def __init__(self, opts, *args, **qdict):
BaseHTTPServer.HTTPServer.__init__(self, *args, **qdict)
import nltk
_SENTENCE_TOKENIZE_MODEL = "tokenizers/punkt/english.pickle"
self.tokenizer = nltk.data.load(_SENTENCE_TOKENIZE_MODEL)
self.speller = aspell.Speller('lang', 'en')
try:
sennabin = unicode(os.environ['SENNAPATH'])
except KeyError:
sennabin = u"/data/tool/senna/"
self.senna = src.tools.senna.SennaWrap(sennabin)
self.funcs = {'split':self.split, 'spell':self.spell, 'pas': self.pas, "score" : self.score}
M_PATH = opts.model_dir
self.model = SklearnClassifier()
self.model.load_model(M_PATH)
self.model.load_fmap(M_PATH)
