python - 使用 beautifulsoup 正则表达式不会像预期的那样匹配

Question

我正在尝试使用正则表达式来匹配带有 class="calendar-days-list2"但不是 class="calendar-days-list2 prev-next-month"的标签。我加载了一段带有包含这两个选项的标签的 HTML 示例。

当我使用 re.findall() 搜索示例 HTML 时，正则表达式匹配我想要的。当我在 beautifulsoup 中使用该示例正则表达式时，它会返回想要的和不需要的类。我不明白这是为什么，有什么想法吗？谢谢!

html = '''<td id="pagestructure_0_pagecontent_0_calendar1_2016_1_7_0" class="calendar-days-list2" width="14%">
       <span class="date-number">7</span>
            <p>
              <img src="/wac/wacassets/images/icons/h1.gif" border="0">
              <a href="http://www.woodruffcenter.org/Commerce/MuseumAdmissions?performanceId=86514">Special Exhibitions</a>
              10:00 AM
            </p>

          <td id="pagestructure_0_pagecontent_0_calendar1_2015_11_29_1"    class="calendar-days-list2 prev-next-month" width="14%"></td>
       '''

soup = BeautifulSoup(html)
# WORKS
print re.findall(r"(calendar\-days\-list2)(?!\sprev\-next\-month)",html), "\n\n"

regex = re.compile(r"(calendar\-days\-list2)(?!\sprev\-next\-month)")
# DOESN'T WORK
tds = soup.find_all("td", {"class": regex})
print tds

输出:

# re.findall                              
['calendar-days-list2'] 

# soup.find_all
[<td class="calendar-days-list2"     id="pagestructure_0_pagecontent_0_calendar1_2016_1_7_0" width="14%">
<span class="date-number">7</span>
<p>
<img border="0" src="/wac/wacassets/images/icons/h1.gif"/>
<a href="http://www.woodruffcenter.org/Commerce/MuseumAdmissions?     performanceId=86514">Special Exhibitions</a>
        10:00 AM
    </p>
</td>, <td class="calendar-days-list2 prev-next-month"       id="pagestructure_0_pagecontent_0_calendar1_2015_11_29_1" width="14%"></td>]

`

Answer 1

regex = re.compile(r"(calendar\-days\-list2)(?!\sprev\-next\-month)")
# DOESN'T WORK
tds = soup.find_all("td", {"class": regex})

这是行不通的，因为正则表达式分别应用于每个类值，而不是整个属性值。这是因为 class 是一个特殊的多值属性。最近有几个与问题相关的帖子:

• BeautifulSoup returns empty list when searching by compound class names
• Include multiple class names in findAll in BeautifulSoup4
• Disable special "class" attribute handling

可能最简单的方法是使用 CSS selector进行完整的 class 属性匹配:

soup.select('[class="calendar-days-list2"]')