html正文提取工具goose的安装及简单使用Demo

1.git clone https://github.com/grangier/python-goose.git

2.cd python-goose

3.sudo pip install -r requirements.txt
此时会报一个安装nltk的错误,执行下面命令单独安装:

sudo apt-get install python-nltk 

4.sudo python setup.py install

 

至此安装完毕!!!!!!!

---------------------------------------------------------

下面付简单的使用demo:

def goose_extraction(response):

    try:

import traceback

        import chardet

        from goose import Goose

        from goose.text import StopWordsChinese

        charset = chardet.detect(response.content)

        coding = charset.get('encoding').lower()  # 网页编码类别：gbk,gb2312,utf-8等

        if coding and coding.startswith(u'gb'):

            codeHtml = response.content.decode("GB18030").encode('utf-8')

        elif coding.startswith(u'utf'):

            codeHtml = response.content

        else:

            codeHtml = response.content.decode(coding, 'ignore')

        g = Goose({'stopwords_class': StopWordsChinese})  #　中文

        article = g.extract(raw_html=codeHtml)

        content = article.cleaned_text

        html = '<div>' + ''.join(['<p>'+con+'</p>\n' for con in content.split('\n\n')]) + '</div>'

        return content, html

    except Exception as e:

        traceback.print_exc(e)