'''
Created on Jul 3, 2017

@author: David
'''

from os.path import isfile,join,splitext
from os import listdir
import codecs

topic_dir = "siwoco/coctaill_topics_lemmatized"
topic_files = [join(topic_dir,f) for f in listdir(topic_dir) if isfile(join(topic_dir,f)) and f.endswith(".list")]

swefn_dir = "siwoco/topic_lists"
swefn_files = [join(swefn_dir,f) for f in listdir(swefn_dir) if isfile(join(swefn_dir,f)) and f.endswith(".syn.list")]

topics = []

topics_fn = []

for topic_file in topic_files:
    arr = []
    with codecs.open(topic_file,"r",encoding="utf-8") as f:
        for l in f:
            arr.append(l.rstrip())
    topics.append(arr)

for swefn_file in swefn_files:
    arr = []
    with codecs.open(swefn_file,"r",encoding="utf-8") as f:
        for l in f:
            arr.append(l.rstrip())
    topics_fn.append(arr)

def get_topics(word):
    topic_distribution = []
    for topic in topics:
        if word in topic:
            topic_distribution.append(1)
        else:
            topic_distribution.append(0)
    return topic_distribution

def get_topics_fn(word):
    topic_distribution = []
    for topic in topics_fn:
        if word in topic:
            topic_distribution.append(1)
        else:
            topic_distribution.append(0)
    return topic_distribution

#print(get_topics_fn("gurka"))
#print(get_topics_fn("tomat"))
#print(get_topics("gurka"))
#print(get_topics("tomat"))