malt parser gives assertion error when using it with nltk

Question

I am using malt parser with python nltk. I have successfully downloaded the training data and updated the latest nltk. When I call the malt parser it gives me an asertion error. Below is the code from python which includes the traceback as well.

 mp = MaltParser("C:/Users/mustufain/Desktop/Python Files/maltparser-1.8.1","C:/Users/mustufain/Desktop/Python Files/maltparser-1.7.2",additional_java_args=['-Xmx512m'])

Traceback (most recent call last):
  File "<pyshell#10>", line 1, in <module>
    mp = MaltParser("C:/Users/mustufain/Desktop/Python Files/maltparser-1.8.1","C:/Users/mustufain/Desktop/Python Files/maltparser-1.7.2",additional_java_args=['-Xmx512m'])
  File "C:\Python34\lib\site-packages\nltk\parse\malt.py", line 131, in __init__
    self.malt_jars = find_maltparser(parser_dirname)
  File "C:\Python34\lib\site-packages\nltk\parse\malt.py", line 72, in find_maltparser
    assert malt_dependencies.issubset(_jars)
AssertionError
>>>

Have you setup: https://github.com/nltk/nltk/wiki/Installing-Third-Party-Software#malt-parser ? — alvas, Feb 19 '16 at 15:30
Do you have ['log4j.jar', 'libsvm.jar', 'liblinear-1.8.jar'] in `C:/Users/mustufain/Desktop/Python Files/maltparser-1.8.1`? — alvas, Feb 19 '16 at 15:32
What is the output of `dir C:/Users/mustufain/Desktop/Python Files/maltparser-1.8.1/` on the command prompt? — alvas, Feb 19 '16 at 15:32
BTW, the second parameter is the trained `.mco` file not another malt-parser, see https://github.com/nltk/nltk/blob/develop/nltk/parse/malt.py#L104 — alvas, Feb 19 '16 at 15:33
Sorry for the multiple comments, please update your question with the above answers to the questions in the comments, and we'll try out best to help you. We need to know these information to better help you with the question. — alvas, Feb 19 '16 at 15:34
My hunch is that this line is causing the problem change the slash to backslash and it should work: https://github.com/nltk/nltk/blob/develop/nltk/parse/malt.py#L69 — alvas, Feb 19 '16 at 15:36
@alvas the second parameter is the path to the mco file it is not another malt-parser. plus I also have the log4j.jar,liblinear-1.8.jar and libvsm.jar in the given path — Mustufain, Feb 20 '16 at 11:20
@alvas output of dir C:/Users/mustufain/Desktop/Python Files/maltparser-1.8.1/ : Invalid switch - "Users" — Mustufain, Feb 20 '16 at 13:07
Ah ha, silly me... In windows, it's `dir C:\Users\mustufain\Desktop\Python Files\maltparser-1.8.1\` What's the output of that? — alvas, Feb 20 '16 at 13:20
@alvas I am not getting an exception now I figured it out as I have changed the name of maltparser-1.8.1.jar to malt.jar file but as I looked through the code of line which was giving me this assertion error I renamed it to maltparser-1.8.1.jar and It went fine — Mustufain, Feb 20 '16 at 14:10
Please look at http://www.howtogeek.com/181774/why-windows-uses-backslashes-and-everything-else-uses-forward-slashes/ — alvas, Feb 22 '16 at 14:18
import nltk from nltk.parse import malt parser = nltk.parse.malt.MaltParser(working_dir="D:/Python Files/maltparser-1.7.2",mco="engmalt.poly-1.7.mco",additional_java_args=['-Xmx512m']) txt = "This is a test sentence" graph=parser.raw_parse(txt) — Mustufain, Feb 22 '16 at 14:20
@alvas I have updated my malt.py file and when I run the above code it gives me this exception : File "C:\Python34\lib\site-packages\nltk\parse\dependencygraph.py", line 378 "The graph doesn't contain a node " UserWarning: The graph doesn't contain a node that depends on the root element. — Mustufain, Feb 22 '16 at 14:21

