知名的pyspider，scrapy就不说了，今天说说这个 looter。
安装
先安装好python3，需要3.6以上，然后执行 pip install looter
λ looter -hLooter, a python package designed for web crawler lovers :)Author: alphardex QQ:2582347430If any suggestion, please contact me.Thank you for cooperation!
Usage: looter genspider <name> [--async] looter shell [<url>] looter (-h | --help | --version)
Options: -h --help Show this screen. --version Show version. --async Use async instead of concurrent.
图片爬虫
λ looter shell https://konachan.com/post
Available objects: url The url of the site you crawled. res The response of the site. tree The element source tree to be parsed.
Available functions: fetch Send HTTP request to the site and parse it as a tree. [has async version] view View the page in your browser. (test rendering) links Get the links of the page. save Save what you crawled as a file. (json or csv)
Examples: Get all the <li> elements of a <ul> table: >>> items = tree.css('ul li')
Get the links with a regex pattern: >>> items = links(res, pattern=r'.*/(jpeg|image)/.*')
For more info, plz refer to documentation: [looter]: https://looter.readthedocs.io/en/latest/ imgs = tree.css('a.directlink::attr(href)').extract()>>> imgs[1:10]['https://konachan.com/jpeg/c67d38b73df6e32199127998fc0f3338/Konachan.com%20-%20283270%20ass%20bed%20blush%20breasts%20clover_%28sakura_gamer%29%20game_cg%20nipples%20pussy_juice%20red_hair%20sakura_gamer%20wanaca%20winged_cloud.jpg', 'https://konachan.com/image/a0952daaf9aa94cd676901203680fec4/Konachan.com%20-%20283269%20aliasing%20anus%20azur_lane%20blush%20breasts%20cum%20gray_hair%20group%20long_hair%20nipples%20nude%20penis%20pussy%20rak_%28kuraga%29%20red_eyes%20twintails%20uncensored.jpg', 'https://konachan.com/image/e8ea71c93a895d87338ebf17e3aef5b3/Konachan.com%20-%20283268%20aliasing%20anthropomorphism%20azur_lane%20blush%20breasts%20gray_hair%20group%20long_hair%20nipples%20nude%20penis%20pussy%20rak_%28kuraga%29%20red_eyes%20sex%20twintails%20uncensored.jpg', 'https://konachan.com/image/8ffb6f968ffe372ea90a339934a9749d/Konachan.com%20-%20283267%20bed%20blush%20brown_eyes%20brown_hair%20condom%20inanaki_shiki%20long_hair%20navel%20no_bra%20open_shirt%20original%20panties%20tie%20underwear.jpg', 'https://konachan.com/jpeg/0d1de5c59eaf6fc717d63912e076de1d/Konachan.com%20-%20283266%20ass%20bed%20black_hair%20brown_eyes%20long_hair%20matsuzaki_miyuki%20original%20ponytail%20shorts.jpg', 'https://konachan.com/jpeg/7b34654c53e43879f20a8fd642c32acc/Konachan.com%20-%20283264%20aqua_eyes%20bed%20blonde_hair%20blush%20breasts%20censored%20dark_skin%20navel%20nipples%20no_bra%20original%20penis%20pubic_hair%20pussy%20sex%20shirt_lift%20spread_legs%20tan_lines.jpg', 'https://konachan.com/image/00a0eb43c07e9361679b5389e284ef7f/Konachan.com%20-%20283263%20ass%20ball%20brown_eyes%20cameltoe%20dress%20erect_nipples%20gray_hair%20kokkoro%20loli%20panties%20pizanuko%20pointed_ears%20princess_connect%21%20underwear%20upskirt%20wristwear.jpg', 'https://konachan.com/jpeg/889214118e9a891c63f0cb759d809775/Konachan.com%20-%20283262%202girls%20animal%20bow%20brown_eyes%20brown_hair%20clouds%20dress%20feathers%20flowers%20gloves%20green_eyes%20headdress%20idolmaster%20loli%20ribbons%20rose%20short_hair%20sky%20tiara.jpg', 'https://konachan.com/image/c7a3f7f9d6a2c1dc17c4c13733f72aed/Konachan.com%20-%20283261%20bikini_top%20black_hair%20blue_eyes%20boots%20chain%20flat_chest%20gloves%20hoodie%20inosia%20kuroi_mato%20long_hair%20magic%20navel%20scar%20shorts%20signed%20sword%20twintails%20weapon.jpg']Path('konachan.txt').write_text('\n'.join(imgs))wget -i konachan.txt
抓取 v2
import timeimport looter as ltfrom pprint import pprintfrom concurrent import futures
domain = 'https://www.v2ex.com'total = []
def crawl(url): tree = lt.fetch(url) items = tree.css('#TopicsNode .cell') for item in items: data = {} data['title'] = item.css('span.item_title a::text').extract_first() data['author'] = item.css('span.small.fade strong a::text').extract_first() data['source'] = f"{domain}{item.css('span.item_title a::attr(href)').extract_first()}" reply = item.css('a.count_livid::text').extract_first() data['reply'] = int(reply) if reply else 0 pprint(data) total.append(data) time.sleep(1)
if __name__ == '__main__': tasklist = [f'{domain}/go/python?p={n}' for n in range(1, 10)] [crawl(task) for task in tasklist] lt.save(total, name='v2ex.csv', sort_by='reply', order='desc')
抓取10页python主题的数据，按照回复数倒序排列
,author,reply,source,title0,chinesehuazhou,127,https://www.v2ex.com/t/562327#reply127,10 行 Python 代码，批量压缩图片 500 张，简直太强大了（内有公号宣传，不喜勿进）1,chinesehuazhou,103,https://www.v2ex.com/t/557286#reply103,len(x) 击败 x.len()，从内置函数看 Python 的设计思想（内有公号宣传，不喜勿进）2,nfroot,73,https://www.v2ex.com/t/555249#reply73,面对 Python 的强大和难用性表示深深的迷茫，莫非打开方式不对？3,css3,58,https://www.v2ex.com/t/554724#reply58,你们用什么工具来管理 Python 的库啊？4,Northxw,54,https://www.v2ex.com/t/558529#reply54,花式反爬之某众点评网5,akmonde,48,https://www.v2ex.com/t/559926#reply48,Python 项目移植到其他机器，要求全 Linux 系统适配6,kayseen,47,https://www.v2ex.com/t/562683#reply47,这道 Python 题目有大神会做吗?7,hellomacos,41,https://www.v2ex.com/t/562413#reply41,老生常谈的问题：如何学好 Python