#!/usr/bin/python
# -*- coding: utf-8 -*-

import sys
import re

if len(sys.argv)<2:
	print("Usage convert_cnec_stanford.py [corpus_in_cnec_xml]")
	exit(0)

filename = sys.argv[1]

ENT = {"P":"PER","pp":"PER","p_":"PER","pf":"PER","ps":"PER","pb":"PER","pm":"PER","pc":"PER","pd":"PER",
       "ia":"ORG","if":"ORG","io":"ORG","ic":"ORG","i_":"ORG",
       "G":"LOC","gc":"LOC","gh":"LOC","gr":"LOC","gq":"LOC","gl":"LOC","gu":"LOC","gt":"LOC","gs":"LOC","g_":"LOC",
	   "A":"LOC","a_":"LOC","ah":"LOC","az":"LOC", "lower": "O", "cap": "O"}

ne_type_re = re.compile(r'<ne type="([^"]*)">', re.U)
markup_split_re = re.compile(r'(<[^>]*>)|(\s*)', re.U)

with open(filename) as f:
	for line in f:
		line = line.strip()
		new_line = u""
		inside = 0
		markup = "O"
		for k, token in enumerate(markup_split_re.split(line)):
			if not token or not token.strip():
				continue
			#print("token",token,inside,markup)
			ne_type = ne_type_re.findall(token)
			if ne_type:
				inside += 1
				if inside== 1:
					if ne_type[0] in ENT.keys():
						markup = ENT[ne_type[0]]
					else:
						markup = "OTHER"
			elif token == "</ne>":
				inside -= 1
				if inside == 0:
					markup = "O"				
			elif not token.startswith("<") or not token.endswith(">"):
				new_line += u"{}\t{}\n".format(token, markup)
		print(new_line)