def run(self, response):
page_1_str = self.page_str + "1"
meta_url = response.meta.get("url")
#Basically the website redirects us to the url and page_1_str, which isn't added to our database
# So we trim that off so we can get the url in our database
this_url = trim_url(response.url, page_1_str)
self.logger.info(f"inside run for {this_url}, meta_url: {meta_url}")
if meta_url != this_url:
self.logger.info(
f"meta_url: {meta_url} != response.url: {response.url}, therefore it must be invalid - skipping"
)
this_url = meta_url
else:
self.scrape_urls(response)
finish_url(self.conn, self.store_id, this_url, scrape_urls=True)
self.logger.info("finishing url - " + this_url)
next_url = get_next_url(self.cursor, 1, self.store_id, True)
if next_url is None:
self.logger.info(
"Next url is none therefore we must be finished ! ")
return
else:
next_request = create_unfiltered_parse_request(
next_url, self.run,
EC.element_to_be_clickable(
(By.CSS_SELECTOR, '#openFulfillmentModalButton')))
self.logger.info(f"got next_url - {next_url}")
yield next_request
def collect_menu(self, response):
self.logger.info("inside collect_menu! ")
self.driver = response.request.meta['driver']
close_modal(self)
change_store_location(self)
departments = self.driver.find_elements_by_css_selector(
'[category-filter="subcategory"]')
for department in departments:
dept_name = department.find_element_by_css_selector(
'[data-test="category-card-"]').text
aisles = department.find_elements_by_css_selector('a')
self.logger.info(f"dept_name: {dept_name}")
self.aisles = aisles
for aisle in aisles:
aisle_name = aisle.text
aisle_url = aisle.get_attribute("href")
category = lookup_category("", dept_name, aisle_name)
store_url(self.conn, aisle_url, self.store_id,
category, dept_name, aisle_name)
#inspect_response(response, self)
self.logger.info("finished collect_menu! ")
finish_url(self.conn, self.store_id, response.url, scrape_urls=True)
request = self.get_next_request()
yield request
def no_pagination(self, failure):
url = failure.request.url
self.logger.info(f"no_pagination for url: {url}, continuing")
finish_url(self.conn, self.store_id, url, set_val=-1, scrape_urls=True)
# TODO add a filter so we don't get the ones with ?page=
request = self.get_next_request()
yield request
def parse(self, response):
# This callback determines if the selected menu is
# at the top of the list, if it is then it adds the urls
# to the list and keeps going
# if its not, then it calls the lua to prepare the page
# for scraping, and then scrapes it
url = response.url
menu = response.css(".category-filter__link")
#submenu = response.css("")
#print ("self.urls - " +str(self.urls))
print("processing response.url - " + response.url)
#print ("menu: ")
#print (menu.getall())
#print ("len(menu): " + str(len(menu)))
#print ("menu[0] : " + menu.get())
#print("name - " + menu[0].css('.category-filter__text ::text').get())
#inspect_response(response,self)
if (len(menu) > 0 and menu[0].css('[aria-current="page"]')):
print(f"inside menu page for url - {url}")
# The top page is active
#print ("menu[0] : [aria-current=page] " + menu[0].css('[aria-current="page"]').get())
