Question

我想从以下xml格式中提取问题（type ='q'）和答案（type ='a'）对作为单个数据点：

<?xml version="1.0" encoding="us-ascii"?>
<transcript id="001" >
<body>
<section name="Q&amp;A">
      <speaker id="0">
        <plist>
          <p>Thank you. We'll now be conducting the question-and-answer session. <mark type="Operator Instructions" /> Thank you. Please go ahead with your question.</p>
        </plist>
      </speaker>
      <speaker id="3" type="q">
        <plist>
          <p>Good morning. First of all, Happy New Year.</p>
        </plist>
      </speaker>
      <speaker id="2" type="a">
        <plist>
          <p>Happy New Year, sir.</p>
        </plist>
      </speaker>
      <speaker id="3" type="q">
        <plist>
          <p>Thank you. How is your pain now?.</p>
        </plist>
      </speaker>
       <speaker id="2" type="a">
            <plist>
              <p>Oh, it's better now. I think i am healing.</p>
            </plist>
          </speaker>
</section>
</body>
</transcript>

即输出应为：['早上好。首先，新年快乐。先生，新年快乐。”，“谢谢。您现在的痛苦如何？哦，现在好了。我想我正在康复。']

有人可以帮我用美丽汤做吗？我当前的代码提取了文档中的所有<p>标签，但是问题在于还有其他部分（“ Q＆A”除外）以及它们的<p>标签也被提取了。

soup = BeautifulSoup(handler, "html.parser")
texts = []
for node in soup.findAll('p'):
    text = " ".join(node.findAll(text=True))
    #text = clean_text(text)
    texts.append(text)

Answer 1

您可以使用findAll('speaker', {"type": "q"})查找问题，并使用findNext("speaker")查找相应的答案。

例如：

from bs4 import BeautifulSoup
soup = BeautifulSoup(handler, "html.parser")
for node in soup.findAll('speaker', {"type": "q"}):
    print( node.find("p").text )
    print( node.findNext("speaker").find("p").text)
    print( "--" )

输出：

Good morning. First of all, Happy New Year.
Happy New Year, sir.
--
Thank you. How is your pain now?.
Oh, it's better now. I think i am healing.
--

Answer 2

您可以分别使用find_all('speaker', type='q')和find_all('speaker', type='a')查找所有问题和所有答案。然后使用zip加入相应的问题及其答案。

代码：

questions = soup.find_all('speaker', type='q')
answers = soup.find_all('speaker', type='a')

for q, a in zip(questions, answers):
    print(' '.join((q.p.text, a.p.text)))

输出：

Good morning. First of all, Happy New Year. Happy New Year, sir.
Thank you. How is your pain now?. Oh, it's better now. I think i am healing.

如果要在列表中使用它，可以使用列表理解：

q_and_a = [' '.join((q.p.text, a.p.text)) for q, a in zip(questions, answers)]
print(q_and_a)
# ['Good morning. First of all, Happy New Year. Happy New Year, sir.',
#  "Thank you. How is your pain now?. Oh, it's better now. I think i am healing."]

根据类型提取内容的精美汤

2 个答案: