jepp-users — Send your questions here

[Jepp-users] Create personalized objects for Java

[Edoardo T. <edo...@gm...>]

Hi all.

Invoking some Python modules from Java i saw that lists and dict  returned
from Python are mapped in Java as Strings.
And that is for me a "problem" because i have to use the data inside lists
and dicts; so one possible, but bad solution is to parse the string and
recreate a Java object.

Mike told me that if i have a list i can map it in Java as an array, using
jarray, and that's ok. I have only to rewrite some python code and that is
not a big problem.
but if i have a Python module that returns something like the dict below or
some other complex data structures:
{
 key = [ val1, val2, val3],
 key2 = [val21, val22, val23],
 key3 {
          key31 = val31,
          key32 = val32,
          key3n = val3n,
          }
}

How can i map it?
Do you think is a possible solution if i write some C modules that maps my
personal objects?

thank you all


-- 
Edoardo Tosca
Mail: edoardo.tosca@{autistici.org, gmail.com}
Skype: edoardotosca

Re: [Jepp-users] Call a function of a Python class in a module

[Edoardo T. <edo...@au...>]

Ok thank you,
I have tried the first solution and it works very well.
I have some more questions, but i will create new threads.


On Thu, Jul 3, 2008 at 5:30 AM, Mike Johnson <mrj...@mi...> wrote:

> On second thought, I didn't fully answer the question. :-)
>
> The first invoke works because Python classes are implemented as
> Callable objects. So the constructor is just a function that returns an
> object that has functions of it's own, more or less.
>
> The second fails because there's no MyUsefulFunction in the scope, it's
> part of MyClass.
>
> If you did something like:
> jep.eval("myclass = MyClass('a string')");
> jep.invoke("myclass.MyUsefulFunction()");
>
> That'd work...
>
> On Wed, 2008-07-02 at 12:27 +0200, Edoardo Tosca wrote:
> > class MyClass:
> >      def  __init__(self, param):
> >          self.myParam = param
> >      def MyUsefulFunction():
> >          mylist = list()
> >          #do something using self.myParam and filling mylist
> >          return mylist
> >
> > In my Java code i call (in a block surrounded with try-catch):
> >    jep.runScript("myclass.py");
> >    jep.invoke("MyClass", "the parameter");
> > //Until here it works fine
> >    jep.invoke("MyUsefulFunction");
>
>
>
> -------------------------------------------------------------------------
> Sponsored by: SourceForge.net Community Choice Awards: VOTE NOW!
> Studies have shown that voting for your favorite open source project,
> along with a healthy diet, reduces your potential for chronic lameness
> and boredom. Vote Now at http://www.sourceforge.net/community/cca08
> _______________________________________________
> Jepp-users mailing list
> Jep...@li...
> https://lists.sourceforge.net/lists/listinfo/jepp-users
>



-- 
Edoardo Tosca
Mail: edoardo.tosca@{autistici.org, gmail.com}
Skype: edoardotosca

Re: [Jepp-users] Call a function of a Python class in a module

[Mike J. <mrj...@mi...>]

On second thought, I didn't fully answer the question. :-)

The first invoke works because Python classes are implemented as
Callable objects. So the constructor is just a function that returns an
object that has functions of it's own, more or less.

The second fails because there's no MyUsefulFunction in the scope, it's
part of MyClass.

If you did something like:
jep.eval("myclass = MyClass('a string')");
jep.invoke("myclass.MyUsefulFunction()");

That'd work...

On Wed, 2008-07-02 at 12:27 +0200, Edoardo Tosca wrote:
> class MyClass:
>      def  __init__(self, param):
>          self.myParam = param
>      def MyUsefulFunction():
>          mylist = list()
>          #do something using self.myParam and filling mylist
>          return mylist
> 
> In my Java code i call (in a block surrounded with try-catch):
>    jep.runScript("myclass.py");
>    jep.invoke("MyClass", "the parameter");
> //Until here it works fine
>    jep.invoke("MyUsefulFunction");

Re: [Jepp-users] Call a function of a Python class in a module

[Mike J. <mrj...@mi...>]

There's not much support there. I'd started down that road (there's a
bunch of source for PyObject and PyClass) but I didn't find it all that
useful and I didn't want to break people's code.

It would be best if you could have a few functions that you'd call from
Java to handle creating classes and whatnot in python. But if you're not
able to do that, you can take a look at jproxy.

Here's an example with a few changes to your script:

from jep import jproxy

class MyClass:
     def  __init__(self, param):
          self.myParam = param
     def MyUsefulFunction(self):
          mylist = list()
          #do something using self.myParam and filling mylist
          return mylist
     def toString(self):
          return self.myParam

def proxyit(param):
     return jproxy(MyClass(param), ['testjep.MyClass'])

interface:
package testjep;

public interface MyClass {
    public Object MyUsefulFunction();
}


java:
Jep jep = new Jep(false);
jep.runScript("getclass.py");
Object ret = jep.invoke("proxyit", "the parameter");
System.out.println("ret = " + ((MyClass) ret).MyUsefulFunction());
jep.close();


Although you'll notice the list returned from MyUseful gets converted to
a string. Lists aren't automatically converted to arrays. You'd want to
go ahead and instantiate a java collection or use the jarray primitive
type.