# therefore we need to scrape the links, and continue searching
# we then need to loop through each other page.
# call parse, and scrape it is not
menu_url = menu[0].css('::attr(href)').get()
menu_name = menu[0].css('.category-filter__text ::text').get()
for item in menu:
heading = item.css('.category-filter__text ::text').get()
scraped_url = item.css('::attr(href)').get()
scraped_url = self.base_url + scraped_url
section = menu_name
subsection = heading
category = lookup_category("", section, subsection)
store_url(self.conn, scraped_url, self.store_id, category,
section, subsection)
elif (len(menu) == 0):
inspect_response(response, self)
finish_url(self.conn, self.store_id, url, True)
print("finishing url - " + url)
next_url = get_next_url(self.cursor, 1, self.store_id, True)
if next_url is not None:
print("got next_url - " + next_url)
yield SplashRequest(
next_url,
self.parse,
endpoint='execute',
dont_filter=True,
args={'lua_source': self.expand_and_scroll_lua})
else:
print("Next url is none therefore we must be finished ! ")
def parse_urls(self, response):
location = response.css('[data-test="store-button"] ::text').get()
self.driver = response.request.meta['driver']
location = self.driver.find_element_by_css_selector(
'[data-test="store-button"]').text
print(f"detected location - {location}")
if location != self.location:
self.change_store_location(response)
self.section_group = response.css(".subcategory.category")
section_group = response.css(".subcategory.category")
for section in section_group:
section_name = section.css(".css-1pita2n ::text").get()
url_nodes = section.css("ul.children a")
for url_node in url_nodes:
subsection_name = url_node.css("::text").get()
url = self.base_url + url_node.css("::attr(href)").get()
store_url(self.conn, url, self.store_id,
lookup_category("", section_name, subsection_name),
section_name, subsection_name)
finish_url(self.conn, self.store_id, response.url)
function = self.parse
item_to_find = '[add-to-cart]'
if len(self.start_urls) != 0:
next_url = self.start_urls.pop()
store_url(self.conn, next_url, self.store_id, "", "", "")
function = self.parse_urls
item_to_find = '[data-test="store-button"]'
#request = self.create_parse_request(next_url,self.parse_urls,EC.element_to_be_clickable((By.CSS_SELECTOR, '[data-test="store-button"]')))
else:
next_url = get_next_url(self.cursor, 1)
# request = self.create_parse_request(next_url,self.parse,EC.element_to_be_clickable((By.CSS_SELECTOR, '[add-to-cart]')))
if next_url is None:
print("No more URLs to parse. Finishing")
return
else:
request = self.create_parse_request(
next_url, function,
EC.element_to_be_clickable((By.CSS_SELECTOR, item_to_find)))
#FIXME these try except blocks don't actually handle timeout exceptions from navigating to the wrong url
try:
yield request
except:
print(f"Parse - Errored out processing request for - {next_url} ")
next_url = get_next_url(self.cursor, 2)
print(f"Parse - Now handling {next_url}")
request = self.create_parse_request(
next_url, self.parse,
EC.element_to_be_clickable((By.CSS_SELECTOR, '[add-to-cart]')))
yield request
def walk_menu(self, response):
# inspect_response(response,self)
self.driver = response.request.meta['driver']
self.logger.info('Inside walk_menu')
start_url = self.driver.current_url
menu_button = self.driver.find_element_by_css_selector(
'[data-automation-id="NavigationBtn"]')
menu_button.click()
time.sleep(.5)
departments = self.driver.find_elements_by_css_selector(
'.NavigationPanel__department___1DF7d button')
for department in departments:
department_name = department.get_attribute('aria-label')
department.click()
time.sleep(.5)
aisles = self.driver.find_elements_by_css_selector(
'.NavigationPanel__aisleLink___309i2')
for aisle in aisles:
url = aisle.get_attribute('href')
aisle_name = aisle.get_attribute('innerText')
# self.department_name = department_name
# self.aisle_name = aisle_name
self.logger.info(
f"department_name: {department_name}, aisle_name: {aisle_name}"
)
category = lookup_category("", department_name, aisle_name)
self.logger.info(f"Storing aisle: {aisle_name}, url: {url}")
store_url(self.conn, url, self.store_id, category,
department_name, aisle_name)
finish_url(self.conn, self.store_id, start_url, scrape_urls=True)
next_url = get_next_url(self.cursor,
1,
store_id=self.store_id,
scrape_urls=True,
filter="aisle=")
if next_url is None:
self.logger.debug(
"Next_url is None therefore we must be finished!")
