from __future__ import with_statement
import re, collections

def words(text):
	#return re.findall("[a-z']+", text.lower())
	return [x for x in re.findall("[a-z']+", text.lower()) if "'" not in x]

def train(features):
    model = collections.defaultdict(lambda : 0.0)
    for f in features:
        model[f] += 1
    return model

def wordfrequency() :
	with open(r"c:\wordfreq.txt") as f :
		lines = [ line.strip() for line in f.readlines() if len(line.strip()) ]
	regex = r"(?P<word>\S+) = (?P<freq>\d+) ?"
	recomp = re.compile(regex)
	d = collections.defaultdict(lambda : 1.0)
	for line in lines :
		for m in recomp.findall(line) :
			word, freq = m[0].lower(), m[1]
			d[word] = float(freq)
	return d

def find_candidate(wf, filename) :
	expected_size = sum(v for v in wf.values())
	with open(filename) as f :
		NWORDS = train(words(f.read()))
	text_size = sum(v for v in NWORDS.values())

	score = dict()
	for w in NWORDS :
		actual_freq, expected_freq = NWORDS[w]/text_size, wf[w]/expected_size
		score[w] = actual_freq / expected_freq
		#print(w, actual_freq, expected_freq, NWORDS[w], wf[w])

	a = sorted(set(score.values()))
	a.reverse()
	google_words = []
	for aa in a[:10] :
		google_words += [k for k,v in score.items() if v == aa]
		if len(google_words) >= 10 : break
	print(google_words)
	print ("http://www.google.com/search?hl=en&q=" + "%2B".join(google_words[:10]))

if __name__ == "__main__" :
	import sys
	wf = wordfrequency()

	for filename in sys.argv[1 :] :
		find_candidate(wf, filename)