Favicon

RE: Absolute Paths and the Resource Manager

Yeah, 
I noticed later that this is supposed to be a URL and not just a path so
if I put file:/// it works.  I don't think that there is any problem
with the way the Resource Manager is getting the files. 

But this doesn't really address the problem that if I have an analysis
engine descriptor that imports a type system then the CDE won't let me
edit the location with the regular tools if the location of the type
system descriptor isn't already on the datapath (if it is imported by
name) or the file location actually exists (if it is imported by
location).  To see this try making a type system descriptor and an
analysis engine in the same directory that references it.  Then delete
the type system descriptor.  Now try to edit the analysis engine
descriptor.  There isn't any way to edit the type system import except
by editing the xml directly and there isn't any way to change the
datapath except by noticing that there is an eclipse project property
UIMA CDE property page.  This happens whether the original type system
was imported by name or location.  It isn't a major gripe but the CDE is
so nice to use it is disconcerting to be forced to edit the XML
directly.

-----Original Message-----
From: Marshall Schor [mailto:msa@...] 
Sent: Friday, January 25, 2008 10:28 PM
To: uima-user@...
Subject: Re: Absolute Paths and the Resource Manager

Hi Frank,

Is the type system <import...> using the location="...." form?  Is that 
(Continue reading)

David Buttler | 1 Feb 20:47

Re: Processing collections as a set of documents

In terms of clustering, we have only done incremental clustering in 
UIMA. Essentially, the clustering component keeps track of all of the 
clusters, and as new documents come in, updates the appropriate cluster. 
Other types of clustering we do externally. 

We are still stuck on UIMA 1.4, so the hacks we use are probably not 
appropriate for the more recent versions. 

Some ideas: if you are processing a parallel corpus (i.e. for machine 
translation), the reader could create a single CAS for each pair of 
documents. Different languages could go in different SOFAs. Then your 
subsequent components would have to know how to deal with the different 
sofas (e.g. by only using an English POS parser on the English sofa)

For multi pass algorithms, I would suggest different pipelines: one to 
collect corpus statistics, and the next pipeline to use them.  If 
incremental statistics are sufficient, then you can just create an 
internal data structure to manage collection statistics as you stream 
through your collection.

Dave

Nicolas Hernandez wrote:
> Hi,
>
> Making my first cpe, I am wondering how to deal with NLP tasks which
> aims at processing several documents (i.e. pair of collection of
> documents considered as a single entity) in a time. I am thinking
> about applications such as (multilingual) text alignement, or term
> extraction based on measures over a corpus, or text clustering (how to
(Continue reading)

Marshall Schor | 1 Feb 22:18

Re: Absolute Paths and the Resource Manager

LeHouillier, Frank D. wrote:
> Yeah, 
> I noticed later that this is supposed to be a URL and not just a path so
> if I put file:/// it works.  I don't think that there is any problem
> with the way the Resource Manager is getting the files. 
>
> But this doesn't really address the problem that if I have an analysis
> engine descriptor that imports a type system then the CDE won't let me
> edit the location with the regular tools if the location of the type
> system descriptor isn't already on the datapath (if it is imported by
> name) or the file location actually exists (if it is imported by
> location).  To see this try making a type system descriptor and an
> analysis engine in the same directory that references it.  Then delete
> the type system descriptor.  Now try to edit the analysis engine
> descriptor.  There isn't any way to edit the type system import except
> by editing the xml directly
This is "philosophical" problem :-).  The CDE is set up to work with an 
internal representation of the descriptor (and the things that it 
imports).  It doesn't have its own code to do this, it instead relies on 
the underlying UIMA framework to read, parse, and and generally build a 
"model" of what the descriptor is doing.  If it's a type system that's 
gone missing, since the available types condition a lot of the 
functionality of the CDE, it gives up being a "nice" editor, and says 
you have to get this fixed at least to the point where it can be read, 
using a plain-old xml editor.

This probably could be improved, if someone wants to "scratch that itch" 
:-).

-Marshall
(Continue reading)

chandra sekhar | 2 Feb 07:59
Picon

Re: A DOUBT IN UIMA

Respected Sir , These are the information you required.