return
self.next_url = next_url
pagination_request = create_parse_request(next_url,
self.handle_pagination,
EC.element_to_be_clickable(
(By.CSS_SELECTOR,
self.PAGE_LOAD)),
errback=self.retry,
meta_url=next_url,
cookies=False)
yield pagination_request
def retry_page(self, failure):
url = failure.request.url
attempt = failure.request.meta['attempt']
attempt = int(attempt) + 1
self.logger.info(
f"retrying url: {url}, on attempt: {attempt}, continuing")
if attempt > self.MAX_RETRIES:
self.logger.warning(f"Failed {url}, {attempt} times. Skipping")
finish_url(self.conn, self.store_id, url, set_val=-1)
attempt = 1
request = self.get_next_request(attempt=attempt)
yield request
def parse(self, response):
page_1_str=self.page_str+"1"
this_url = trim_url(response.url,page_1_str)
print (f"inside parse for {this_url}")
self.scrape_urls(response)
# Only scrape pages that have the page_str in the url.
if this_url.find(self.page_str) != -1:
print (f"scraping for {this_url}")
items = response.css('product-item-v2')
print(f"length of items - {len(items)}")
metadata=get_url_metadata(self.cursor,this_url)
section=metadata[1]
subsection=metadata[2]
for item in items:
name = item.css('.product-title ::text').get()
price_strings = item.css('.product-price ::text').getall()
price = clean_string(price_strings[-1],['$'])
ppu = item.css('.product-price-qty ::text').get()
unit = self.collect_units(name)
#inspect_response(response,self)
if unit == "OZ" or unit == "LB":
ounces = self.collect_ounces(name)
else:
ounces = 0
print (f"yielding - {name}, {price}, {ppu}, {ounces}, {unit}")
yield{
"name": name,
"price": price,
"ounces": ounces,
"unit": unit,
"price-per-unit": ppu,
"url": this_url,
"section": section,
"subsection": subsection
}
#Basically the website redirects us to the url and page_1_str, which isn't added to our database
# So we trim that off so we can get the url in our database
finish_url(self.conn,self.store_id,this_url)
print("finishing url - " + this_url)
next_url = get_next_url(self.cursor, 1)
if next_url is None:
print ("Next url is none therefore we must be finished ! ")
return
else:
next_request = create_parse_request(next_url,
self.check_location,
EC.element_to_be_clickable((By.CSS_SELECTOR,'#openFulfillmentModalButton')))
print(f"got next_url - {next_url}")
yield next_request
def parse(self, response):
time.sleep(1)
url = response.url
print(f"inside parse for {url}")
PRODUCTS_CSS = '#product-main'
metadata = get_url_metadata(self.cursor, url)
if metadata is None:
print("Could not find metadata for url - " + url + " - skipping")
finish_url(self.conn, self.store_id, url)
return
section = metadata[1]
subsection = metadata[2]
products = response.css(PRODUCTS_CSS)
for product in products:
name = product.css('.product-name ::text').get()
name = name.replace("'", "")
raw_price = product.css('.product-price ::text').get()
price = re.findall(
"[0-9]+.[0-9]*", raw_price
)[0] # This will filter out any of the $'s and other text in the price'
quantity = product.css('.product-quantity ::text').get()
index_split = quantity.find('|')
ppu = quantity[index_split + 1:]
amount = quantity[:index_split]
ounces = self.collect_ounces(amount)
unit = self.collect_unit(amount)
print(f"yielding - {name}, {price}, {ppu}, {ounces}, {unit}")
yield {
"name": name,
"price": price,
"ounces": ounces,
"unit": unit,
"price-per-unit": ppu,
"url": url,
"section": section,
"subsection": subsection
}
check_subsection_amount(self.cursor, url)
finish_url(self.conn, self.store_id, url)
def parse(self, response):
self.driver = response.request.meta['driver']
close_modal(self)
change_store_location(self)
url = response.url
metadata = get_url_metadata(self.cursor, url)
section = metadata[1]
subsection = metadata[2]
#check if it has a next button,
items = response.css('.cell-content-wrapper')
for item in items:
name = item.css('.cell-title-text ::text').get()
name = clean_string(name, ['\"'])
price = item.css('[data-test="amount"] .css-19m8h51 ::text').get()
