简单小练习_微博爬取

# start_chrome -> input_date -> scroll_down-> find_cards_info -> save -> find_next (goto)
from selenium import webdriver
from selenium.webdriver.common.keys import Keys
import time
import csv
import os

# 运行前先下载 chrome driver,下载地址是：https://sites.google.com/a/chromium.org/chromedriver/downloads，点击【Latest Release: ChromeDriver x.xx】进入下载

def start_chrome():
    driver = webdriver.Chrome(executable_path='./chromedriver')  # Windows 需写成'./chromedriver.exe'
    driver.start_client()
    return driver
# weibo.com/sxbg? + start_t ...

def q(st,et):
    return f'?is_ori=1&key_word=&start_time={st}&end_time={et}&is_search=1&is_searchadv=1#_0'

def scroll_down():
    html_page = driver.find_element_by_tag_name('html')
    # ...
    # form > input
    for i in range(15):
        print(i)
        html_page.send_keys(Keys.END)
        time.sleep(0.6)

def find_cards_info():
    cards_sel = 'div.WB_feed_detail'
    cards     = driver.find_elements_by_css_selector(cards_sel)
    info_list = []

    for card in cards:
        content_sel = 'div.WB_text.W_f14'
        time_sel    = 'div.WB_from.S_txt2'
        link_sel    = 'div.WB_from.S_txt2 > a:nth-child(1)'

        content     = card.find_element_by_css_selector(content_sel).text
        time        = card.find_element_by_css_selector(time_sel).text
        link        = card.find_element_by_css_selector(link_sel).get_attribute('href')

        info_list.append([content,time,link])
        # [[1,2,3],[4,5,6]...]
    return info_list

def find_next():
    next_sel  = 'a.page.next'
    next_page = driver.find_elements_by_css_selector(next_sel)
    if next_page:
        return next_page[0].get_attribute('href')

def save(info_list,name):
    full_path = './' + name + '.csv' # 2018-01-02~2018-03-05.csv
    if os.path.exists(full_path):
        with open(full_path,'a') as f:
            writer = csv.writer(f)
            writer.writerows(info_list)
            print('Done')
    else:
        with open(full_path,'w+') as f:
            writer = csv.writer(f)
            writer.writerows(info_list)
            print('Done')


def run_crawler(base,duration):
    # 2018-01-02~2018-03-05
    if not base.endswith('feedtop'):
        st, et = duration.split('~')
        driver.get(base+q(st,et))
    else:
        driver.get(base)
    time.sleep(5)
    scroll_down()
    time.sleep(5)
    info_list = find_cards_info()
    save(info_list,duration)
    next_page = find_next()
    if next_page:
        run_crawler(next_page,duration)

base = 'https://weibo.com/erick'   #这里可以替换成你想爬取的微博用户
driver = start_chrome()
input()
run_crawler(base, '2017-06-20~2018-02-03')   #这里可以替换成你想爬取的时间段