On Wed, 2008-07-02 at 12:27 +0200, Edoardo Tosca wrote:
> Hello
> I have found this awesome library that will be useful for my master
> thesis.
> Anyway i have some questions.
> The first one related to the mail subject is:
> Can i invoke a Python function of a Python class?
> 
> here is a simple example:
> 
> class MyClass:
>      def  __init__(self, param):
>          self.myParam = param
>      def MyUsefulFunction():
>          mylist = list()
>          #do something using self.myParam and filling mylist
>          return mylist
> 
> In my Java code i call (in a block surrounded with try-catch):
>    jep.runScript("myclass.py");
>    jep.invoke("MyClass", "the parameter");
> //Until here it works fine
>    jep.invoke("MyUsefulFunction");
> 
> Thank you all
> 
> edo
> 
> 
> 
> 
> 
> 
> 
> -------------------------------------------------------------------------
> Sponsored by: SourceForge.net Community Choice Awards: VOTE NOW!
> Studies have shown that voting for your favorite open source project,
> along with a healthy diet, reduces your potential for chronic lameness
> and boredom. Vote Now at http://www.sourceforge.net/community/cca08
> _______________________________________________ Jepp-users mailing list Jep...@li... https://lists.sourceforge.net/lists/listinfo/jepp-users

[Jepp-users] Call a function of a Python class in a module

[Edoardo T. <edo...@au...>]

Hello
I have found this awesome library that will be useful for my master thesis.
Anyway i have some questions.
The first one related to the mail subject is:
Can i invoke a Python function of a Python class?

here is a simple example:

class MyClass:
     def  __init__(self, param):
         self.myParam = param
     def MyUsefulFunction():
         mylist = list()
         #do something using self.myParam and filling mylist
         return mylist

In my Java code i call (in a block surrounded with try-catch):
   jep.runScript("myclass.py");
   jep.invoke("MyClass", "the parameter");
//Until here it works fine
   jep.invoke("MyUsefulFunction");

Thank you all

edo

Re: [Jepp-users] problem importing scipy

[Andy G. <and...@es...>]

Hi Jon,

great - it works ! Sometimes the answer is closer to home than one 
thinks ;-)

I did not know the trick about PYTHONVERBOSE. I will definitely make use 
of this next time I am stuck.

I will look at enthought. Distributing scipy for Linux platforms looks 
like an interesting i.e. difficult, task.

See you soon

Andy

Jon Wright wrote:
> Andy,
>
> You can see what is happening with python imports inside of jep if you 
> export an environment variable PYTHONVERBOSE=1. This is equivalent to 
> doing
> $ python -v
>
> In this case there was some confusion about jep wanting .pyo files and 
> needing root permission to create them. I did:
>
> $ sudo python -O -c 'import scipy'
>
> ... which created the .pyo files in the scipy installation and then it 
> seemed to load fine.
>
> $ java -cp jep.jar jep.Run console.py
>>>> import scipy
>>>>
>
> It also could be worth trying to see what happed if the optmise flag 
> is turned off in jepp, eg: 'Py_OptimizeFlag = 0' in 
> pyembed.c:pyembed_startup(). There may be eventually issues for other 
> packages which are using zipimports (eg easy_install) that don't have 
> the pyo files. I didn't think python actually optimises much anyway.
>
> For scipy it would be worth looking at the enthought python 
> distribution in case you need to distribute scipy 
> (http://www.enthought.com/products/epd.php) to someone who doesn't 
> have ubuntu and a root password (eg: me on coral ;)
>
> Best,
>
> Jon
>
>
>
>
>
> Andy Gotz wrote:
>> Hi,
>>
>> I am trying to import scipy in jepp but I get a "Package not found" 
>> error. I think it is related to the shared libraries not being found 
>> but I have not dug deeper. Has anyone managed to import scipy from 
>> jepp yet, how ?
>>
>> I am running Java 6, latest version Jepp from svn, scipy 0.6.0 on 
>> Ubuntu 8.04.
>>
>> Any help appreciated.
>>
>> Andy
>>
>> ------------------------------------------------------------------------- 
>>
>> This SF.net email is sponsored by: Microsoft Defy all challenges. 
>> Microsoft(R) Visual Studio 2008. 
>> http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
>> _______________________________________________
>> Jepp-users mailing list
>> Jep...@li...
>> https://lists.sourceforge.net/lists/listinfo/jepp-users

Re: [Jepp-users] problem importing scipy

[Jon W. <wr...@es...>]

Andy,

You can see what is happening with python imports inside of jep if you 
export an environment variable PYTHONVERBOSE=1. This is equivalent to doing
$ python -v

In this case there was some confusion about jep wanting .pyo files and 
needing root permission to create them. I did:

$ sudo python -O -c 'import scipy'

... which created the .pyo files in the scipy installation and then it 
seemed to load fine.

$ java -cp jep.jar jep.Run console.py
>>> import scipy
>>>

It also could be worth trying to see what happed if the optmise flag is 
turned off in jepp, eg: 'Py_OptimizeFlag = 0' in 
pyembed.c:pyembed_startup(). There may be eventually issues for other 
packages which are using zipimports (eg easy_install) that don't have 
the pyo files. I didn't think python actually optimises much anyway.