price = convert_dollars(price)
quantity = item.css('[data-test="amount"] .css-cpy6p ::text').get()
unit = item.css('.cell-product-size ::text').get()
ounces = convert_to_ounces(unit)
ppu = item.css('[data-test="per-unit-price"] ::text').get()
ppu = convert_ppu(ppu)
self.logger.info(
f"name - {name}, price - {price}, quantity - {quantity}, ounces - {ounces}, ppu - {ppu}, url - {url}, section - {section}, subsection - {subsection} "
)
#inspect_response(response,self)
yield {
"name": name,
"price": price,
"ounces": ounces,
"unit": unit,
"price-per-unit": ppu,
"url": url,
"section": section,
"subsection": subsection
}
finish_url(self.conn, self.store_id, url)
request = self.get_next_request()
yield request
def handle_pagination(self, response):
self.logger.info('inside handle_pagination')
url = self.driver.current_url
next_button = self.driver.find_elements_by_css_selector(
self.NEXT_BUTTON_SELECTOR)
# inspect_response(response,self)
if len(next_button) != 0:
next_page_url = get_next_pagination(self.PAGE_STRING, url)
metadata = get_url_metadata(self.cursor, url)
category = metadata[0]
section = metadata[1]
subsection = metadata[2]
quantity = self.driver.find_element_by_css_selector(
'.Title__browseTotalCount___OWylh').get_attribute('innerText')
quantity = re.findall('[0-9]+', quantity)[0]
store_url(self.conn,
next_page_url,
self.store_id,
category,
section,
subsection,
grocery_quantity=quantity)
finish_url(self.conn, self.store_id, url, scrape_urls=True)
next_url = get_next_url(self.cursor,
1,
store_id=self.store_id,
scrape_urls=True,
filter="aisle=")
if next_url is None:
self.logger.debug(
"Next_url is None therefore we must be finished!")
return
request = create_parse_request(next_url,
self.handle_pagination,
EC.element_to_be_clickable(
(By.CSS_SELECTOR, self.PAGE_LOAD)),
errback=self.retry,
meta_url=next_url,
cookies=False)
yield request
def handle_pagination(self, response):
# if it has a page-last class, read that content, and interprolate
# else, get the last pager, page and interprolate
self.logger.info("Inside handle_pagination")
close_modal(self)
change_store_location(self)
base_url = response.url
string_location=base_url.find(self.page_string)
if string_location != -1:
base_url = base_url[:string_location]
pag_last = self.driver.find_elements_by_css_selector(
'.pagination-last.pager-item')
if pag_last:
final_page_number = pag_last[0].text
else:
last_page = self.driver.find_elements_by_css_selector(
'.pagination-page.pager-item')[-1]
final_page_number = last_page.text
final_page_number = int(final_page_number)
metadata = get_url_metadata(self.cursor, base_url)
category = metadata[0]
section = metadata[1]
subsection = metadata[2]
for page_num in range(1, final_page_number+1):
# Something like -
# https://shop.wegmans.com/shop/categories/94 ?page= 13
page_url = base_url + self.page_string + str(page_num)
store_url(self.conn, page_url, self.store_id,
category, section, subsection)
self.logger.info(f"finished handling pagination for {base_url}")
finish_url(self.conn, self.store_id, response.url, scrape_urls=True)
request = self.get_next_request()
yield request
def parse(self, response):
# This callback determines if the selected menu is
# at the top of the list, if it is then it adds the urls
# to the list and keeps going
# if its not, then it calls the lua to prepare the page
# for scraping, and then scrapes it
url = response.url
menu = response.css(".category-filter__link")
#submenu = response.css("")
#print ("self.urls - " +str(self.urls))
print("processing response.url - " + response.url)
#print ("menu: ")
#print (menu.getall())
#print ("len(menu): " + str(len(menu)))
#print ("menu[0] : " + menu.get())
#print("name - " + menu[0].css('.category-filter__text ::text').get())
#inspect_response(response,self)
if (len(menu) > 0 and menu[0].css('[aria-current="page"]')):
print(f"inside menu page for url - {url}")
# The top page is active
#print ("menu[0] : [aria-current=page] " + menu[0].css('[aria-current="page"]').get())