The entire stack trace when I got the error:
C:\Program Files\IBM\uima\bin>setlocal

C:\Program Files\IBM\uima\bin>call "C:\Program
Files\IBM\uima\bin\setUimaClassPa
th"

C:\Program Files\IBM\uima\bin>set UIMA_CLASSPATH=;C:\Program
Files\IBM\uima\docs
\examples\resources;C:\Program Files\IBM\uima\lib\uima_core.jar;C:\Program
Files
\IBM\uima\lib\uima_cpe.jar;C:\Program
Files\IBM\uima\lib\uima_jcas_builtin_types
.jar;C:\Program Files\IBM\uima\lib\uima_tools.jar;C:\Program
Files\IBM\uima\lib\
uima_examples.jar;C:\Program
Files\IBM\uima\lib\uima_adapter_messaging.jar;C:\Pr
ogram Files\IBM\uima\lib\uima_adapter_soap.jar;C:\Program
Files\IBM\uima\lib\uim
a_adapter_vinci.jar;C:\Program Files\IBM\uima\lib\uima_search.jar;C:\Program
Fil
es\IBM\uima\lib\juru.jar;C:\Program
Files\IBM\uima\lib\siapi.jar;\webapps\axis\W
EB-INF\lib\activation.jar;\webapps\axis\WEB-INF\lib\axis.jar;\webapps\axis\WEB-I
NF\lib\commons-discovery.jar;\webapps\axis\WEB-INF\lib\commons-
discovery-0.2.jar
;\webapps\axis\WEB-INF\lib\commons-logging.jar
;\webapps\axis\WEB-INF\lib\commons
(Continue reading)

Marshall Schor | 2 Feb 13:57

Re: A DOUBT IN UIMA

Hi -

 From examining this information, I can see in the lines below that the 
-cp parameter to the Java call *do not* include the location where you 
have your ProductAnnotator class.

So, Java cannot find the class, and reports the message

Annotator class "com.backstopmedia.uima.tutorial.ProductAnnotator" was not found.

Any time you have classes you want Java to be able find and load, you 
have to put the location of these classes in Java's class path.

There are many ways to fix this.  One way is to set an environment 
variable, CLASSPATH, to have the path to your classes.  In the way you 
are starting up things, I can see that the "setUimaClassPath" batch file 
is being called; it will append the paths it finds from CLASSPATH to the 
path passed to Java.

To do this in your case, in the command window you use to start things, 
type this command to set the environment variable:

set CLASSPATH=C:\Program Files\IBM\uima\docs\examples\bin\com\backstopmedia\uima\tutorial

Also, please verify that this is the path where the 
ProductAnnotator.class file is.

If you do this, then everytime you start the documentAnalyzer in this 
command window, it will include this path when starting Java.

(Continue reading)

chandra sekhar | 2 Feb 15:20
Picon

Re: A DOUBT IN UIMA

Respected Sir ,  I set my classpath as you suggested in your reply in
uima-user archieve. I set path variable in my windows. now my path variable
looks like this:
"C:\Program
Files\IBM\uima\bin";%SystemRoot%\system32;%SystemRoot%;%SystemRoot%\System32\Wbem;C:\Program
Files\Java\jdk1.5.0\bin\;C:\Program
Files\IBM\uima\docs\examples\bin\com\backstopmedia\uima\tutorial

eventhogh I set the path to the location , where my
ProductAnnotator.classfile  is located , I am getting the same error.
Sir ,
Mr.Lally in

http://www-128.ibm.com/developerworks/forums/thread.jspa?threadID=138977&tstart=0

suggested : " in the manifest "MANIFEST.MF" file of the plug-in that
contains the class "de.group.uima.temp.PersonNameAnnotator", insert the
following line:
code <http://www-128.ibm.com/developerworks/forums/>Eclipse-RegisterBuddy:
com.ibm.uima.runtime[/code]  "

for my situation it is : com.ibm.uima.ProductNumberAnnotator.

*I dont know how to find a plugin ,which contains
com.ibm.uima.ProductNumberAnnotator class.*

please help me.
Marshall Schor | 2 Feb 17:58

Re: A DOUBT IN UIMA

Hi -

In your response below, it appears you may have confused "PATH" and 
"CLASSPATH".  These are different things.  At the beginning of your 
response you say you set your "classpath" as suggested, but the example 
you give is not for the "CLASSPATH" but is instead for the "PATH" variable.

