class TaobaoSpider(scrapy.Spider):
    name = 'taobao'
    # allowed_domains = ['www.taobao.com']
    base_url = ['https://s.taobao.com/search?q=']
    pages = 100
    re_headers = {
        'user-agent': 'Mozilla/5.0 (Windows NT 6.1; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/65.0.3325.181 Safari/537.36',
        'referer': 'https://www.taobao.com/',
        'accept-encoding': 'gzip, deflate, b',
    }
    i = 1

    def start_requests(self):
        keys = self.settings.get('KEYS')#獲取要搜索的關鍵詞
        self.browser,list = register()#這里調用selenium登錄的方法并返回browser和一個cookies
        self.browser.get(self.base_url[0]+keys)#使用browser登錄淘寶商品搜索頁面
        self.browser.execute_script("window.scrollTo(0, document.body.scrollHeight)")#使用execute_script執行js操作，這里是下拉到最底下
        url_i = self.browser.current_url#獲取selenium界面當前的url用來錯誤處理
        html = self.browser.page_source#獲取源代碼
        yield scrapy.Request(url=self.base_url[0]+keys,headers=self.re_headers,cookies=list,callback=self.parse,meta={'html':html,'i':self.i,'url':url_i})

    def parse(self, response):
        time.sleep(5)#等待時間，可調
        html = response.meta.get('html')
        i = response.meta.get("i")
        url_i = response.meta.get("url")
        i +=1
        if i > 100:#因為翻一百頁，到了之后就不在執行循環
            return
        try:
            soup = BeautifulSoup(html,'html.parser')
            lists = soup.select('#mainsrp-itemlist > div > div > div > div')
            for list in lists:#這一段是解析數據
                item = TaobaoSItem()
                url = list.select('a[class="pic-link J_ClickStat J_ItemPicA"]')[0].attrs.get('href','')
                name = list.select("a[class='J_ClickStat']")[0].get_text().strip()
                name = data_cleaning(name)
                price = list.select('div[class="price g_price g_price-highlight"] strong')[0].get_text()
                num = list.select('div[class="deal-cnt"]')[0].get_text()
                shop_name = list.select("a[class='shopname J_MouseEneterLeave J_ShopInfo']")[0].get_text().strip()
                shop_name = data_cleaning(shop_name)
                item['url'] = url
                item['name'] = name
                item['price'] = price
                item['num'] = num
                item['shop_name'] = shop_name
                yield item
            button = self.browser.find_elements(By.XPATH,'//a[@class="J_Ajax num icon-tag"]')[-1]#這里是獲取點擊下一頁的，因為到第二頁以后會有二個一樣class的，一個是上一頁，一個是下一頁。
            button.click()#點擊進入下一頁
            time.sleep(random.random()*2)
            self.browser.execute_script("window.scrollTo(0, document.body.scrollHeight)")#下拉操作
            html = self.browser.page_source
            yield scrapy.Request(url=response.url,headers=self.re_headers,callback=self.parse,meta={'html':html,'i':i,'url':url_i},dont_filter=True)
        except Exception as e:#如果被淘寶抓到就重新登錄，用保存的url在接著獲取數據
            time.sleep(10)
            print(e)
            self.browser.close()
            self.browser,list = register()
            self.browser.get(url=url_i)
            time.sleep(random.random()*2)
            self.browser.execute_script("window.scrollTo(0, document.body.scrollHeight)")
            html = self.browser.page_source
            yield scrapy.Request(url=response.url,headers=self.re_headers,callback=self.parse,meta={'html':html,'i':i,'url':url_i},dont_filter=True)

    def close(spider, reason):#這是結束時執行的函數，用來關掉開啟的瀏覽器進程
        spider.browser.close()

def data_cleaning(data):#這是清洗數據的
    if ' ' in data:
        data = re.sub(' ', '', data)
    if "'" in data:
        data = re.sub("'", '', data)
    if r'\n' in data:
        data = re.sub(r'\\n', '', data)
    return data

def register():#這是登錄的函數，主要
    while True: #因為淘寶能夠識別出selenium，有時我們會登錄失敗，會重新登錄
        browser = webdriver.FirefoxOptions()
        browser.add_argument('-headless') #無頭瀏覽器
        browser = webdriver.Firefox(firefox_options=browser) 
        # browser = webdriver.Firefox()
        browser.get('https://login.taobao.com/member/login.jhtml')#進入登錄頁面try:
            input = WebDriverWait(browser, 10).until(
                EC.presence_of_element_located((By.CLASS_NAME, 'forget-pwd.J_Quick2Static'))) #因為登錄頁面有時候是掃碼登錄，使用需要我們點擊切換到密碼登錄
            input.click()
        except Exception as e: #因為頁面有時是直接密碼登錄，使用如果直接是密碼登錄就不需要點擊
            print(e)
        user = browser.find_element(By.ID, 'TPL_username_1')#找到賬號輸入框
        password = browser.find_element(By.ID, 'TPL_password_1')#密碼輸入框
        user.send_keys(USER) #輸入賬號并等待一下
        time.sleep(random.random() * 2)
        password.send_keys(PASSWORD)#輸入密碼并等待一下
        time.sleep(random.random() * 1)
        browser.execute_script("Object.defineProperties(navigator,{webdriver:{get:() => false}})") #淘寶對selenium的識別主要是通過navigator.webdriver,使用selenium的瀏覽器api顯示的是True，所有我們改成fALSE就可以過淘寶的檢測
        action = ActionChains(browser)
        time.sleep(random.random() * 1)
        butt = browser.find_element(By.ID, 'nc_1_n1z')
        browser.switch_to.frame(browser.find_element(By.ID, '_oid_ifr_'))
        browser.switch_to.default_content()
        action.click_and_hold(butt).perform()
        action.reset_actions()
        action.move_by_offset(285, 0).perform()#輸入賬號密碼后會有一個滑動驗證
        time.sleep(random.random() * 1)
        button = browser.find_element(By.ID, 'J_SubmitStatic')#登錄按鈕
        time.sleep(random.random() * 2)
        button.click()
        time.sleep(random.random() * 2)
        # browser.get('https://www.taobao.com/')
        cookie = browser.get_cookies()#獲取cookies，原本想selenium實現登錄，其他使用scrapy來，但是淘寶的商品搜索頁的js找不到加上時間不夠就沒寫了。
        list = {}#scrapy攜帶的cookies需要字典類型的
        for cookiez in cookie:
            name = cookiez['name']
            value = cookiez['value']
            list[name] = value
        if len(list) > 10:
            break
        else:
            browser.close()
    return browser,list

class TaobaoSPipeline(object):
    def open_spider(self,spider): #scrapy打開時啟動，這里是打開或者新建一個txt文件，文件路徑是當前路徑
        self.f = open('淘寶店鋪數據.txt','w')
    def process_item(self, item, spider):#數據保存以字典的形式，也可以改成數據庫或者csv
        data = {}
        data['url'] = item['url']
        data['name'] = item['name']
        data['price'] = item['price']
        data['num'] = item['num']
        data['shop_name'] = item['shop_name']
        self.f.write(str(data)+'\n')
        return item
    def close_spider(self,spider):#scrapy結束時啟動，用來關掉文件。
        self.f.close()