For scipy it would be worth looking at the enthought python distribution 
in case you need to distribute scipy 
(http://www.enthought.com/products/epd.php) to someone who doesn't have 
ubuntu and a root password (eg: me on coral ;)

Best,

Jon





Andy Gotz wrote:
> Hi,
> 
> I am trying to import scipy in jepp but I get a "Package not found" 
> error. I think it is related to the shared libraries not being found but 
> I have not dug deeper. Has anyone managed to import scipy from jepp yet, 
> how ?
> 
> I am running Java 6, latest version Jepp from svn, scipy 0.6.0 on Ubuntu 
> 8.04.
> 
> Any help appreciated.
> 
> Andy
> 
> -------------------------------------------------------------------------
> This SF.net email is sponsored by: Microsoft 
> Defy all challenges. Microsoft(R) Visual Studio 2008. 
> http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
> _______________________________________________
> Jepp-users mailing list
> Jep...@li...
> https://lists.sourceforge.net/lists/listinfo/jepp-users

[Jepp-users] problem importing scipy

[Andy G. <and...@es...>]

Hi,

I am trying to import scipy in jepp but I get a "Package not found" 
error. I think it is related to the shared libraries not being found but 
I have not dug deeper. Has anyone managed to import scipy from jepp yet, 
how ?

I am running Java 6, latest version Jepp from svn, scipy 0.6.0 on Ubuntu 
8.04.

Any help appreciated.

Andy

Re: [Jepp-users] Why is jep a final class?

[Mike J. <mrj...@mi...>]

It's implemented here:

http://jepp.svn.sourceforge.net/viewvc/jepp/trunk/jep/src/jep/JepScriptEngine.java?revision=366&view=markup

You might not want to use that one though. The way the jsr was designed
means I had to run jep in interactive mode, which is much slower than
running scripts. The bummer about it is they decided to send me a Reader
instance instead of a file. I understand why, but it kills a lot of
performance for those languages that byte compile. I went so far as
writing the script out to a temp file first.

By script engine, I didn't mean you necessarily had to use the
javax.script stuff, but you can make your own, too.

On Fri, 2008-05-16 at 08:43 +0200, Jonathan Wright wrote:
> Hi Mike,
> 
> The reason for inheritance is that I normally speak python, and so the 
> concepts of 'API' and 'safety' are just alien :-) A locally designed api 
> would need to be maintained and also need its own javadocs and 
> unittests. It's not just laziness - imagine that next week you add a 
> cool new feature - with inheritance we can use it straightaway. If we 
> break things by overriding a method with inheritance we can still 
> switch  back to using a vanilla jep object instead of our noisy_jep, 
> even at runtime.
> 
> Your "script engine" idea sounds interesting - I had not heard of them 
> until now. Google shows there is a "JSR 223" which specifies an 
> "interface" for scripting engines? Jepp seems to be listed in the "for 
> info only" section at https://scripting.dev.java.net/ ? This 
> javax.script magic looks like the sort of API which we should be very 
> happy with. I'm glad we didn't write our own api already ;-) 
> 
> In case there is a lot of work to do in burying jepp inside a JSR 223 
> interface, we could try to help out...?
> 
> All the best,
> 
> Jon
> 
> 
> 
> Mike Johnson wrote:
> > Hi,
> >
> > It's final simply because there's a lot of nuanced action going on in there and I didn't want to have to worry about methods being overridden. Quite simply, I didn't think it'd be useful so I didn't bother designing for it. And hey, you're the first. :-)
> >
> > Anyway, you could download the source and compile without the final keyword. You don't even have to recompile the c library, so that's pretty easy.
> >
> > Although, I gotta ask why you wouldn't make your own script engine / class / whatever and then call Jep? You'd be much safer from future API changes.
> >
> >
> >   
> >> On Thu May 15 12:14 , Jonathan Wright <wr...@es...> wrote:Dear Mike, and Jepp-users,
> >>
> >> The "final" word in
> >>
> >> " public final class Jep { ..."
> >>
> >> ...apparently stops us from inheriting from Jep? I was hoping to find a 
> >> way to keep log of the python interpreter commands, eg:
> >>
> >> class noisy_jep extends Jep {
> >>    //...
> >>    public boolean eval(String str) throws JepException {
> >>      logger.info("jep.eval of: "+str); // record what we did
> >>      return super.eval(str);
> >>    }
> >>    // also set etc...
> >> }
> >>
> >>
> >> Is there some reason for making Jep a final class? Wikipedia only 
> >> suggested efficiency and security, but since we're calling python code, 
> >> those don't seem to matter. Is it to do with the JNI?
> >>
> >> Can anyone suggest a better way to record what commands have passed 
> >> through Jep? The aim is to have an eclipse gui driving some scientific 
> >> python codes with a record / playback macro (eg python script) type 
> >> functionality.
> >>
> >> Thanks,
> >>
> >> Jon
> >>
> >> PS: I made it to the end of the mail without complaining that private 
> >> and final exist as keywords in Java at all.
> >>
> >> -------------------------------------------------------------------------
> >> This SF.net email is sponsored by: Microsoft 
> >> Defy all challenges. Microsoft(R) Visual Studio 2008. 
> >> http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
> >> _______________________________________________
> >> Jepp-users mailing list
> >> Jep...@li...
> >> https://lists.sourceforge.net/lists/listinfo/jepp-users
> >>
> >>     
> >
> >   
> 
> 
> -------------------------------------------------------------------------
> This SF.net email is sponsored by: Microsoft 
> Defy all challenges. Microsoft(R) Visual Studio 2008. 
> http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
> _______________________________________________
> Jepp-users mailing list
> Jep...@li...
> https://lists.sourceforge.net/lists/listinfo/jepp-users

Re: [Jepp-users] Why is jep a final class?

