Integration Python Script in Knime and proper execution of python function

Hello all.. I encountered an issue where I have a function written in Python which is basiclly a Hebrew Stemmer, and i want to use terms or documents from the "Bag of Words" node, and apply the stemmer on the input features.

I have the functin which calls a local server for the stemming process:

import urllib
import pandas

def lemmatize(text):
    text_encoded = text.encode("utf8")
    params = urllib.urlencode({'text' : text_encoded})#.encode("utf8")})
    while didnt_got_tags:
            f = urllib.urlopen("HTTP://", params)
            words = f.readlines()
            for word in words:
                if 'HTTP ERROR' in word:
                    print 'error'
            didnt_got_tags = False
            print 'error :('
    words = f.readlines()
    lemmas = []
    for word in words:
        if not word == '\n':
            splt = word.split('\t')
            lemmatized = remove_nums(splt[2].split('^')[-1])
            pos = tostring1(int(splt[1])).split(',')[0].split(':')[1].split('-')[0]
            lemmas.append((lemmatized, pos))
    return lemmas

I know that i need to create additional column in the input_table which will contain the stemmed document or term:

input_table['lemmatize'] = lemmatize(input_table['Term']


However, i keep getting errors... anyone can understand why?


could you please provide the errors and the KNIME log?

An example workflow would be helpful, too (if possible).