amazon_get_book_info.py

import sys
import re
import amazonproduct
import pprint
import pickle
import lxml.objectify
import numpy as np
from lxml import etree
from datetime import datetime
import socket
import urllib2


def item_search(api):
	items = api.item_search('Books', Title="Benjamin Franklin", Author = 'Walter Isaacson')
	# print dir(items)
	print len(items)
	for book in items:
		try:
			print '%s: "%s"' % (book.ItemAttributes.Author,
	                        	book.ItemAttributes.Title)
			pprint.pprint(dir(book.ItemAttributes))
		except AttributeError:
			pass

def get_page_numbers(api, asin):
	# result = api.item_lookup('B006H3MIV8', ResponseGroup = 'Tracks')
	try:
		result = api.item_lookup(asin, ResponseGroup = 'Large')
	except (socket.timeout, urllib2.URLError):
		return None
	# print str(result)
	for book in result:
		result_string = etree.tostring(book, pretty_print = True)
		for line in result_string.split('\n'):
			if 'NumberOfPages' in line:
				match = re.search('(\d+)', line)
				return int(match.group(0))


def get_all_page_numbers(res_list, api):
	for book in res_list:
		book['page_numbers_list'] = []
		for asin in book['asin_list']:
			number = get_page_numbers(api, asin)
			if number != None:
				book['page_numbers_list'].append(number)
		print book['page_numbers_list']
		book['median_page_numbers'] = np.median(np.array(book['page_numbers_list']))
		print book['median_page_numbers']

def parse_book_orders(inhandle):
	res_list = []
	with open(inhandle) as fi:
		lines = fi.readlines()
		lines = [x.strip() for x in lines]
		date_re = re.compile('([0-9]+ [A-Z][a-z]+ [0-9]+)')
		for x, line in enumerate(lines):
			if line.startswith('Order details'):
				res_dict = {}
				this_date = date_re.search(line)
				# print line
				# print this_date.group(0)
				# print datetime.strptime(this_date.group(0), '%d %B %Y')
				res_dict['date'] = datetime.strptime(this_date.group(0), '%d %B %Y')
				# print lines[x + 2]
				res_dict['title'] = lines[x + 1].split(' (')[0]
				# print title
				if len(lines[x + 2].split(', ')) > 1:
					author_surname = lines[x + 2].split(', ')[0]
					author_firstname = lines[x + 2].split(', ')[1]
					author = author_firstname + ' ' + author_surname
				else:
					author = lines[x + 2]
				res_dict['author'] = author
				res_list.append(res_dict)
	return res_list

def get_aisn(res_list, api):
	for book in res_list:
		try:
			results = api.item_search('Books', Title = book['title'], Author = book['author'])
			print book['title'], len(results)
		except UnicodeDecodeError:
			print 'problem with', book['title']
			pass
		book['asin_list'] = []
		for x in results:
			result_string = etree.tostring(x, pretty_print = True)
			for line in result_string.split('\n'):
				if '<ASIN>' in line:
					match = re.search('>([A-Z, 0-9])\w+<', line)
					asin = match.group(0)
					asin = asin.strip('>')
					asin = asin.strip('<')
					book['asin_list'].append(asin)
	return res_list

def print_data_for_r(res_list):
	dates = [x['date'].date() for x in res_list]
	dates = sorted(dates)
	for book in res_list:
		try:
			end_date = dates[dates.index(book['date'].date()) + 1]
			print '\t'.join(map(str, [book['title'], book['median_page_numbers'], book['date'].date(), end_date]))
		except IndexError:
			print '\t'.join(map(str, [book['title'], book['median_page_numbers'], book['date'].date()]))



# item_lookup(api)
def main():
	api = amazonproduct.API()
	res_list = parse_book_orders('/path/to/kindle_books')
	res_list = get_aisn(res_list, api)
	get_all_page_numbers(res_list, api)
	pickle.dump(res_list, open('res_list.pick', 'wb'))
	# res_list = pickle.load(open('res_list.pick', 'r'))
	print_data_for_r(res_list)




if __name__ == '__main__':
	main()