score 2 · Answer 1 · edited Feb 23 '16 at 21:24

2

TL;DR (In PYTHON3!!):

import urllib.request
urllib.request.urlretrieve('http://www.maltparser.org/mco/english_parser/engmalt.poly-1.7.mco', 'C:\\Users\\mustufain\\Desktop\\engmalt.poly-1.7.mco')
urllib.request.urlretrieve('http://maltparser.org/dist/maltparser-1.8.1.zip', 'C:\\Users\\mustufain\\Desktop\\maltparser-1.8.1.zip')
zfile = zipfile.ZipFile('C:\\Users\\mustufain\\Desktop\\maltparser-1.8.1.zip')
zfile.extractall('C:\\Users\\mustufain\\Desktop\\maltparser-1.8.1\\')

Then:

from nltk.parse import malt
mp = malt.MaltParser('C:\\Users\\mustufain\\Desktop\\maltparser-1.8.1\\', "C:\\Users\\mustufain\\Desktop\\engmalt.poly-1.7.mco")
mp.parse_one('I shot an elephant in my pajamas .'.split()).tree()

edited Feb 23 '16 at 21:24

L3viathan

26,748
2
58
81

answered Feb 22 '16 at 14:24

alvas

115,346
109
446
738

Thanks @L3viathan for the edit! There's an extensive answer on: https://github.com/nltk/nltk/issues/1294 too. – alvas Feb 23 '16 at 21:26
Maybe you should link that in your answer as well then. Greet Saarbrücken from me! – L3viathan Feb 23 '16 at 21:28
@L3viathan No problem ;P – alvas Feb 23 '16 at 21:29

alvas · Answer 2 · 2016-02-22T14:20:38.607

If all the download and environmental variable setup is done correctly, most probably it's how file/dir path are splitted in the nltk.parse.malt.py, at https://github.com/nltk/nltk/blob/develop/nltk/parse/malt.py#L69 that splits the directory and filename specifically for linux:

def find_maltparser(parser_dirname):
    """
    A module to find MaltParser .jar file and its dependencies.
    """
    if os.path.exists(parser_dirname): # If a full path is given.
        _malt_dir = parser_dirname
    else: # Try to find path to maltparser directory in environment variables.
        _malt_dir = find_dir(parser_dirname, env_vars=('MALT_PARSER',))
    # Checks that that the found directory contains all the necessary .jar
    malt_dependencies = ['','','']
    _malt_jars = set(find_jars_within_path(_malt_dir))
    _jars = set(jar.rpartition('/')[2] for jar in _malt_jars)
    malt_dependencies = set(['log4j.jar', 'libsvm.jar', 'liblinear-1.8.jar'])

    assert malt_dependencies.issubset(_jars)
    assert any(filter(lambda i: i.startswith('maltparser-') and i.endswith('.jar'), _jars))
    return list(_malt_jars)

The bug has been fixed and in the process of merging at https://github.com/nltk/nltk/pull/1292

Changing this line:

_jars = set(jar.rpartition('/')[2] for jar in _malt_jars)

to this should solve your problem =)

_jars = set(os.path.split(jar)[1] for jar in _malt_jars)

For the answer not related to the code itself but how you have setup the environment variables or downloaded and saved the malt parser directories or files, see https://github.com/nltk/nltk/issues/1294

It didnot work I have changed the line in malt.py and restart it, it still gives me an assertion erorr when i load the malt-parser — Mustufain, Feb 20 '16 at 11:22
It gives me this new assertion error after changing the line in malt.py: assert any(filter(lambda i: i.startswith('maltparser-') and i.endswith('.jar'), _jars)) AssertionError — Mustufain, Feb 20 '16 at 11:27
It is throwing exception at this line : malt_dependencies = set(['log4j.jar', 'libsvm.jar', 'liblinear-1.8.jar']) — Mustufain, Feb 20 '16 at 12:00

malt parser gives assertion error when using it with nltk

2 Answers2

Linked

Related