# -*- coding: utf-8 -*-
#!/usr/bin/python3.5 -S

import time

#pip install --upgrade 3to2
#pip install --upgrade language-check

start_time = time.time()
import os
os.system('clear')

# http://polyglot.readthedocs.io/en/latest/Installation.html
import NP
#import CalcDocSimilarity
import GensimCalcSimilarity
import polyglot
from polyglot.text import Text, Word
import markovify
from langdetect import detect
import spacy
import os.path
import pprint
import codecs
import re
import numpy
import string
import sys  # import sys package, if not already imported
from textblob_de import TextBlobDE
from unidecode import unidecode

allowedRule			= [u'DET',u'ADJ',u'NOUN',u'VERB',u'ADV',u'ADJ']
####allowedRule				= [u'DET',u'ADJ',u'NOUN']
####allowedRule			= [u'ADJ',u'NOUN']

allowedRuleString		= "".join(allowedRule).replace("PUNCT","")


def createNERBasedSamples(text, markov, ndEntity):
	if (not ndEntity or not markov or not text or len(ndEntity) <= 0):
		return False
		
	#markov += "In Berlin und Umgebung"
	goodNERFlagAny = any(x in markov for x in ndEntity)
	goodNERFlagAll = all(x in markov for x in ndEntity)
	'''
	print("\n -> Good Markov Result (NER-Based):")
	print(goodNERFlagAny)
	print(ndEntity)
	print(markov)
	print("\n\n")
	sys.exit(0)
	'''
	if goodNERFlagAny: 
		simScore 		= GensimCalcSimilarity.CalcDocSimilarity(text, markov)
		syntaxScore 	= GensimCalcSimilarity.CalcWordSimilarity(text, markov, train=False)
		syntaxScore2 	= GensimCalcSimilarity.CalcCosineSimilarity(text, markov)
		print("###############################"), 
		print(" -> Similarity Score (NER-Based):"), 
		print(simScore),
		print("\n -> Good Markov Result (NER-Based):")
		print(markov)
		print("\n\n")
		
		with codecs.open("sample_nerbased.log",'a',encoding='utf-8') as f:
			f.write("\nsimScore: -> ")
			f.write('% 6.2f' % simScore)
			f.write("\n -> ")
			'''
			f.write("\ntriesMarkov: ")
			f.write('% 6.2f' % triesMarkov)
			f.write("\n  ---> max_overlap_ratioMarkov: ")
			f.write('% 6.2f' % max_overlap_ratioMarkov)
			f.write("\n  ---> max_overlap_totalMarkov: ")
			f.write('% 6.2f' % max_overlap_totalMarkov)
			f.write("\n  ---> ####### -> ")
			'''
			f.write(markov)
			f.write("\n")
			f.close()
		#time.sleep(3)
		
		return True
	else:
		return False
		#print(count), 
		#print(" -> No Markov Result (Noun Phrase)")
	return False


def createGrammarBasedSamples(text, markov, nlp):
	if (not nlp or not markov or not text):
		return False
	
	posTagMarkov 	= []
	toks1 			= nlp(markov)

	for token1 in toks1:
		posTagMarkov.append(token1.pos_)

	posTagMarkovString		= "".join(posTagMarkov).replace("PUNCT","")
	
	if allowedRuleString in posTagMarkovString:
	#	print(count)
	#	print("\n")
	#	print("Match: (Rule-Based-Grammar)")
	#	print("Allowed Rule"),
	#	print(allowedRule)
	#	print("Postag Markov:"),
	#	print(posTagMarkov)

		simScore 		= GensimCalcSimilarity.CalcDocSimilarity(text, markov)
		syntaxScore 	= GensimCalcSimilarity.CalcWordSimilarity(text, markov, train=False)
		syntaxScore2 	= GensimCalcSimilarity.CalcCosineSimilarity(text, markov)
		
		#print("###############################"), 
		#print(" -> Similarity Score (Rule-Based-Grammar):"), 
		#print(simScore),
		print("\n -> Good Markov Result (Rule-Based-Grammar):")
		print(markov)
		print("\n\n")

		with codecs.open("sample_rulebased.log",'a',encoding='utf-8') as f:
			f.write("\nsimScore: -> ")
			f.write('% 6.2f' % simScore)
			f.write("\n -> ")
			'''
			f.write("\ntriesMarkov: ")
			f.write('% 6.2f' % triesMarkov)
			f.write("\n  ---> max_overlap_ratioMarkov: ")
			f.write('% 6.2f' % max_overlap_ratioMarkov)
			f.write("\n  ---> max_overlap_totalMarkov: ")
			f.write('% 6.2f' % max_overlap_totalMarkov)
			f.write("\n  ---> ####### -> ")
			'''
			f.write(markov)
			f.write("\n")
			f.close()
		#time.sleep(3)
		
		return True
	else:
		return False
		#print(count), 
		#print(" -> No Markov Result (Rule-Based-Grammar)")
	return False

def createNounPhraseSamples(text, markov, nounPhrases, nlp):
	if (not nlp or not markov or not text or not nounPhrases):
		return False
	
	iNP 	= NP.tokenize(markov, nlp)  
	goodFlag = False
	#flache liste von iNP -> immer nur ein Element, und da stecken dann
	nPList 	= nounPhrases[0]
	tmpList = iNP[0]
	
	goodFlagAny = any(x in tmpList for x in nPList)
	goodFlagAll = all(x in tmpList for x in nPList)
	
	if goodFlagAll:
		#print("Match (Noun Phrase MARKOV):"),
		#print(tmpList)
		#print("Match (Noun Phrase STANDARD):"),
		#print(nPList)
		#print("Match ANY (Noun Phrase):"),
		#print(goodFlagAny),
		#print("Match ALL (Noun Phrase):"),
		#print(goodFlagAll)
		#
		
		simScore 		= GensimCalcSimilarity.CalcDocSimilarity(text, markov)
		syntaxScore 	= GensimCalcSimilarity.CalcWordSimilarity(text, markov, train=False)
		syntaxScore2 	= GensimCalcSimilarity.CalcCosineSimilarity(text, markov)
		
		print("###############################"), 
		print(" -> Similarity Score (Noun-Phrase):"), 
		print(simScore),
		print("\n -> Good Markov Result (Noun-Phrase):")
		print(markov)
		print("\n\n")
		
		with codecs.open("sample_nounphrase.log",'a',encoding='utf-8') as f:
			f.write("\nsimScore: -> ")
			f.write('% 6.2f' % simScore)
			f.write("\n -> ")
			'''
			f.write("\ntriesMarkov: ")
			f.write('% 6.2f' % triesMarkov)
			f.write("\n  ---> max_overlap_ratioMarkov: ")
			f.write('% 6.2f' % max_overlap_ratioMarkov)
			f.write("\n  ---> max_overlap_totalMarkov: ")
			f.write('% 6.2f' % max_overlap_totalMarkov)
			f.write("\n  ---> ####### -> ")
			'''
			f.write(markov)
			f.write("\n")
			f.close()
		#time.sleep(3)
		
		return True
	else:
		1
		#print(count), 
		#print(" -> No Markov Result (Noun Phrase)")
		return False
	return False