Source Code for Module nltk.classify.mallet

 1  # Natural Language Toolkit: Interface to Mallet Machine Learning Package 
 2  # 
 3  # Copyright (C) 2008 NLTK Project 
 4  # Author: Edward Loper <[email protected]> 
 5  # URL: <http://nltk.org> 
 6  # For license information, see LICENSE.TXT 
 7  # 
 8  # $Id: naivebayes.py 2063 2004-07-17 21:02:24Z edloper $ 
 9   
10  """ 
11  A set of functions used to interface with the external U{Mallet 
12  <http://mallet.cs.umass.edu/>} machine learning package.  Before 
13  C{mallet} can be used, you should tell NLTK where it can find the 
14  C{mallet} package, using the L{config_mallet()} function.  Typical 
15  usage: 
16   
17      >>> import nltk 
18      >>> nltk.config_mallet('.../path/to/mallet') 
19  """ 
20  __docformat__ = 'epytext en' 
21   
22  import os 
23  import os.path 
24   
25  from nltk.internals import find_binary, java 
26   
27  ###################################################################### 
28  #{ Configuration 
29  ###################################################################### 
30   
31  _mallet_home = None 
32  _mallet_classpath = None 
33 -def config_mallet(mallet_home=None): 
34      """ 
35      Configure NLTK's interface to the C{mallet} machine learning 
36      package. 
37   
38      @param mallet_home: The full path to the C{mallet} directory.  If 
39          not specified, then nltk will search the system for a 
40          C{mallet} directory; and if one is not found, it will raise a 
41          C{LookupError} exception. 
42      @type mallet_home: C{string} 
43      """ 
44      global _mallet_home, _mallet_classpath 
45       
46      # We don't actually care about this binary -- we just use it to 
47      # make sure we've found the right directory. 
48      mallethon_bin = find_binary( 
49          'mallet', mallet_home, 
50          env_vars=['MALLET',  'MALLET_HOME'], 
51          binary_names=['mallethon'], 
52          url='http://mallet.cs.umass.edu>') 
53      # Record the location where mallet lives. 
54      bin_dir = os.path.split(mallethon_bin)[0] 
55      _mallet_home = os.path.split(bin_dir)[0] 
56      # Construct a classpath for using mallet. 
57      lib_dir = os.path.join(_mallet_home, 'lib') 
58      if not os.path.isdir(lib_dir): 
59          raise ValueError('While configuring mallet: directory %r ' 
60                           'not found.' % lib_dir) 
61      _mallet_classpath = ':'.join([os.path.join(lib_dir, filename) 
62                                    for filename in sorted(os.listdir(lib_dir)) 
63                                    if filename.endswith('.jar')]) 
64   
65       
66 -def call_mallet(cmd, classpath=None, stdin=None, stdout=None, stderr=None, 
67                  blocking=True): 
68      """ 
69      Call L{nltk.internals.java()} with the given command, and with the 
70      classpath modified to include both C{nltk.jar} and all the C{.jar} 
71      files defined by Mallet. 
72   
73      See L{nltk.internals.java()} for parameter and return value 
74      descriptions. 
75      """ 
76      if _mallet_classpath is None: 
77          config_mallet() 
78       
79      # Set up the classpath 
80      if classpath is None: 
81          classpath = _mallet_classpath 
82      else: 
83          classpath += ':' + _mallet_classpath 
84      # Delegate to java() 
85      return java(cmd, classpath, stdin, stdout, stderr, blocking) 
86