Overview 概览

tinyCrawl 是一个微型的爬虫框架，具有以下特点：

• 简单轻巧，没有任何第三方包的依赖
• checkpoint断点续爬
• 支持多线程爬取
• 内置日志功能
• 使用简单

Documentation 文档

访问官方说明文档

Installation 安装

pip install tinyCrawl

How to use 如何使用

tinyCrawl 支持2种运行方法

• By using funciton 函数式：定义爬虫的方法 task()，实例化 BaseCrawl(iter_url, iter_num_range, thread_num)，调用 run 方法执行
• By inheritance 继承式：继承 BaseCrawl(iter_url, iter_num_range, thread_num) ，重写 crawl() 和 sink() 方法，其中 crawl()类似于上一方法中的爬虫方法task()，定义爬取单页的爬虫代码， sink()是将结果输出的方法，最后执行 main() 方法执行总程序

By using funciton 函数式

# -*- coding: utf-8 -*-

from tinyCrawl import BaseCrawl, RowContainer

from urllib.request import urlopen
from lxml import etree

# 定义xpath
song_name_xpath = '//div[@class="song-name"]/a/text()'
singer_xpath = '//div[@class="singers"]/a[1]/text()'
album_xpath = '//div[@class="album"]/a[1]/text()'

def task(url):
    """
    定义爬取单页的爬虫代码

    """
    # 定义数据存放的容器，容器名字就是最后爬取结果存放字典的self.out的key
    song_name_list = RowContainer("song name")
    singer_list = RowContainer("singer")
    album_list = RowContainer("album")

    page = urlopen(url).read().decode("utf-8", 'ignore')
    parse = etree.HTML(page)
    for _song_name, _singer, _album in zip(parse.xpath(song_name_xpath),
                                           parse.xpath(singer_xpath),
                                           parse.xpath(album_xpath)):
       	# 将数据append进指定容器中
        song_name_list.append(str(_song_name))
        singer_list.append(str(_singer))
        album_list.append(str(_album))


# 第一个参数是链接地址，需通过 %s 定义页数等迭代的参数
# 第二个参数是迭代的范围，
# 第三个参数是启用的线程数，大于1就是多线程，等于1就是单线程
bc = BaseCrawl("http://example.com/?page=%s", range(1, 5), 3)
# 输入task的对象，开始执行程序
bc.run(task)
# 执行完毕后，通过out属性，获取结果
print(bc.out)

By inheritance 继承式

from tinyCrawl import BaseCrawl, RowContainer
from urllib.request import urlopen
from lxml import etree
import pandas as pd

# 需继承BaseCrawl类，覆写crawl和sink方法
class Scratch(BaseCrawl):
    def __init__(self, iter_url, iter_num_range, thread_num):
        super().__init__(iter_url, iter_num_range, thread_num)

    # 覆写crawl方法
    def crawl(self, url):
    # 定义数据存放的容器，容器名字就是最后爬取结果存放字典的self.out的key
    song_name_list = RowContainer("song name")
    singer_list = RowContainer("singer")
    album_list = RowContainer("album")

    page = urlopen(url).read().decode("utf-8", 'ignore')
    parse = etree.HTML(page)
    for _song_name, _singer, _album in zip(parse.xpath(song_name_xpath),
                                           parse.xpath(singer_xpath),
                                           parse.xpath(album_xpath)):
       	# 将数据append进指定容器中
        song_name_list.append(str(_song_name))
        singer_list.append(str(_singer))
        album_list.append(str(_album))

    # 覆写sink方法，将爬取的结果输出
    def sink(self):
        # self.out是字典结构的结果，可以直接输入pandas存为dataframe
        recent_music = pd.DataFrame(self.out)
        recent_music.to_csv("D:/tmptest.csv", index=0)


if __name__ == '__main__':
    mc = Scratch("http://example.com/?page=%s", range(1, 5), 3)
    # 调用main函数执行程序
    mc.main()

output：

2021-01-10 16:18:36,944 - base.py - __init__ - [line:30] - INFO: Checkpoint path: D:\breakpoint_page.txt
2021-01-10 16:18:38,539 - base.py - __source - [line:119] - INFO: Now is running on multithread mode, total thread num is `3`
2021-01-10 16:18:38,539 - base.py - __source - [line:126] - INFO: Total iteration num: 4
2021-01-10 16:18:38,541 - base.py - _multi_thread_wrap - [line:59] - INFO: ThreadPoolExecutor-1_0 now is processing: http://example.com/?page=1
2021-01-10 16:18:38,541 - base.py - _multi_thread_wrap - [line:59] - INFO: ThreadPoolExecutor-1_1 now is processing: http://example.com/?page=2
2021-01-10 16:18:38,542 - base.py - _multi_thread_wrap - [line:59] - INFO: ThreadPoolExecutor-1_2 now is processing: http://example.com/?page=3
2021-01-10 16:18:41,544 - base.py - __task_done - [line:115] - INFO: ThreadPoolExecutor-1_1 task finished; (Time took: 3.0009s)
2021-01-10 16:18:41,544 - base.py - __task_done - [line:115] - INFO: ThreadPoolExecutor-1_0 task finished; (Time took: 3.0019s)
2021-01-10 16:18:41,544 - base.py - __task_done - [line:115] - INFO: ThreadPoolExecutor-1_2 task finished; (Time took: 3.0009s)
2021-01-10 16:18:41,545 - base.py - _multi_thread_wrap - [line:59] - INFO: ThreadPoolExecutor-1_1 now is processing: http://example.com/?page=4
2021-01-10 16:18:44,551 - base.py - __task_done - [line:115] - INFO: ThreadPoolExecutor-1_1 task finished; (Time took: 3.0022s)
2021-01-10 16:18:44,551 - base.py - __source - [line:151] - INFO: All done. (Time took: 6.0102s)