'''
Created on Dec 6, 2016

@author: David
'''
from pattern import padded_ngrams
import codecs
import sys
file = "f:/parll data/data/svalex-ttm.csv"
out = "f:/parll data/data/trigram-padded.dist"

f = codecs.open(file, "r", "utf-8")
o = codecs.open(out, "w", "utf-8")

dist = {}

        
def cefrmap(cefr):
    if (cefr == "A1"):
        return 1
    if (cefr == "A2"):
        return 2
    if (cefr == "B1"):
        return 3
    if (cefr == "B2"):
        return 4
    if (cefr == "C1"):
        return 5
    if (cefr == "C2"):
        return 6
    return 0

for line in f:
    lemma,pos,cefr = line.rstrip().split("\t")
    bigram = padded_ngrams(lemma, 2)
    trigram= padded_ngrams(lemma, 3)
    mcefr = cefrmap(cefr) - 1
    for a,b,c in trigram:
        
        gram = a+b+c
        if (not gram in dist.keys()):
            dist[gram] = [0,0,0,0,0]
        
        dist[gram][mcefr] += 1
    
    #print(bigram)
    #sys.exit()
print(dist)
for key,val in dist.items():
    o.write("{}\t{}\n".format(key, "\t".join(str(v) for v in val)))