# therefore we need to scrape the links, and continue searching
# we then need to loop through each other page.
# call parse, and scrape it is not
menu_url = menu[0].css('::attr(href)').get()
menu_name = menu[0].css('.category-filter__text ::text').get()
for item in menu:
heading = item.css('.category-filter__text ::text').get()
scraped_url = item.css('::attr(href)').get()
scraped_url = self.base_url + scraped_url
section = menu_name
subsection = heading
category = lookup_category("", section, subsection)
store_url(self.conn, scraped_url, self.store_id, category,
section, subsection)
#self.section_dict[url]=(menu_name, heading)
#if self.urls.count(url) == 0:
# self.urls.append(url)
#urls=menu.css('::attr(href)').getall()
# Remove the the first(this) page from list to parse
#urls.pop()
#self.urls.extend(urls)
#print("urls to scrape - " + str(self.urls))
#print("local urls - " + str(urls))
"""
while len(self.urls) != 0:
url = self.urls.pop()
self.processedUrls.append(url)
#url = self.base_url + url_suffix
#print ("urls - " + str(self.urls))
#print ("pulling from url - " + url)
#print ("urls lengths - " + str(len(self.urls)))
yield SplashRequest(url,
self.parse,
endpoint='execute',
args={'lua_source': self.expand_and_scroll_lua})
"""
elif (len(menu) == 0):
inspect_response(response, self)
else:
#we are on a subpage, so now we can start scraping
#
GROCERY_SELECTOR = '.grid-item'
NAME_SELECTOR = '.small-type.detail-card-description ::text'
PRICE_SELECTOR = '.price ::text'
PRICE_PER_UNIT_SELECTOR = '.sub-headline.detail-card-subtext ::text'
metadata = get_url_metadata(self.cursor, url)
section = metadata[0]
subsection = metadata[1]
print("subpage - scraping " + url + ", from section - " + section)
for grocery in response.css(GROCERY_SELECTOR):
self.name = grocery.css(NAME_SELECTOR).extract_first()
self.price = grocery.css(PRICE_SELECTOR).extract_first()
if self.price is not None:
self.price = self.price.replace('*', '').replace('$', '')
self.ppu = grocery.css(PRICE_PER_UNIT_SELECTOR).extract_first()
if self.ppu is not None:
self.ppu = convert_ppu(self.ppu)
#inspect_response(response, self)
#parse the ounces off of the name
yield {
'name': self.name,
'price': self.price,
'price-per-unit': self.ppu,
'section': section,
'subsection': subsection,
'url': response.url
}
finish_url(self.conn, self.store_id, url)
print("finishing url - " + url)
next_url = get_next_url(self.cursor, 1)
if next_url is not None:
print("got next_url - " + next_url)
yield SplashRequest(
next_url,
self.parse,
endpoint='execute',
dont_filter=True,
args={'lua_source': self.expand_and_scroll_lua})
else:
print("Next url is none therefore we must be finished ! ")
def parse(self, response):
url = response.url
self.logger.info(f"Inside parse for {url}")
GROCERY_SELECTOR = '[data-automation-id="productTile"]'
SPONSORED_SELECTOR = '[data-automation-id="sponsoredProductTile"]'
GROCERIES_SELECTOR = GROCERY_SELECTOR + ',' + SPONSORED_SELECTOR
metadata=get_url_metadata(self.cursor,url)
section=metadata[1]
subsection=metadata[2]
for grocery in response.css(GROCERIES_SELECTOR):
NAME_SELECTOR = '[data-automation-id="name"] ::attr(name)'
name = grocery.css(NAME_SELECTOR).extract_first()
#parse the ounces off of the name
decimal_regex = "([\d]+[.]?[\d]*|[.\d]+)"
ounces = re.findall(decimal_regex + "\s*o(?:z|unces?)",
name, re.IGNORECASE)
pounds = re.findall(decimal_regex + "\s*(?:pound|lb)s?",
name, re.IGNORECASE)
count = re.findall("([\d]+)\s*(?:c(?:t|ount)|p(?:k|ack))",
name, re.IGNORECASE)
self.ounce = ounces
self.pounds = pounds
self.count = count
#Check if the arrays returned from re.findall are empty
if ounces:
ounces = parse_float(ounces[0])
else:
ounces = 0
if pounds:
pounds = parse_float(pounds[0])
else:
pounds = 0
if count:
count = parse_float(count[0])
else:
count = 0
if pounds != 0:
ounces = 16*pounds
elif count != 0:
ounces *= count
# inspect_response(response,self)
SALEPRICE_SELECTOR = '[data-automation-id="salePrice"] ::text'
PRICE_SELECTOR = '[data-automation-id="price"] ::text'
PRICE_PER_UNIT_SELECTOR = '[data-automation-id="price-per-unit"] ::text'
name=grocery.css(NAME_SELECTOR).extract_first()
name=clean_string(name,"\"")
ounces=ounces
pounds=pounds
count=count
price=str(handle_none(grocery.css(SALEPRICE_SELECTOR).extract_first())).replace('$','')
ppu=convert_ppu(grocery.css(PRICE_PER_UNIT_SELECTOR).extract_first())
yield {
'name': name,
'ounces': ounces,
'pounds': pounds,
'count': count,
'price': price,
'price-per-unit': ppu,
'section': section,
'subsection': subsection,
'url': url,
}
finish_url(self.conn,self.store_id,url)
next_url=get_next_url(self.cursor,1,store_id=self.store_id,filter="aisle=")
print(f"next_url - {next_url}")
if next_url is None:
print ("No more urls - finishing")
else:
request = create_parse_request(next_url,
self.parse,
EC.element_to_be_clickable(
(By.CSS_SELECTOR, '[aria-current="page"]')),
meta_url=next_url)
yield request
def parse(self, response):
page_1_str = self.page_str + "1"
meta_url = response.meta.get('url')
this_url = response.url #trim_url(response.url,page_1_str)
trimmed_url = trim_url(response.url, page_1_str)
self.logger.info(
f"inside parse for meta_url: {meta_url}, response.url: {response.url}"
)
# Only scrape pages that have the page_str in the url.
if this_url.find(self.page_str) == -1:
self.logger.info(
f"Skipping {this_url} because it couldn't find {self.page_str}"
)
elif meta_url != response.url:
self.logger.info(
f"meta_url: {meta_url} != response.url: {response.url}, and so we are finishing stale {meta_url}"
)
this_url = meta_url
else:
self.logger.info(f"scraping for {this_url}")
items = response.css('product-item-v2')
self.logger.info(f"length of items - {len(items)}")
## FIXME For some reason the middleware is returning an empty response for all of the urls that reach here.
#inspect_response(response,self)
metadata = get_url_metadata(self.cursor, trimmed_url)
if len(metadata) != 3:
self.logger.info(
f"Couldn't detect metadata: {metadata}, for trimmed_url: {trimmed_url}, defaulting to empty"
)
section = ""
subsection = ""
else:
section = metadata[1]
subsection = metadata[2]
for item in items:
name = item.css('.product-title ::text').get()
price_string = item.css('.product-price').get()
price = re.findall("\$([0-9]+\.[0-9]+)", price_string)[0]
ppu = item.css('.product-price-qty ::text').get()
unit = self.collect_units(name)
if unit == "OZ" or unit == "LB":
ounces = self.collect_ounces(name)
else:
ounces = 0
self.logger.info(
f"yielding - {name}, {price}, {ppu}, {ounces}, {unit}")
yield {
"name": name,
"price": price,
"ounces": ounces,
"unit": unit,
"price-per-unit": ppu,
"url": this_url,
"section": section,
"subsection": subsection
}
#Basically the website redirects us to the url and page_1_str, which isn't added to our database
# So we trim that off so we can get the url in our database
finish_url(self.conn, self.store_id, this_url)
self.logger.info("finishing url - " + this_url + ", store_id: ",
self.store_id)
