def test_path_to_file_uri(self):
if os.name == 'nt':
self.assertEqual(path_to_file_uri("C:\\windows\clock.avi"),
"file:///C:/windows/clock.avi")
else:
self.assertEqual(path_to_file_uri("/some/path.txt"),
"file:///some/path.txt")
fn = "test.txt"
x = path_to_file_uri(fn)
self.assert_(x.startswith('file:///'))
self.assertEqual(file_uri_to_path(x).lower(), os.path.abspath(fn).lower())
def parse(self, response):
base_url = get_base_url(response)
hxs = HtmlXPathSelector(response)
products_list = []
if file_uri_to_path(response.url) == '/documents/specials/':
products_list.extend(
hxs.select('//a[@class="specialMore"]'
'/@href').extract())
category = u'Specials'
else:
products_list.extend(
hxs.select('//ul[@class="subdepartment_list"]'
'/li/a/@href').extract())
category = file_uri_to_path(response.url)\
.split('/')[2].replace('+', ' ')
for url in products_list:
yield Request(urljoin_rfc(base_url, url),
callback=self.parse_products,
meta={'category': category})
def test_file_uri_to_path(self):
if os.name == 'nt':
self.assertEqual(file_uri_to_path("file:///C:/windows/clock.avi"),
"C:\\windows\clock.avi")
uri = "file:///C:/windows/clock.avi"
uri2 = path_to_file_uri(file_uri_to_path(uri))
self.assertEqual(uri, uri2)
else:
self.assertEqual(file_uri_to_path("file:///path/to/test.txt"),
"/path/to/test.txt")
self.assertEqual(file_uri_to_path("/path/to/test.txt"),
"/path/to/test.txt")
uri = "file:///path/to/test.txt"
uri2 = path_to_file_uri(file_uri_to_path(uri))
self.assertEqual(uri, uri2)
self.assertEqual(file_uri_to_path("test.txt"),
"test.txt")
def parse_product(self, response):
hxs = HtmlXPathSelector(response)
options = response.xpath('//*[contains(@class, "config-product-option")]'
'//option[contains(@class, "single-option")]/@value').extract()
if options:
for option in options:
param = '?conf=%(option)s' % ({'option': option})
yield Request(urljoin_rfc(response.url,
file_uri_to_path(response.url) + param),
callback=self.parse_product)
dinamic_name = join(hxs.select('//*[@class="dynamic-name"]/text()')
.extract())
stock = hxs.select('//div[@id="product_type_data"]/div[@class="availability"]/span[contains(@class, "spr-icon-dispo-big-10") or contains(@class, "spr-icon-dispo-big-50")]')
categories = hxs.select('//div[@class="breadcrumbs"]//a/text()')[1:].extract()
if dinamic_name:
name = dinamic_name
else:
name = join(hxs.select('//li[@class="home"]//strong/text()')
.extract())
if not name:
name = join(hxs.select('//div[@id="product_description"]'
'//div[@class="etiquette-title"]/span'
'/span/text()').extract())
if not name:
name = join(hxs.select('//li[@class="home"]'
'//*[@class="store-main-color"]/text()')
.extract())
if not name:
name = join(hxs.select('//div[@class="product-name"]//h1/text()').extract())
sku = join(hxs.select('//div[@class="sku-product"]/span[@class="sku"]/text()')
.extract()).strip()
if not 'outilmania.fr' in response.url:
price_url = hxs.select('//div[@class="price-box"]'
'//img/@src').extract()
if len(price_url) > 1:
price_url = hxs.select('//div[@class="price-box"]'
'//*[contains(@class, "special-price") '
'or contains(@class, "normal-price")]'
'//img/@src').extract()
price_no_vat = response.xpath('//script/text()').re('"productPrice":"(.+?)"')
if price_no_vat:
price = (Decimal(price_no_vat[0]) * Decimal('1.2')).quantize(Decimal('0.01'))
else:
if price_url:
params = {
'url': price_url[0],
'resize': '200',
'mode': '',
'blur':'1',
'format':'float'}
# doing OCR decoding in 3 different modes to increase accuracy
prices = []
price = '0.00'
attempt = 0
while len(prices) == 0 and attempt < self.max_ocr_retries:
attempt += 1
self.log('OCR: {}, attempt {}'.format(price_url[0], attempt))
for mode in ('6', '7', '8'):
params_copy = params.copy()
params_copy['mode'] = mode
params_copy_encoded = urllib.urlencode(params_copy)
ocr_service_url = "http://148.251.79.44/ocr/get_price_from_image?%s" % params_copy_encoded
self.log('>>> GET PRICE => %s' % ocr_service_url)
f = urllib.urlopen(ocr_service_url)
jdata = json.loads(f.read())
self.log(str(jdata))
if len(jdata['price']) > 0:
prices.append(jdata['price'])
try:
price = self._select_price(prices)
except:
# self.errors.append("Price error, posibly ocr error on " + response.url)
pass
price = price.encode('utf-8')
price = price.replace(" ", "").replace(",", ".")
log.msg(str(price), log.DEBUG)
else:
return
else:
faction = hxs.select('//form[@id="product_addtocart_form"]/@action').extract()[0]
image_url = hxs.select('//div[@class="product-image-gallery"]/img[@id="image-main"]/@src').extract()
yield Request(faction,
meta={'name': name,
'url': response.url,
'sku': sku,
'identifier': sku,
'image_url': image_url,
'categories': categories},
callback=self.parse_outilmania_price,
dont_filter=True)
return
if sku and price:
loader = ProductLoader(response=response, item=Product())
loader.add_value('name', name)
loader.add_value('url', response.url)
loader.add_value('price', price)
loader.add_value('sku', sku)
loader.add_value('identifier', sku)
loader.add_xpath('image_url', '//div[@class="product-image-gallery"]/img[@id="image-main"]/@src')
loader.add_value('stock', 1)
for category in categories:
loader.add_value('category', category)
yield loader.load_item()
def __init__(self, uri):
self.path = file_uri_to_path(uri)
def download_request(self, request, spider):
filepath = file_uri_to_path(request.url)
body = open(filepath, 'rb').read()
respcls = responsetypes.from_args(filename=filepath, body=body)
return respcls(url=request.url, body=body)
