Source code for nltk.tag.hunpos

# -*- coding: utf-8 -*-
# Natural Language Toolkit: Interface to the HunPos POS-tagger
# Copyright (C) 2001-2017 NLTK Project
# Author: Peter Ljunglöf <>
#         Dávid Márk Nemeskey <> (modifications)
#         Attila Zséder <> (modifications)
# URL: <>
# For license information, see LICENSE.TXT

A module for interfacing with the HunPos open-source POS-tagger.

import os
from subprocess import Popen, PIPE

from nltk.internals import find_binary, find_file
from nltk.tag.api import TaggerI
from nltk import compat

_hunpos_url = ''

_hunpos_charset = 'ISO-8859-1'
"""The default encoding used by hunpos: ISO-8859-1."""

[docs]class HunposTagger(TaggerI): """ A class for pos tagging with HunPos. The input is the paths to: - a model trained on training data - (optionally) the path to the hunpos-tag binary - (optionally) the encoding of the training data (default: ISO-8859-1) Example: >>> from nltk.tag import HunposTagger >>> ht = HunposTagger('en_wsj.model') >>> ht.tag('What is the airspeed of an unladen swallow ?'.split()) [('What', 'WP'), ('is', 'VBZ'), ('the', 'DT'), ('airspeed', 'NN'), ('of', 'IN'), ('an', 'DT'), ('unladen', 'NN'), ('swallow', 'VB'), ('?', '.')] >>> ht.close() This class communicates with the hunpos-tag binary via pipes. When the tagger object is no longer needed, the close() method should be called to free system resources. The class supports the context manager interface; if used in a with statement, the close() method is invoked automatically: >>> with HunposTagger('en_wsj.model') as ht: ... ht.tag('What is the airspeed of an unladen swallow ?'.split()) ... [('What', 'WP'), ('is', 'VBZ'), ('the', 'DT'), ('airspeed', 'NN'), ('of', 'IN'), ('an', 'DT'), ('unladen', 'NN'), ('swallow', 'VB'), ('?', '.')] """ def __init__(self, path_to_model, path_to_bin=None, encoding=_hunpos_charset, verbose=False): """ Starts the hunpos-tag executable and establishes a connection with it. :param path_to_model: The model file. :param path_to_bin: The hunpos-tag binary. :param encoding: The encoding used by the model. Unicode tokens passed to the tag() and tag_sents() methods are converted to this charset when they are sent to hunpos-tag. The default is ISO-8859-1 (Latin-1). This parameter is ignored for str tokens, which are sent as-is. The caller must ensure that tokens are encoded in the right charset. """ self._closed = True hunpos_paths = ['.', '/usr/bin', '/usr/local/bin', '/opt/local/bin', '/Applications/bin', '~/bin', '~/Applications/bin'] hunpos_paths = list(map(os.path.expanduser, hunpos_paths)) self._hunpos_bin = find_binary( 'hunpos-tag', path_to_bin, env_vars=('HUNPOS_TAGGER',), searchpath=hunpos_paths, url=_hunpos_url, verbose=verbose ) self._hunpos_model = find_file( path_to_model, env_vars=('HUNPOS_TAGGER',), verbose=verbose) self._encoding = encoding self._hunpos = Popen([self._hunpos_bin, self._hunpos_model], shell=False, stdin=PIPE, stdout=PIPE, stderr=PIPE) self._closed = False def __del__(self): self.close()
[docs] def close(self): """Closes the pipe to the hunpos executable.""" if not self._closed: self._hunpos.communicate() self._closed = True
def __enter__(self): return self def __exit__(self, exc_type, exc_value, traceback): self.close()
[docs] def tag(self, tokens): """Tags a single sentence: a list of words. The tokens should not contain any newline characters. """ for token in tokens: assert "\n" not in token, "Tokens should not contain newlines" if isinstance(token, compat.text_type): token = token.encode(self._encoding) self._hunpos.stdin.write(token + b"\n") # We write a final empty line to tell hunpos that the sentence is finished: self._hunpos.stdin.write(b"\n") self._hunpos.stdin.flush() tagged_tokens = [] for token in tokens: tagged = self._hunpos.stdout.readline().strip().split(b"\t") tag = (tagged[1] if len(tagged) > 1 else None) tagged_tokens.append((token, tag)) # We have to read (and dismiss) the final empty line: self._hunpos.stdout.readline() return tagged_tokens
# skip doctests if Hunpos tagger is not installed
[docs]def setup_module(module): from nose import SkipTest try: HunposTagger('en_wsj.model') except LookupError: raise SkipTest("HunposTagger is not available")