# We only want requests that have the page= string in it because they have the groceries,
# Also currently we're getting some urls in our database for locations that don't match our default_store_number
# So filter those out too.
next_url = get_next_url(self.cursor,
1,
self.store_id,
filter=f"{self.store_number}%page=")
if next_url is None:
self.logger.info(
"Next url is none therefore we must be finished ! ")
return
else:
next_request = create_unfiltered_parse_request(
next_url, self.parse,
EC.element_to_be_clickable(
(By.CSS_SELECTOR, 'product-item-v2')))
self.logger.info(f"got next_url - {next_url}")
yield next_request
def parse(self, response):
GROCERY_SELECTOR = '[data-automation-id="productTile"]'
SPONSORED_SELECTOR = '[data-automation-id="sponsoredProductTile"]'
GROCERIES_SELECTOR = GROCERY_SELECTOR + ',' + SPONSORED_SELECTOR
NEXT_BUTTON = '[data-automation-id="nextButton"]'
# Handle pagination
url = response.url
print (f"working on url - {url}")
metadata=get_url_metadata(self.cursor,url)
section=metadata[1]
subsection=metadata[2]
next_page=response.css(NEXT_BUTTON).get()
if next_page is not None:
#inspect_response(response,self)
page_string="&page="
page_str_len=len(page_string)
next_page_url=get_next_pagination(page_string,url)
store_url(self.conn,next_page_url, self.store_id, lookup_category("",section,subsection) ,section, subsection)
for grocery in response.css(GROCERIES_SELECTOR):
NAME_SELECTOR = '[data-automation-id="name"] ::attr(name)'
self.name = grocery.css(NAME_SELECTOR).extract_first()
#parse the ounces off of the name
decimal_regex = "([\d]+[.]?[\d]*|[.\d]+)"
self.ounces = re.findall(decimal_regex + "\s*o(?:z|unces?)",
self.name, re.IGNORECASE)
self.pounds = re.findall(decimal_regex + "\s*(?:pound|lb)s?",
self.name, re.IGNORECASE)
self.count = re.findall("([\d]+)\s*(?:c(?:t|ount)|p(?:k|ack))",
self.name, re.IGNORECASE)
#Check if the arrays returned from re.findall are empty
if self.ounces:
self.ounces = parse_float(self.ounces[0])
else:
self.ounces = 0
if self.pounds:
self.pounds = parse_float(self.pounds[0])
else:
self.pounds = 0
if self.count:
self.count = parse_float(self.count[0])
else:
self.count = 0
if self.pounds != 0:
self.ounces = 16*self.pounds
elif self.count != 0:
self.ounces *= self.count
# inspect_response(response,self)
SALEPRICE_SELECTOR = '[data-automation-id="salePrice"] ::text'
PRICE_SELECTOR = '[data-automation-id="price"] ::text'
PRICE_PER_UNIT_SELECTOR = '[data-automation-id="price-per-unit"] ::text'
name=grocery.css(NAME_SELECTOR).extract_first()
name=clean_string(name,"\"")
ounces=self.ounces
pounds=self.pounds
count=self.count
price=str(handle_none(grocery.css(SALEPRICE_SELECTOR).extract_first())).replace('$','')
ppu=convert_ppu(grocery.css(PRICE_PER_UNIT_SELECTOR).extract_first())
url=response.url
yield {
'name': name,
'ounces': ounces,
'pounds': pounds,
'count': count,
'price': price,
'price-per-unit': ppu,
'section': section,
'subsection': subsection,
'url': url,
}
finish_url(self.conn,self.store_id,url)
next_url=get_next_url(self.cursor,1)
print(f"next_url - {next_url}")
if next_url is None:
print ("No more urls - finishing")
else:
yield SplashRequest(next_url,
self.parse,
endpoint='render.html',
args={
'wait': 10,
'section': section,
'subsection': subsection
})
def parse(self, response):
url = response.url
finish_url(self.conn, self.store_id, url)
items = response.css('.cell-content-wrapper')
metadata = get_url_metadata(self.cursor, url)
section = metadata[1]
subsection = metadata[2]
#check if it has a next button,
next_page = response.css('.pagination-next:not(.disabled)').get()
if next_page is not None:
#inspect_response(response,self)
page_string = "?page="
page_str_len = len(page_string)
i = url.find(page_string)
#if yes, check url if it has a page part on it
if i == -1:
#if no, add ?page=2 to it
next_url = url + page_string + "2"
else:
#if yes, extract page and add 1
page_number = i + page_str_len
current_page = int(url[page_number:])
next_page = current_page + 1
next_url = url[:page_number] + str(next_page)