Can you confirm you set the "CLASSPATH", and not the "PATH"?

One way to check is to open up a command window and type:

echo %CLASSPATH%

Please do this if you still have troubles, and include this in your 
response note.

chandra sekhar wrote:
> Respected Sir ,  I set my classpath as you suggested in your reply in
> uima-user archieve. I set path variable in my windows. now my path variable
> looks like this:
> "C:\Program
> Files\IBM\uima\bin";%SystemRoot%\system32;%SystemRoot%;%SystemRoot%\System32\Wbem;C:\Program
> Files\Java\jdk1.5.0\bin\;C:\Program
> Files\IBM\uima\docs\examples\bin\com\backstopmedia\uima\tutorial
>
>
> eventhogh I set the path to the location , where my
> ProductAnnotator.classfile  is located , I am getting the same error.
> Sir ,
> Mr.Lally in
(Continue reading)

chandra sekhar | 2 Feb 19:06
Picon

A DOUBT IN UIMA

Respected Sir , I set CLASSPATH variable as you suggested ,

now CLASSPATH  look like this :

CLASSPATH
:%SystemRoot%\system32;%SystemRoot%;%SystemRoot%\System32\Wbem;C:\\Program
Files\\Java\\jdk1.5.0\\bin\;C:\\Program
Files\\IBM\\uima\\docs\\examples\\bin\\com\\backstopmedia\\uima\\tutorial;UIMA_HOME//lib//uima_core.jar

eventhough i am getting same error as previously specified.

Sir , Mr.Lally in

http://www-128.ibm.com/developerworks/forums/thread.jspa?threadID=138977&tstart=0

suggested : " in the manifest "MANIFEST.MF" file of the plug-in that
contains the class "de.group.uima.temp.PersonNameAnnotator", insert the
following line:
code <http://www-128.ibm.com/developerworks/forums/>Eclipse-RegisterBuddy:
com.ibm.uima.runtime[/code]  "

for my situation it is : com.ibm.uima.ProductNumberAnnotator.

*I dont know how to find a plugin ,which contains
com.ibm.uima.ProductNumberAnnotator class.*

please help me.
Adam Lally | 2 Feb 19:55
Picon
Gravatar

Re: A DOUBT IN UIMA

On Feb 2, 2008 1:06 PM, chandra sekhar <sekhar.daiict@...> wrote:
> Respected Sir , I set CLASSPATH variable as you suggested ,
>
> now CLASSPATH  look like this :
>
> CLASSPATH
> :%SystemRoot%\system32;%SystemRoot%;%SystemRoot%\System32\Wbem;C:\\Program
> Files\\Java\\jdk1.5.0\\bin\;C:\\Program
> Files\\IBM\\uima\\docs\\examples\\bin\\com\\backstopmedia\\uima\\tutorial;UIMA_HOME//lib//uima_core.jar
>

Your classpath should contain

C:\Program Files\IBM\uima\docs\examples\bin

instead of

C:\\Program Files\\IBM\\uima\\docs\\examples\\bin\\com\\backstopmedia\\uima\\tutorial

This is a basic Java question, not a UIMA question.  If you aren't
very familiar with using Java then I highly recommend Sun's Java
tutorial: http://java.sun.com/docs/books/tutorial/.

-Adam

Marshall Schor | 2 Feb 20:08

Re: A DOUBT IN UIMA

Hi -

The link below that you give is an error due to an entirely different 
situation that does not apply in your case.  There is no "plugin" in 
your case to find.

The class path value looks wrong in several aspects.  It should only 
have Java Jar files and directories with Java .class files.  It should 
not have the colon (":") character as its first character.

If your files (including the .class files) are in this directory:

c:\Program Files\IBM\uima\docs\examples\bin\com\backstopmedia\uima\tutorial

and Java is trying to load the class        
com.backstopmedia.uima.tutorial.ProductAnnotator

then your class path should be:

c:\Program Files\IBM\uima\docs\examples\bin

Please try setting this as the (only) value of your CLASSPATH. 

Check it by typing into a command window:

echo %CLASSPATH%      you should get the response:

c:\Program Files\IBM\uima\docs\examples\bin

and nothing else. 
(Continue reading)


Gmane