[Jonathan W. <wr...@es...>]

Hi Mike,

The reason for inheritance is that I normally speak python, and so the 
concepts of 'API' and 'safety' are just alien :-) A locally designed api 
would need to be maintained and also need its own javadocs and 
unittests. It's not just laziness - imagine that next week you add a 
cool new feature - with inheritance we can use it straightaway. If we 
break things by overriding a method with inheritance we can still 
switch  back to using a vanilla jep object instead of our noisy_jep, 
even at runtime.

Your "script engine" idea sounds interesting - I had not heard of them 
until now. Google shows there is a "JSR 223" which specifies an 
"interface" for scripting engines? Jepp seems to be listed in the "for 
info only" section at https://scripting.dev.java.net/ ? This 
javax.script magic looks like the sort of API which we should be very 
happy with. I'm glad we didn't write our own api already ;-) 

In case there is a lot of work to do in burying jepp inside a JSR 223 
interface, we could try to help out...?

All the best,

Jon



Mike Johnson wrote:
> Hi,
>
> It's final simply because there's a lot of nuanced action going on in there and I didn't want to have to worry about methods being overridden. Quite simply, I didn't think it'd be useful so I didn't bother designing for it. And hey, you're the first. :-)
>
> Anyway, you could download the source and compile without the final keyword. You don't even have to recompile the c library, so that's pretty easy.
>
> Although, I gotta ask why you wouldn't make your own script engine / class / whatever and then call Jep? You'd be much safer from future API changes.
>
>
>   
>> On Thu May 15 12:14 , Jonathan Wright <wr...@es...> wrote:Dear Mike, and Jepp-users,
>>
>> The "final" word in
>>
>> " public final class Jep { ..."
>>
>> ...apparently stops us from inheriting from Jep? I was hoping to find a 
>> way to keep log of the python interpreter commands, eg:
>>
>> class noisy_jep extends Jep {
>>    //...
>>    public boolean eval(String str) throws JepException {
>>      logger.info("jep.eval of: "+str); // record what we did
>>      return super.eval(str);
>>    }
>>    // also set etc...
>> }
>>
>>
>> Is there some reason for making Jep a final class? Wikipedia only 
>> suggested efficiency and security, but since we're calling python code, 
>> those don't seem to matter. Is it to do with the JNI?
>>
>> Can anyone suggest a better way to record what commands have passed 
>> through Jep? The aim is to have an eclipse gui driving some scientific 
>> python codes with a record / playback macro (eg python script) type 
>> functionality.
>>
>> Thanks,
>>
>> Jon
>>
>> PS: I made it to the end of the mail without complaining that private 
>> and final exist as keywords in Java at all.
>>
>> -------------------------------------------------------------------------
>> This SF.net email is sponsored by: Microsoft 
>> Defy all challenges. Microsoft(R) Visual Studio 2008. 
>> http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
>> _______________________________________________
>> Jepp-users mailing list
>> Jep...@li...
>> https://lists.sourceforge.net/lists/listinfo/jepp-users
>>
>>     
>
>   

Re: [Jepp-users] Why is jep a final class?

[Mike J. <mrj...@mi...>]

Hi,

It's final simply because there's a lot of nuanced action going on in there and I didn't want to have to worry about methods being overridden. Quite simply, I didn't think it'd be useful so I didn't bother designing for it. And hey, you're the first. :-)

Anyway, you could download the source and compile without the final keyword. You don't even have to recompile the c library, so that's pretty easy.

Although, I gotta ask why you wouldn't make your own script engine / class / whatever and then call Jep? You'd be much safer from future API changes.


>On Thu May 15 12:14 , Jonathan Wright <wr...@es...> wrote:Dear Mike, and Jepp-users,
>
>The "final" word in
>
>" public final class Jep { ..."
>
>...apparently stops us from inheriting from Jep? I was hoping to find a 
>way to keep log of the python interpreter commands, eg:
>
>class noisy_jep extends Jep {
>    //...
>    public boolean eval(String str) throws JepException {
>      logger.info("jep.eval of: "+str); // record what we did
>      return super.eval(str);
>    }
>    // also set etc...
>}
>
>
>Is there some reason for making Jep a final class? Wikipedia only 
>suggested efficiency and security, but since we're calling python code, 
>those don't seem to matter. Is it to do with the JNI?
>
>Can anyone suggest a better way to record what commands have passed 
>through Jep? The aim is to have an eclipse gui driving some scientific 
>python codes with a record / playback macro (eg python script) type 
>functionality.
>
>Thanks,
>
>Jon
>
>PS: I made it to the end of the mail without complaining that private 
>and final exist as keywords in Java at all.
>
>-------------------------------------------------------------------------
>This SF.net email is sponsored by: Microsoft 
>Defy all challenges. Microsoft(R) Visual Studio 2008. 
>http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
>_______________________________________________
>Jepp-users mailing list
>Jep...@li...
>https://lists.sourceforge.net/lists/listinfo/jepp-users
>

[Jepp-users] Why is jep a final class?

[Jonathan W. <wr...@es...>]

Dear Mike, and Jepp-users,

The "final" word in

" public final class Jep { ..."

...apparently stops us from inheriting from Jep? I was hoping to find a 
way to keep log of the python interpreter commands, eg:

class noisy_jep extends Jep {
    //...
    public boolean eval(String str) throws JepException {
      logger.info("jep.eval of: "+str); // record what we did
      return super.eval(str);
    }
    // also set etc...
}


Is there some reason for making Jep a final class? Wikipedia only 
suggested efficiency and security, but since we're calling python code, 
those don't seem to matter. Is it to do with the JNI?

