#!/usr/bin/env python
# A programmatic interface to the Oxford English Dictionary's online edition.
# Adam Sampson <ats@offog.org>

import os, re, HTMLParser, sha
from urlparse import urljoin

base_url = "http://80-dictionary.oed.com.chain.kent.ac.uk"
cookie_jar = "/tmp/oed.cookies"

class TagFinder(HTMLParser.HTMLParser):
	def __init__(self, lookfor):
		HTMLParser.HTMLParser.__init__(self)
		self.lookfor = lookfor
		self.found = []

	def handle_starttag(self, tag, attrs):
		if tag not in self.lookfor:
			return
		self.found.append((tag, dict(attrs)))

def get_page(url):
	cache_name = "/tmp/oed.cache." + sha.new(url).hexdigest()
	try:
		f = open(cache_name)
		#print "Reading from cache " + url
		data = f.read()
		f.close()
		return data
	except IOError:
		pass

	#print "Fetching " + url
	f = os.popen("wget -q -O - --load-cookies=" + cookie_jar + " --save-cookies=" + cookie_jar + " '" + url + "'")
	data = f.read()
	f.close()

	f = open(cache_name, "w")
	f.write(data)
	f.close()

	return data

def get_frames(url):
	data = get_page(url)
	tf = TagFinder(["frame"])
	tf.feed(data)
	frames = {}
	for (tag, attrs) in tf.found:
		frames[attrs["name"]] = urljoin(url, attrs["src"])
	return frames

def strip_tags(s):
	s = re.sub(r'<[^>]*>', "", s)
	s = re.sub(r'\s+', " ", s)
	def fix(m):
		s = m.group(1)
		if s == "amp":
			return "&"
		elif s == "lt":
			return "<"
		elif s == "gt":
			return ">"
		elif s[0] == "#":
			if s[1] == "x":
				return chr(int(s[2:], 16))
			else:
				return chr(int(s[1:]))
	s = re.sub(r'&([^;]+);', fix, s)
	return s.strip()

def parse_entry(maindata, hdrdata):
	m = re.search(r'(?ms)<FONT SIZE="\+2">(.*?)</FONT>', hdrdata)
	word = strip_tags(m.group(1))

	ms = re.findall(r'(?ms)<!--start_def-->(.*?)<!--end_def-->', maindata)
	definition = strip_tags(" ".join(ms))

	return (word, definition)

def search(word):
	fseturl = base_url + "/cgi/findword?query_type=word&queryword=" + word + "&find=find"
	frames = get_frames(fseturl)
	if frames["Main frame"].find("nearest_to") != -1:
		# No results.
		return []
	maindata = get_page(frames["Main frame"])

	results = []
	if maindata.find("<TITLE>OED Online main entry text frame") != -1:
		# Only one result -- this page.
		hdrdata = get_page(frames["Header frame"])
		results.append(parse_entry(maindata, hdrdata))
	else:
		# Multiple results.
		for m in re.findall(r'<A HREF="(/cgi/entry[^"]*)">', maindata):
			rseturl = urljoin(frames["Main frame"], m)
			rframes = get_frames(rseturl)

			rmaindata = get_page(rframes["Main frame"])
			rhdrdata = get_page(rframes["Header frame"])
			results.append(parse_entry(rmaindata, rhdrdata))

	return results

def wordwrap(text, width, prefix = ""):
	text += " "
	n = 0
	while n < len(text):
		p = -1
		w = width
		while p == -1:
			p = text.rfind(" ", n, w + n)
			w += width
		print prefix + text[n:p]
		n = p + 1

if __name__ == "__main__":
	import sys
	if len(sys.argv) == 2:
            want = sys.argv[1]
        else:
            want = "hello"
	for word, definition in search(want):
		print word
		wordwrap(definition, 70, "  ")