Preserve newlines in toot corpus

The original code was already trying to do this, but in a way that
Beautiful Soup ended up stripping out. This way preserves the newlines
properly, which will prevent the bot from smooshing together words
accidentally.

functions.py

@@ -80,10 +80,10 @@ def extract_toot(toot):
 	toot = html.unescape(toot)  # convert HTML escape codes to text
 	soup = BeautifulSoup(toot, "html.parser")
 	for lb in soup.select("br"):  # replace <br> with linebreak
-		lb.name = "\n"
+		lb.append('\n')
 
 	for p in soup.select("p"):  # ditto for <p>
-		p.name = "\n"
+		p.append('\n\n')
 
 	for ht in soup.select("a.hashtag"):  # convert hashtags from links to text
 		ht.unwrap()