Can anyone suggest a better way to record what commands have passed 
through Jep? The aim is to have an eclipse gui driving some scientific 
python codes with a record / playback macro (eg python script) type 
functionality.

Thanks,

Jon

PS: I made it to the end of the mail without complaining that private 
and final exist as keywords in Java at all.

Re: [Jepp-users] java.lang.UnsatisfiedLinkError: no jep in java.library.path

[<Pau...@no...>]

We're running on Windows and the failures are not restricted to specific
machines but happen occasionally on various machines.

Our init scripts construct a PATH for the rest of the tools to include a
number of executables and libs, including Jepp, which are bundled in our
tools package. As a result the PATH can vary depending on where users
locate the tools package. This might be a source of problems but as
mentioned I have even experienced this occurring and then running
exactly the same command worked without changing any other variables.

paul
 

>-----Original Message-----
>From: ext Andy Gotz [mailto:and...@es...] 
>Sent: Friday, March 28, 2008 10:54 PM
>To: Mackay Paul (Nokia-D-MSW/Vancouver)
>Cc: jep...@li...
>Subject: Re: [Jepp-users] java.lang.UnsatisfiedLinkError: no 
>jep in java.library.path
>
>Hi,
>
>we are using Jepp from inside eclipse rcp applications on 
>Linux, Windows and Macintosh and we have packaged it as part 
>of a plugin. We set the Bundle-NativeCode  property in the 
>plugin MANIFEST. We have never encountered this problem except 
>when Python is not found or if on Macintosh the symbolic link 
>is missing in /Library/Extension/Java. In all other cases jepp 
>works like a charm. What platform are you having problems with ?
>
>Andy
>
>Pau...@no... wrote:
>> Hi,
>>  
>> We've started using Jepp and find it very useful. 
>Occasionally we are 
>> experiencing an UnsatisfiedLinkError when the Jep class 
>tries to load 
>> the jep library. This is sporadic and not obviously reproducible.
>>  
>> Is anyone aware of any reasons why this might be happening? It does 
>> work most of the time so it is not that we do not have the 
>libs on the 
>> java.library.path. One thing is that we dont install Jepp on every 
>> machine but have taken a copy and included it with other tools.
>>  
>> thanks
>>
>> paul
>>
>>  
>> 
>----------------------------------------------------------------------
>> --
>>
>> 
>----------------------------------------------------------------------
>> --- Check out the new SourceForge.net Marketplace.
>> It's the best place to buy or sell services for just about anything 
>> Open Source.
>> 
>http://ad.doubleclick.net/clk;164216239;13503038;w?http://sf.net/marke
>> tplace
>> 
>----------------------------------------------------------------------
>> --
>>
>> _______________________________________________
>> Jepp-users mailing list
>> Jep...@li...
>> https://lists.sourceforge.net/lists/listinfo/jepp-users
>>   
>
>

Re: [Jepp-users] java.lang.UnsatisfiedLinkError: no jep in java.library.path

[Andy G. <and...@es...>]

Hi,

we are using Jepp from inside eclipse rcp applications on Linux, Windows 
and Macintosh and we have packaged it as part of a plugin. We set the 
Bundle-NativeCode  property in the plugin MANIFEST. We have never 
encountered this problem except when Python is not found or if on 
Macintosh the symbolic link is missing in /Library/Extension/Java. In 
all other cases jepp works like a charm. What platform are you having 
problems with ?

Andy

Pau...@no... wrote:
> Hi,
>  
> We've started using Jepp and find it very useful. Occasionally we are 
> experiencing an UnsatisfiedLinkError when the Jep class tries to load 
> the jep library. This is sporadic and not obviously reproducible.
>  
> Is anyone aware of any reasons why this might be happening? It does 
> work most of the time so it is not that we do not have the libs on the 
> java.library.path. One thing is that we dont install Jepp on every 
> machine but have taken a copy and included it with other tools.
>  
> thanks
>
> paul
>
>  
> ------------------------------------------------------------------------
>
> -------------------------------------------------------------------------
> Check out the new SourceForge.net Marketplace.
> It's the best place to buy or sell services for
> just about anything Open Source.
> http://ad.doubleclick.net/clk;164216239;13503038;w?http://sf.net/marketplace
> ------------------------------------------------------------------------
>
> _______________________________________________
> Jepp-users mailing list
> Jep...@li...
> https://lists.sourceforge.net/lists/listinfo/jepp-users
>   

Re: [Jepp-users] java.lang.UnsatisfiedLinkError: no jep in java.library.path

[Mike J. <mrj...@mi...>]

Hi,

>We've started using 
>Jepp and find it very useful.

Awesome. :-)

> Occasionally we are experiencing an 
>UnsatisfiedLinkError when the Jep class tries to load the jep library. This is 
>sporadic and not obviously reproducible.

Do you mean you get that exception on some machines while others work, or that on the same machine you get that sporadically when loading?

That doesn't really make sense to me -- if it works once it should work every time unless something on the OS changed. One thing to keep in mind though, that it might not be jep.so that it couldn't find. If the linker fails to find anything jep depends on it'll fail with the same message. Using ldd on the library (depends.exe on Windows will do this I think) will tell you what the linker will try to use.

I find it's easier to not depend on java.library.path but to put jep and dependencies in a system location...

