#!/usr/bin/env python
# -*- coding: utf-8 -*-
#
# Quick Hack to convert a Kanji list (from STDIN) to Heisig
# "Remembering the Kanji" numbers
#

# Modules
import urllib2
import sys
import re
import time
import subprocess

out = open("kanji_rtk_fetch.py-output", "w")

for kanji in sys.stdin.readlines():
    kanji = kanji.strip()   # No newlines
    print "Fetching " + kanji + " ..."
    url = "http://jisho.org/kanji/details/" + kanji +  "/"
    print "   " + url
    time.sleep(3)
# Does not get all Kanji ... wget does the job better
#    f = urllib2.urlopen(url)
#    data = f.read()
    child = subprocess.Popen(["wget", "-O", "-", url], stdout=subprocess.PIPE)
    data = ""
    for line in child.stdout:
      data = data + line
    child.wait()
    # Veeery dump parsing, maybe it helps :-)
    # Look for
    # <td class="dic_ref">1568</td>\
    # <span title="James Heisig">Remembering The Kanji</span>
    # print data
    kanjinum=-1
    for rx in re.finditer('<td class="dic_ref">(\d+)</td>',data):
        # print "   FIND " + rx.group(1)
        if data[rx.end():rx.end()+69] == '<td class="dic_name"><span title="James Heisig">Remembering The Kanji':
            print "Match found"
            # We found our match
            kanjinum=rx.group(1)
            break
    # Write the result
    if kanjinum is -1:
        print "   Error fetching " + kanji + "\n"
	out.write("Error fetchung " + kanji + "\n")
    else:
	print "   " + kanjinum + " " + kanji + "\n"
	out.write(kanjinum + " " + kanji + "\n")

print "Done"
print "Output in kanji_rtk_fetch.py-output"