markov startwords

Chatbot.py

@@ -2,6 +2,7 @@ import asyncio
 import discord
 from discord.ext import commands
 import numpy as np
+import random
 
 class Chatbot(object):
 	"""Een heuse chatbot. Under construction..."""
@@ -10,11 +11,12 @@ class Chatbot(object):
 		self.bot = bot
 		self.voice_states = {}
 		self.END_OF_MESSAGE = "<EOM>"
+		self.START_OF_MESSAGE = "<SOM>"
 		try:
 			np.load('markov_dict.npy')
 		except FileNotFoundError:
 			#If there's no initial file, make a new one:
-			initFile = {'pairs':{'Hoi':{self.END_OF_MESSAGE : 1}}, 'trairs':{}}
+			initFile = {'pairs':{self.START_OF_MESSAGE:{'Hoi':1},'Hoi':{self.END_OF_MESSAGE:1}}, 'trairs':{}}
 			np.save("markov_dict.npy", initFile)
 		except:
 			print("ERROR! - Could not open or create Markov chat file")
@@ -41,11 +43,10 @@ class Chatbot(object):
 		newPairDict = wordDict['pairs']
 		newTrairDict = wordDict['trairs']
 		
-		message = message + " " + self.END_OF_MESSAGE
+		message = self.START_OF_MESSAGE + " " + message + " " + self.END_OF_MESSAGE
 		corpus = message.split()
 		pairs = self.makePairs(corpus)
 		trairs = self.makeTrairs(corpus)
-		print(trairs)
 
 		#Add wordcounts to pair dict
 		for word_1, word_2 in pairs:
@@ -84,7 +85,8 @@ class Chatbot(object):
 		iterator = channel.history(limit=nrOfMessages, before=None, after=None, reverse=False, around=None)
 		for i in range(nrOfMessages):
 			msg = yield from iterator.next()
-			messageContent = msg.content.strip()
+			messageContent = msg.content.replace('(', '').replace(')', '').replace('"', '')
+			messageContent = messageContent.strip()
 			if messageContent == "":
 				pass #print("-- Bericht is leeg. Negeer.")
 			elif msg.author.bot:
@@ -124,11 +126,17 @@ class Chatbot(object):
 		#print(wordDict)
 		
 		#Pick a random first word (Which is not an end-of-message)
-		#TODO pick a word which responds to the previous message in the channel
-		first_word = np.random.choice(list(wordPairDict.keys()))
-		while first_word == self.END_OF_MESSAGE: #or first_word.islower() ???
+		#Usually, we start with a regular <SOM>, but sometimes we don't
+		a = random.randint(1,10)
+		if a == 1:
 			first_word = np.random.choice(list(wordPairDict.keys()))
+			while first_word == self.END_OF_MESSAGE:
+				first_word = np.random.choice(list(wordPairDict.keys()))
+		else:
+			first_word = self.START_OF_MESSAGE
 		chain = [first_word]
+		#TODO pick a word which responds to the previous message in the channel
+		
 
 		#Go through the chain:
 		for i in range(1, MAX_N_WORDS):
@@ -151,7 +159,10 @@ class Chatbot(object):
 			probs = weights / weights.sum()
 			chain.append(np.random.choice(possible_words, p=probs))
 
-		chain = chain[:-1] #remove <EOM>
+		if chain[0] == self.START_OF_MESSAGE:
+			chain = chain[1:-1] #remove <SOM> and <EOM>
+		else:
+			chain = chain[:-1] #remove <EOM>
 			
 		#print(' '.join(chain))
 		yield from ctx.channel.send(' '.join(chain))