#then add to self.urls
store_url(self.conn, next_url, self.store_id,
lookup_category("", section, subsection), section,
subsection)
for item in items:
name = item.css('.cell-title-text ::text').get()
name = clean_string(name, ['\"'])
price = item.css('[data-test="amount"] .css-19m8h51 ::text').get()
price = convert_dollars(price)
quantity = item.css('[data-test="amount"] .css-cpy6p ::text').get()
unit = item.css('.cell-product-size ::text').get()
ounces = convert_to_ounces(unit)
ppu = item.css('[data-test="per-unit-price"] ::text').get()
ppu = convert_ppu(ppu)
print(
f"name - {name}, price - {price}, quantity - {quantity}, ounces - {ounces}, ppu - {ppu}, url - {url}, section - {section}, subsection - {subsection} "
)
#inspect_response(response,self)
yield {
"name": name,
"price": price,
"ounces": ounces,
"unit": unit,
"price-per-unit": ppu,
"url": url,
"section": section,
"subsection": subsection
}
next_url = get_next_url(self.cursor, 1)
if next_url is None:
print("No more URLs to parse. Finishing")
return
request = self.create_parse_request(
next_url, self.parse,
EC.element_to_be_clickable((By.CSS_SELECTOR, '[add-to-cart]')))
if next_url is not None:
try:
yield request
except:
print(
f"Parse - Errored out processing request for - {next_url} "
)
next_url = get_next_url(self.cursor, 2)
print(f"Parse - Now handling {next_url}")
request = self.create_parse_request(
next_url, self.parse,
EC.element_to_be_clickable(
(By.CSS_SELECTOR, '[add-to-cart]')))
yield SeleniumRequest(url=next_url,
callback=self.parse,
wait_time=50,
wait_until=EC.element_to_be_clickable(
(By.CSS_SELECTOR,
'.button.full.cart.add')))
def parse(self, response):
# This callback determines if the selected menu is
# at the top of the list, if it is then it adds the urls
# to the list and keeps going
# if its not, then it calls the lua to prepare the page
# for scraping, and then scrapes it
url = response.url
menu = response.css(".category-filter__link")
#submenu = response.css("")
#self.logger.info ("self.urls - " +str(self.urls))
self.logger.info("processing response.url - " + response.url)
#self.logger.info ("menu: ")
#self.logger.info (menu.getall())
#self.logger.info ("len(menu): " + str(len(menu)))
#self.logger.info ("menu[0] : " + menu.get())
#self.logger.info("name - " + menu[0].css('.category-filter__text ::text').get())
#inspect_response(response,self)
if not (len(menu) > 0 and menu[0].css('[aria-current="page"]')):
#we are on a subpage, so now we can start scraping
# TODO check to see if we should just scrape all pages?
GROCERY_SELECTOR = '.grid-item'
NAME_SELECTOR = '.small-type.detail-card-description ::text'
PRICE_SELECTOR = '.price ::text'
PRICE_PER_UNIT_SELECTOR = '.sub-headline.detail-card-subtext ::text'
metadata = get_url_metadata(self.cursor, url)
if metadata is None:
self.logger.debug(f"Metadata is none for {url}")
metadata = ["", ""]
section = metadata[0]
subsection = metadata[1]
self.logger.info("subpage - scraping " + url +
", from section - " + section)
for grocery in response.css(GROCERY_SELECTOR):
self.name = grocery.css(NAME_SELECTOR).extract_first()
self.price = grocery.css(PRICE_SELECTOR).extract_first()
if self.price is not None:
self.price = self.price.replace('*', '').replace('$', '')
self.ppu = grocery.css(PRICE_PER_UNIT_SELECTOR).extract_first()
if self.ppu is not None:
self.ppu = convert_ppu(self.ppu)
#inspect_response(response, self)
#parse the ounces off of the name
yield {
'name': self.name,
'price': self.price,
'price-per-unit': self.ppu,
'section': section,
'subsection': subsection,
'url': response.url
}
finish_url(self.conn, self.store_id, url)
self.logger.info("finishing url - " + url)
next_url = get_next_url(self.cursor, 1, store_id=self.store_id)
if next_url is not None:
self.logger.info("got next_url - " + next_url)
yield SplashRequest(
next_url,
self.parse,
endpoint='execute',
dont_filter=True,
args={'lua_source': self.expand_and_scroll_lua})
else:
self.logger.info(
"Next url is none therefore we must be finished ! ")