(Sorry for the formatting, on a web client.)
- Mike

[Jepp-users] java.lang.UnsatisfiedLinkError: no jep in java.library.path

[<Pau...@no...>]

Hi,
 
We've started using Jepp and find it very useful. Occasionally we are
experiencing an UnsatisfiedLinkError when the Jep class tries to load
the jep library. This is sporadic and not obviously reproducible.
 
Is anyone aware of any reasons why this might be happening? It does work
most of the time so it is not that we do not have the libs on the
java.library.path. One thing is that we dont install Jepp on every
machine but have taken a copy and included it with other tools.
 
thanks

paul 

 

Re: [Jepp-users] Jepp on MacOSX

[Andy G. <and...@es...>]

Hi,

oops ! I made the common mistake of not reading the README again !

The answer is you need the symlink in /Library/Java/Extensions. To quote 
the README :

make a symlink from libjep.dynlib to /Library/Java/Extentions, like:
                sudo ln -s `pwd`/.libs/libjep.dylib \
                    /Library/Java/Extensions/libjep.jnilib

I wonder why this is different on MacOSX ?

Thanks !

Andy

Mike Johnson wrote:
> Hi,
>
> Did you check the README included with the source?
>
> On Tue, 2008-03-18 at 11:59 +0100, andy gotz wrote:
>   
>> Hi,
>>
>> I am using jepp on Linux, Windows and MacOSX and am very happy with it. 
>> Thanks !
>>
>> I have a small question concerning jepp on macosx. Despite the fact that 
>> I have it working on my mac mini at home I have forgotten how I setup my 
>> environment to make it work. I have installed it on a different machine 
>> now (both are running Leopard) and now I cannot instantiate Jep anymore. 
>> Can anyone remind of the recipe and environment variables required to 
>> make Jep run under a MacOSX environment ?
>>
>> Your help will save me a lot of time.
>>
>> Thanks
>>
>> Andy
>>
>> -------------------------------------------------------------------------
>> This SF.net email is sponsored by: Microsoft
>> Defy all challenges. Microsoft(R) Visual Studio 2008.
>> http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
>> _______________________________________________
>> Jepp-users mailing list
>> Jep...@li...
>> https://lists.sourceforge.net/lists/listinfo/jepp-users
>>     

Re: [Jepp-users] Jepp on MacOSX

[Mike J. <mrj...@mi...>]

Hi,

Did you check the README included with the source?

On Tue, 2008-03-18 at 11:59 +0100, andy gotz wrote:
> Hi,
> 
> I am using jepp on Linux, Windows and MacOSX and am very happy with it. 
> Thanks !
> 
> I have a small question concerning jepp on macosx. Despite the fact that 
> I have it working on my mac mini at home I have forgotten how I setup my 
> environment to make it work. I have installed it on a different machine 
> now (both are running Leopard) and now I cannot instantiate Jep anymore. 
> Can anyone remind of the recipe and environment variables required to 
> make Jep run under a MacOSX environment ?
> 
> Your help will save me a lot of time.
> 
> Thanks
> 
> Andy
> 
> -------------------------------------------------------------------------
> This SF.net email is sponsored by: Microsoft
> Defy all challenges. Microsoft(R) Visual Studio 2008.
> http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
> _______________________________________________
> Jepp-users mailing list
> Jep...@li...
> https://lists.sourceforge.net/lists/listinfo/jepp-users

[Jepp-users] Jepp on MacOSX

[andy g. <and...@es...>]

Hi,

I am using jepp on Linux, Windows and MacOSX and am very happy with it. 
Thanks !

I have a small question concerning jepp on macosx. Despite the fact that 
I have it working on my mac mini at home I have forgotten how I setup my 
environment to make it work. I have installed it on a different machine 
now (both are running Leopard) and now I cannot instantiate Jep anymore. 
Can anyone remind of the recipe and environment variables required to 
make Jep run under a MacOSX environment ?

Your help will save me a lot of time.

Thanks

Andy

Re: [Jepp-users] [mailinglist] Re: New Windows binaries ?

[Uwe S. <usc...@mi...>]

Mike Johnson schrieb:
> Aight, there are new installers available. Let me know if there are
> problems...
>
>
Hi,
installing jepp works. I can use "invoke", "runScript" and "eval" without any problems. But one issue is still unsolved:

http://sourceforge.net/mailarchive/message.php?msg_name=473C2EFB.8080205%40mineway.de

When I create a jep.array instance in Python and pass it back
to Java the Type is "Object" and I can not cast it to "Object[]",
"Integer[]" or so. I allways get an ClassCastException.

Greetings, Uwe




-- 
Dr. rer. nat. Uwe Schmitt
F&E Mathematik
 
mineway GmbH
Science Park 2
D-66123 Saarbrücken
 
Telefon: +49 (0)681 8390 5334
Telefax: +49 (0)681 830 4376
 
usc...@mi...
www.mineway.de
 
Geschäftsführung: Dr.-Ing. Mathias Bauer
Amtsgericht Saarbrücken HRB 12339

Re: [Jepp-users] New Windows binaries ?

[Mike J. <mrj...@mi...>]

Aight, there are new installers available. Let me know if there are
problems...


