1

I am trouble getting the Maltparser to work with Python NLTK.

Here is my code so far:

import nltk

os.environ["MALT_PARSER"] = "C:/Python34/maltparser-1.8.1"
os.environ["MALTPARSERHOME"] = "C:/Python34/maltparser-1.8.1"

parser8 = nltk.parse.malt.MaltParser(
...        working_dir="C:/Python34/maltparser-1.8.1", mco="engmalt.poly-1.7", 
...        additional_java_args=['-Xmx512m'])

txt = "This is a test sentence"

parser8.raw_parse(txt)

I have downloaded and selected to used a pre-trained model.

This is the response I get: 

runfile('C:/Anaconda/Lib/site-packages/nltk/malt2.py', wdir='C:/Anaconda/Lib/site-packages/nltk')
Traceback (most recent call last):

  File "<ipython-input-38-73069e4ee673>", line 1, in <module>
    runfile('C:/Anaconda/Lib/site-packages/nltk/malt2.py', wdir='C:/Anaconda/Lib/site-packages/nltk')

  File "C:\Anaconda\lib\site-packages\spyderlib\widgets\externalshell\sitecustomize.py", line 580, in runfile
    execfile(filename, namespace)

  File "C:/Anaconda/Lib/site-packages/nltk/malt2.py", line 14, in <module>
    parser8.raw_parse(txt)

  File "C:\Anaconda\lib\site-packages\nltk\parse\malt.py", line 139, in raw_parse
    return self.parse(words, verbose)

  File "C:\Anaconda\lib\site-packages\nltk\parse\malt.py", line 126, in parse
    return self.parse_sents([sentence], verbose)[0]

  File "C:\Anaconda\lib\site-packages\nltk\parse\malt.py", line 114, in parse_sents
    return self.tagged_parse_sents(tagged_sentences, verbose)

  File "C:\Anaconda\lib\site-packages\nltk\parse\malt.py", line 194, in tagged_parse_sents
    "code %d" % (' '.join(cmd), ret))

Exception: MaltParser parsing (java -Xmx512m -jar C:/Python34/maltparser-1.8.1\malt.jar -w C:/Python34/maltparser-1.8.1 -c engmalt.poly-1.7.mco -i C:\Python34\maltparser-1.8.1\malt_input.conllqgpbye -o C:\Python34\maltparser-1.8.1\malt_output.conllib1nx0 -m parse) failed with exit code 2

I have followed all the advice on this post How to use malt parser in python nltk.

Specifically:

-I downloaded the latest version of MaltParser.

-Using Pip, I uninstalled and re installed NLTK to get the latest version, which includes the addition in malt/py that allows 'additional_java_args' to be added as a parameter.

-I renamed the jar file to 'malt.jar'.

-I set an environmental variable pointing both MALT_PARSER and MALTPARSERHOME to the working directory.

-I've tried both the linear and poly pre-trained models.

-The code for malt.py can be found here http://www.nltk.org/_modules/nltk/parse/malt.html

If there isn't an apparent solution, how can I continue to debug this myself?

It seems that there's some slash (/ ) inconsistency with the exception raised. Nothing I do seems to fix it though.

Community
  • 1
  • 1
Santosh Gupta
  • 33
  • 1
  • 5
  • setting environment and all is messed up, the malt.py under development resolves that, but there's still some outstanding issues, see https://github.com/nltk/nltk/issues/943 – alvas Apr 09 '15 at 02:28
  • Note that the latest version of NLTK has this module patched up nicely, see http://stackoverflow.com/questions/33015326/maltparser-giving-error-in-nltk – alvas Oct 08 '15 at 15:10
  • also try and just run "java -Xmx512m -jar C:/Python34/maltparser-1.8.1\malt.jar -w C:/Python34/maltparser-1.8.1 -c engmalt.poly-1.7.mco -i C:\Python34\maltparser-1.8.1\malt_input.conllqgpbye -o C:\Python34\maltparser-1.8.1\malt_output.conllib1nx0 -m parse It may say wrong jave version as it should be 1.8 or higher – brianray Dec 06 '15 at 22:54

0 Answers0