Task_Extraction

#The coding was done in jupyter notebook installed along the Anaconda Distributions, hence, many libraries are already present

#importing necessary libraries
import nltk
from nltk import pos_tag, word_tokenize
from nltk.corpus import stopwords
from nltk.tag import RegexpTagger
import re
nltk.download('punkt')
nltk.download('averaged_perceptron_tagger')

#Making a Custom Tagger based on NLTK
tag1 = RegexpTagger([              
(r'.*today$', 'CD'),
(r'.*tonight$', 'CD'),
(r'.*tomorrow$', 'CD'),
(r'.*Monday$', 'CD'),
(r'.*Tuesday$', 'CD'),
(r'.*Wednesday$', 'CD'),
(r'.*Thursday$', 'CD'),
(r'.*Friday$', 'CD'),
(r'.*Saturday$', 'CD'),
(r'.*Sunday$', 'CD'),
(r'.*morning$', 'CD'),
(r'.*afternoon$', 'CD'),
(r'.*evening$', 'CD'),
(r'.*night$', 'CD'),
(r'.*monday$', 'CD'),
(r'.*tuesday$', 'CD'),
(r'.*wednesday$', 'CD'),
(r'.*thursday$', 'CD'),
(r'.*friday$', 'CD'),
(r'.*saturday$', 'CD'),
(r'.*sunday$', 'CD'),
(r'.*ll$','MD')
])
def tg(sent):
    ne=list()
    sent=nltk.word_tokenize(sent)
    for i in sent:
        if tag1.tag(nltk.word_tokenize(i))[0][1]==None:
            ne.append(tuple([nltk.pos_tag(nltk.word_tokenize(i))[0][0],nltk.pos_tag(nltk.word_tokenize(i))[0][1]]))
        else:
            ne.append(tuple([tag1.tag(nltk.word_tokenize(i))[0][0],tag1.tag(nltk.word_tokenize(i))[0][1]]))
    return ne
   
#Function to chunk out the part of a sentence satisfying the defined pattern
def extract_NN(sent):
    grammar = """
    NP:{<.*>*<MD|PDT|VB|VBG|NN.*|PRP.*><.*>*<IN><.*>*<CD><.*>*}
       {<IN><.*>*<CD><.*>*<MD|PDT|VB|VBG|NN.*|PRP.*><.*>*}
    """
    chunker = nltk.RegexpParser(grammar)
    ne = set()
    chunk = chunker.parse(tg(sent))
    for tree in chunk.subtrees(filter=lambda t: t.label() == 'NP'):
        ne.add(' '.join([child[0] for child in tree.leaves()]))
    if len(ne)>0:
        print(ne)
    return ne
 
#Getting Input
text="""
Hey driver, please be early tomorrow, I have my flight at 6 in the morning. 
"""

#Dividing the text into sentences
sentences = re.split(r' *[\.\?!,][\'"\)\]]* *', text)

#Printing all the tasks from all the sentences
for stuff in sentences:
    extract_NN(stuff)