方法一：beautifulsoup+pd.read_html

r = self.session.get(url_d, params=params_d, headers=self.headers)
# 解析网页
soup = BeautifulSoup(r.text, 'lxml')
content = soup.select('body > div > div > div > div > div.ibox-content > table')[0]  # [0]将返回的list改为bs4类型，参数为selector选择器
tbl = pd.read_html(content.prettify(), header=0)[0]
tbl.to_excel('test.xlsx', index=False)

缺点：对于首行不是标题或者合并居中单元格的网页数据不友好

方法二：etree.HTML+xpath

html = etree.HTML(r.content.decode('utf-8'))
table = html.xpath("//table/tbody")[0]  # XPath定位到表格，因为页面只有一个表格，所以直接//table，# 如果有多个表格，如取第二个表格，则写为//table[1] 偏移量为1 。我们不取表头信息，所以从tr[3]开始取，返回一个列表
short_info = {
                ''.join(table.xpath(".//tr[1]/td[2]/b/text()")): ''.join(table.xpath(".//tr[1]/td[3]/text()")),
                ''.join(table.xpath(".//tr[1]/td[4]/b/text()")): ''.join(table.xpath(".//tr[1]/td[5]/text()")),
                ''.join(table.xpath(".//tr[1]/td[6]/b/text()")): ''.join(table.xpath(".//tr[1]/td[7]/text()")),
                ''.join(table.xpath(".//tr[1]/td[8]/b/text()")): ''.join(table.xpath(".//tr[1]/td[9]/text()"))}
# 获取文本的另一种方法：当值为空时，列表索引[0]报错，而上面这种方法则没有这种问题
short_info = {
                table.xpath(".//tr[1]/td[2]/b/text()")[0]: table.xpath(".//tr[1]/td[3]/text()")[0],
                table.xpath(".//tr[1]/td[4]/b/text()")[0]: table.xpath(".//tr[1]/td[5]/text()")[0],
                table.xpath(".//tr[1]/td[6]/b/text()")[0]: table.xpath(".//tr[1]/td[7]/text()")[0],
                table.xpath(".//tr[1]/td[8]/b/text()")[0]: table.xpath(".//tr[1]/td[9]/text()")[0]}
cols = html.xpath("//table/tbody/tr[3]/td")
cols = [''.join(x.xpath(".//text()")) for x in cols]  # 标题行
data = []
rows = html.xpath("//table/tbody/tr")[3:]  # 数据行，从第4个开始
for row in rows:
    row = row.xpath(".//td")
    r_data = []
    for cell in row:
        val = ''.join(cell.xpath(".//text()"))
        val = val.replace('\n', "").strip()  # 删除换行符、空格
        try:
            val = float(val)
        except:
            pass
        r_data.append(val)
    data.append(r_data)
# 创建dataframe
df = pd.DataFrame(data=data, columns=cols)