are using IDV 5.0u2 and 5.1u2 try to read GrADs binary grid files
> > > through control file (CTL File).
> > > The Ctl looks like this and it works fine using Grads to open it.
> > >
> > > both 5.0 and 5.1 can open ctl file fine and plot fine.
> > > only issue remain is the time dimension
> > > no matter how I change the tdef line, "12z20MAY2015"
> > > it always show starting from 1970-01-01-08 CST in IDV
> > > in this case:
> > > 1970-01-01-08 CST
> > > 1970-01-01-09 CST
> > > 1970-01-01-10 CST...
> > >
> > > If I change "60mn" to "1dy", it will appear like
> > > 1970-01-01-08 CST
> > > 1970-01-02-08 CST
> > > 1970-01-03-08 CST...
> > >
> > > If I change "49", IDV will report error.
> > >
> > > Seems to me that the IDV does not get any infomation from "12z20MAY2015"
> > > in the line TDEF in CTL file.
> > >
> > > If anyone has any experience please do not hestitate to share with me.
> > > Many thanks.
> >
> >
> > Again, there is nothing wrong with the syntax of time definition. I am not
> > GrADS expert, you need to find someone to help you figure out the other
> > part of this ctl file. I used the attached file from
> > DaiKan, and change the time definition to "12z20MAY2015", the result in the
> > IDV is exactly what I expected.
> >
> > DaiKan, the attached file works correctly in the IDV 5.1u2.
> >
> > Yuan
> > >
> > > -------------------------------------------
> > > dset ^postvar2015052012
> > > options sequential big_endian
> > > title post output from grapes
> > > undef 9.999E+20
> > > xdef 831 linear 102.0000 0.0300
> > > ydef 661 linear 17.0000 0.0300
> > > zdef 26 levels
> > > 1000.000000
> > > 975.0000000
> > > ...... to save space here
> > > 20.00000000
> > > 10.00000000
> > > tdef 49 linear 12z20MAY2015 60mn
> > > vars 30
> > > u 26 0 u_wind
> > > v 26 0 v_wind
> > > ...... to save space here
> > > tslb 4 0 tslb
> > > mslb 4 0 mslb
> > > endvars
> > > ------------------------------------------
> > > please see to the attachments for more details.
> > >
> > > https://www.unidata.ucar.edu/mailing_lists/archives/idvusers/2011/msg00042.html
> > >
> > >
> >
> > Ticket Details
> > ===================
> > Ticket ID: RMI-290185
> > Department: Support IDV
> > Priority: Normal
> > Status: Closed
> >
>
>
> --
> Dr. Jian TANG
> Central Meteorological Office
> National Meteorological Center of CMA
>
> No. 46, Zhongguancun South Street, Haidian District, Beijing
> P. R. China, 100081
>
>
>
>
>
>
Ticket Details
===================
Ticket ID: RMI-290185
Department: Support IDV
Priority: Normal
Status: OpenAttachment:
LocaleTest.class
Description: Binary data