On Fri, 2008-02-22 at 08:37 +0100, Uwe Schmitt wrote:
> Am 22.02.2008 um 05:02 schrieb Mike Johnson:
> 
> > Sorry guys, I've been out of town. I'll try to build and release a new
> > Jep this weekend...
> 
> Thank you very much.
> Greetings, Uwe
> 
> >
> >
> >
> > On Thu, 2008-02-21 at 23:23 +0100, Jon Wright wrote:
> >> Dear Uwe,
> >>
> >> I think there is something buried in svn on the fable project:
> >>
> >> http://fable.svn.sourceforge.net/viewvc/fable/gui/trunk/jep/lib/ 
> >> win32/
> >>
> >> ... and jep.jar is one or two dirs up. Thought we posted a batch file
> >> for building on windows here a while ago? Do you have this problem  
> >> with
> >> the build we're using? It apparently works for the things we use it  
> >> for
> >> at least.
> >>
> >> Best,
> >>
> >> Jon
> >>
> >> Uwe Schmitt wrote:
> >>> Uwe Schmitt schrieb:
> >>>>> Hi,
> >>>>>
> >>>>> I tried building jepp from actual sources, but
> >>>>> did not succeed.
> >>>>> Do you plan to provide new binary versions for python 2.5
> >>>>> ?
> >>>>>
> >>>>> Greetings, Uwe.
> >>>>>
> >>>
> >>>> Meanwhile I succeeded in building an actuar jar for python 2.5 and
> >>> jdk > 1.5.
> >>>> If anybody is interested feel free to contact me.
> >>>
> >>> I have to draw back my anouncmenet. "Jep j = new Jep(...)" works  
> >>> fine,
> >>> even "j.eval(...)" and "j.getValue(...)" makes no problems, but
> >>> "j.runScript(..)" crashes the Java virtual machine:
> >>>
> >>>    #
> >>>    # An unexpected error has been detected by HotSpot Virtual  
> >>> Machine:
> >>>    #
> >>>    #  EXCEPTION_ACCESS_VIOLATION (0xc0000005) at pc=0x7c928fea,
> >>> pid=4724, tid=8696
> >>>    #
> >>>    # Java VM: Java HotSpot(TM) Client VM (1.5.0_14-b03 mixed mode)
> >>>    # Problematic frame:
> >>>    # C  [ntdll.dll+0x18fea
> >>>
> >>> Any hints ?
> >>>
> >>> Greetings, Uwe.
> >>>
> >>
> >> -------------------------------------------------------------------------
> >> This SF.net email is sponsored by: Microsoft
> >> Defy all challenges. Microsoft(R) Visual Studio 2008.
> >> http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
> >> _______________________________________________
> >> Jepp-users mailing list
> >> Jep...@li...
> >> https://lists.sourceforge.net/lists/listinfo/jepp-users
> >
> 
> 
> -------------------------------------------------------------------------
> This SF.net email is sponsored by: Microsoft
> Defy all challenges. Microsoft(R) Visual Studio 2008.
> http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
> _______________________________________________
> Jepp-users mailing list
> Jep...@li...
> https://lists.sourceforge.net/lists/listinfo/jepp-users

Re: [Jepp-users] New Windows binaries ?

[Uwe S. <usc...@mi...>]

Am 22.02.2008 um 05:02 schrieb Mike Johnson:

> Sorry guys, I've been out of town. I'll try to build and release a new
> Jep this weekend...

Thank you very much.
Greetings, Uwe

>
>
>
> On Thu, 2008-02-21 at 23:23 +0100, Jon Wright wrote:
>> Dear Uwe,
>>
>> I think there is something buried in svn on the fable project:
>>
>> http://fable.svn.sourceforge.net/viewvc/fable/gui/trunk/jep/lib/ 
>> win32/
>>
>> ... and jep.jar is one or two dirs up. Thought we posted a batch file
>> for building on windows here a while ago? Do you have this problem  
>> with
>> the build we're using? It apparently works for the things we use it  
>> for
>> at least.
>>
>> Best,
>>
>> Jon
>>
>> Uwe Schmitt wrote:
>>> Uwe Schmitt schrieb:
>>>>> Hi,
>>>>>
>>>>> I tried building jepp from actual sources, but
>>>>> did not succeed.
>>>>> Do you plan to provide new binary versions for python 2.5
>>>>> ?
>>>>>
>>>>> Greetings, Uwe.
>>>>>
>>>
>>>> Meanwhile I succeeded in building an actuar jar for python 2.5 and
>>> jdk > 1.5.
>>>> If anybody is interested feel free to contact me.
>>>
>>> I have to draw back my anouncmenet. "Jep j = new Jep(...)" works  
>>> fine,
>>> even "j.eval(...)" and "j.getValue(...)" makes no problems, but
>>> "j.runScript(..)" crashes the Java virtual machine:
>>>
>>>    #
>>>    # An unexpected error has been detected by HotSpot Virtual  
>>> Machine:
>>>    #
>>>    #  EXCEPTION_ACCESS_VIOLATION (0xc0000005) at pc=0x7c928fea,
>>> pid=4724, tid=8696
>>>    #
>>>    # Java VM: Java HotSpot(TM) Client VM (1.5.0_14-b03 mixed mode)
>>>    # Problematic frame:
>>>    # C  [ntdll.dll+0x18fea
>>>
>>> Any hints ?
>>>
>>> Greetings, Uwe.
>>>
>>
>> -------------------------------------------------------------------------
>> This SF.net email is sponsored by: Microsoft
>> Defy all challenges. Microsoft(R) Visual Studio 2008.
>> http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
>> _______________________________________________
>> Jepp-users mailing list
>> Jep...@li...
>> https://lists.sourceforge.net/lists/listinfo/jepp-users
>

Re: [Jepp-users] New Windows binaries ?

