到目前为止我已经尝试过的

Question

我下面有HTML文件，其中包含来自PDF文件的bbox信息：

<flow>
  <block xMin="53.879997" yMin="369.965298" xMax="63.939976" yMax="380.991433">
    <line xMin="53.879997" yMin="369.965298" xMax="63.939976" yMax="380.991433">
      <word xMin="53.879997" yMin="369.965298" xMax="63.939976" yMax="380.991433">10</word>
    </line>
  </block>
</flow>
<flow>
  <block xMin="53.879997" yMin="417.965298" xMax="63.939976" yMax="428.991433">
    <line xMin="53.879997" yMin="417.965298" xMax="63.939976" yMax="428.991433">
      <word xMin="53.879997" yMin="417.965298" xMax="63.939976" yMax="428.991433">20</word>
    </line>
  </block>
</flow>
<flow>
  <block xMin="111.351361" yMin="369.965298" xMax="134.220382" yMax="380.991433">
    <line xMin="111.351361" yMin="369.965298" xMax="134.220382" yMax="380.991433">
      <word xMin="111.351361" yMin="369.965298" xMax="116.331548" yMax="380.991433">1</word>
      <word xMin="121.909358" yMin="369.965298" xMax="134.220382" yMax="380.991433">PC</word>
    </line>
  </block>
</flow>

上方是以下单词的边界框区域：10 20 1 PC

在原始文档中，它是这样写的：

10 1 PC
20

因此，我想解析上述HTML文件并提取 all <line>标签，然后按yMin值对所有标签进行排序。上面的最终输出将是：10 1 PC 20。

到目前为止我已经尝试过的

我还不太远，因为我还在学习Python。我正在使用BeautifulSoup4：

with open("test.html", "r") as f:
    contents = f.read()
    soup = BeautifulSoup(contents, 'lxml')

    for line in soup.find_all("line", attrs={"ymin":True}):
        print(line.get('ymin'))

上面只是打印出每个标签及其内容。

我不确定如何对行标签进行排序。

任何帮助将不胜感激。

Answer 1

您可以将BeautifulSoup与soup.find_all一起使用：

from bs4 import BeautifulSoup as soup
r = [i.find_all('word') for i in sorted(soup(html, 'html.parser').find_all('line'), key=lambda x:float(x['ymin']))]
result = [i.text for b in r for i in b]

输出：

['10', '1', 'PC', '20']

Answer 2

尝试下面的代码。可以定义平均值，然后检查平均值。

from bs4 import BeautifulSoup
html='''<flow>
  <block xMin="53.879997" yMin="369.965298" xMax="63.939976" yMax="380.991433">
    <line xMin="53.879997" yMin="369.965298" xMax="63.939976" yMax="380.991433">
      <word xMin="53.879997" yMin="369.965298" xMax="63.939976" yMax="380.991433">10</word>
    </line>
  </block>
</flow>
<flow>
  <block xMin="53.879997" yMin="417.965298" xMax="63.939976" yMax="428.991433">
    <line xMin="53.879997" yMin="417.965298" xMax="63.939976" yMax="428.991433">
      <word xMin="53.879997" yMin="417.965298" xMax="63.939976" yMax="428.991433">20</word>
    </line>
  </block>
</flow>
<flow>
  <block xMin="111.351361" yMin="369.965298" xMax="134.220382" yMax="380.991433">
    <line xMin="111.351361" yMin="369.965298" xMax="134.220382" yMax="380.991433">
      <word xMin="111.351361" yMin="369.965298" xMax="116.331548" yMax="380.991433">1</word>
      <word xMin="121.909358" yMin="369.965298" xMax="134.220382" yMax="380.991433">PC</word>
    </line>
  </block>
</flow>'''

soup=BeautifulSoup(html,'lxml')
pricemin=soup.select_one('line[yMin]')['ymin']
list1=[]
list_last=[]
for item in soup.select('line[yMin]'):
    if float(pricemin) < float(item['ymin']):

         for w in item.select('word'):
             list_last.append(w.text)
    else:
        for w in item.select('word'):
            list1.append(w.text)

print(list1+list_last)

输出：

['10', '1', 'PC', '20']

要打印此

print(' '.join(list1+list_last))

输出：

10 1 PC 20

用漂亮的汤解析并排序html标签

到目前为止我已经尝试过的

2 个答案: