topic_spider/find_source/crawler/analysis/DomAnalysis.py

import re

from bs4 import BeautifulSoup

from crawler.analysis.FilterUrl import FilterUrl


class DomAnalysis(FilterUrl):
    """
    Beautiful Soup将复杂HTML文档转换成一个复杂的树形结构,每个节点都是Python对象,所有对象可以归纳为4种:
        Tag
        NavigableString
        BeautifulSoup
        Comment
    """

    def __init__(self, dom, base_url):
        self.soup = BeautifulSoup(dom, "lxml")
        self.url = base_url
        self.pattern = re.compile("href=([a-zA-Z0-9'\"+?=.%/_]*)")

    def show_html(self):
        #https://www.crummy.com/software/BeautifulSoup/bs3/documentation.zh.html 发现prettify是u字符
        print(self.soup.prettify().encode('utf-8', 'ignore'))

    def _is_input_with_onclick(self, tag):
        return (tag.name == 'input') and (tag.get('type')=='button') and tag.has_attr('onclick')

    def get_urls(self, **kwargs):
        urls = []
        # 静态页面链接解析 和 javascript动态解析
        for tag in self.soup.find_all('a'):
            if self.judge(tag.get('href'), **kwargs):
                urls.append(self.filter(tag.get('href')))

        # 自动交互. 这里采用静态解析的思路提取交互式生成的链接
        for tag in self.soup.find_all():
            if self._is_input_with_onclick(tag):
                for item in re.findall(self.pattern, tag.get('onclick')):
                    if self.judge(self.onclick_filter(item), **kwargs):
                        urls.append(self.filter(self.onclick_filter(item)))
        return urls

    def get_items(self, **kwargs):
        items = []

        def _extract():
            name = (tag.text if len(tag.text) != 0 else None or tag.parent.text)
            name = "".join(name.split())
            if len(name) > 50:
                name = "{:.50s}".format(name)
            if tag.get('href') is None:
                return
            try:
                href = self.filter(tag.get('href'))
            except ValueError:
                return
            data = {'name': name, 'host': href}
            if data not in items:
                items.append(data)

        for tag in self.soup.find_all('a'):
            if self.judge(tag.get('href'), **kwargs):
                _extract()

        for tag in self.soup.find_all():
            if self._is_input_with_onclick(tag):
                for item in re.findall(self.pattern, tag.get('onclick')):
                    if self.judge(self.onclick_filter(item), **kwargs):
                        _extract()
        return items