[Mike J. <mrj...@mi...>]

That's usually caused by mixing C runtimes... It crashes on string
functions or file accesses.

Because Java uses MSVC, you have to compile using whatever version they
use...

On Thu, 2008-02-21 at 16:58 +0100, Uwe Schmitt wrote:
> Uwe Schmitt schrieb:
> >
> >> Hi,
> >>
> >> I tried building jepp from actual sources, but
> >> did not succeed.
> >> Do you plan to provide new binary versions for python 2.5
> >> ?
> >>
> >> Greetings, Uwe.
> >>
> 
>  > Meanwhile I succeeded in building an actuar jar for python 2.5 and 
> jdk > 1.5.
>  > If anybody is interested feel free to contact me.
> 
> I have to draw back my anouncmenet. "Jep j = new Jep(...)" works fine,
> even "j.eval(...)" and "j.getValue(...)" makes no problems, but
> "j.runScript(..)" crashes the Java virtual machine:
> 
>     #
>     # An unexpected error has been detected by HotSpot Virtual Machine:
>     #
>     #  EXCEPTION_ACCESS_VIOLATION (0xc0000005) at pc=0x7c928fea, 
> pid=4724, tid=8696
>     #
>     # Java VM: Java HotSpot(TM) Client VM (1.5.0_14-b03 mixed mode)
>     # Problematic frame:
>     # C  [ntdll.dll+0x18fea
> 
> Any hints ?
> 
> Greetings, Uwe.
> 

Re: [Jepp-users] New Windows binaries ?

[Mike J. <mrj...@mi...>]

Sorry guys, I've been out of town. I'll try to build and release a new
Jep this weekend...


On Thu, 2008-02-21 at 23:23 +0100, Jon Wright wrote:
> Dear Uwe,
> 
> I think there is something buried in svn on the fable project:
> 
> http://fable.svn.sourceforge.net/viewvc/fable/gui/trunk/jep/lib/win32/
> 
> ... and jep.jar is one or two dirs up. Thought we posted a batch file 
> for building on windows here a while ago? Do you have this problem with 
> the build we're using? It apparently works for the things we use it for 
> at least.
> 
> Best,
> 
> Jon
> 
> Uwe Schmitt wrote:
> > Uwe Schmitt schrieb:
> >>> Hi,
> >>>
> >>> I tried building jepp from actual sources, but
> >>> did not succeed.
> >>> Do you plan to provide new binary versions for python 2.5
> >>> ?
> >>>
> >>> Greetings, Uwe.
> >>>
> > 
> >  > Meanwhile I succeeded in building an actuar jar for python 2.5 and 
> > jdk > 1.5.
> >  > If anybody is interested feel free to contact me.
> > 
> > I have to draw back my anouncmenet. "Jep j = new Jep(...)" works fine,
> > even "j.eval(...)" and "j.getValue(...)" makes no problems, but
> > "j.runScript(..)" crashes the Java virtual machine:
> > 
> >     #
> >     # An unexpected error has been detected by HotSpot Virtual Machine:
> >     #
> >     #  EXCEPTION_ACCESS_VIOLATION (0xc0000005) at pc=0x7c928fea, 
> > pid=4724, tid=8696
> >     #
> >     # Java VM: Java HotSpot(TM) Client VM (1.5.0_14-b03 mixed mode)
> >     # Problematic frame:
> >     # C  [ntdll.dll+0x18fea
> > 
> > Any hints ?
> > 
> > Greetings, Uwe.
> > 
> 
> -------------------------------------------------------------------------
> This SF.net email is sponsored by: Microsoft
> Defy all challenges. Microsoft(R) Visual Studio 2008.
> http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
> _______________________________________________
> Jepp-users mailing list
> Jep...@li...
> https://lists.sourceforge.net/lists/listinfo/jepp-users

Re: [Jepp-users] New Windows binaries ?

[Jon W. <wr...@es...>]

Dear Uwe,

I think there is something buried in svn on the fable project:

http://fable.svn.sourceforge.net/viewvc/fable/gui/trunk/jep/lib/win32/

... and jep.jar is one or two dirs up. Thought we posted a batch file 
for building on windows here a while ago? Do you have this problem with 
the build we're using? It apparently works for the things we use it for 
at least.

Best,

Jon

Uwe Schmitt wrote:
> Uwe Schmitt schrieb:
>>> Hi,
>>>
>>> I tried building jepp from actual sources, but
>>> did not succeed.
>>> Do you plan to provide new binary versions for python 2.5
>>> ?
>>>
>>> Greetings, Uwe.
>>>
> 
>  > Meanwhile I succeeded in building an actuar jar for python 2.5 and 
> jdk > 1.5.
>  > If anybody is interested feel free to contact me.
> 
> I have to draw back my anouncmenet. "Jep j = new Jep(...)" works fine,
> even "j.eval(...)" and "j.getValue(...)" makes no problems, but
> "j.runScript(..)" crashes the Java virtual machine:
> 
>     #
>     # An unexpected error has been detected by HotSpot Virtual Machine:
>     #
>     #  EXCEPTION_ACCESS_VIOLATION (0xc0000005) at pc=0x7c928fea, 
> pid=4724, tid=8696
>     #
>     # Java VM: Java HotSpot(TM) Client VM (1.5.0_14-b03 mixed mode)
>     # Problematic frame:
>     # C  [ntdll.dll+0x18fea
> 
> Any hints ?
> 
> Greetings, Uwe.
>