我正在使用Python中的BeautifulSoup将许多HTML文件大规模转换为XML。
一个示例HTML文件如下所示:
<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN"
"http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
<!-- this is an HTML comment -->
<!-- this is another HTML comment -->
<html xmlns="http://www.w3.org/1999/xhtml">
<head>
...
<!-- here is a comment inside the head tag -->
</head>
<body>
...
<!-- Comment inside body tag -->
<!-- Another comment inside body tag -->
<!-- There could be many comments in each file and scattered, not just 1 in the head and three in the body. This is just a sample. -->
</body>
</html>
<!-- This comment is the last line of the file -->
我想好了如何找到doctype并用标记<doctype>...</doctype>
替换它,但注释让我很沮丧。我想用<comment>...</comment>
替换HTML注释。在这个示例HTML中,我能够替换前两个HTML注释,但html
标记内的任何内容和结束HTML标记后的最后一个注释我都不能。
这是我的代码:
file = open ("sample.html", "r")
soup = BeautifulSoup(file, "xml")
for child in soup.children:
# This takes care of the first two HTML comments
if isinstance(child, bs4.Comment):
child.replace_with("<comment>" + child.strip() + "</comment>")
# This should find all nested HTML comments and replace.
# It looks like it works but the changes are not finalized
if isinstance(child, bs4.Tag):
re.sub("(<!--)|(<!--)", "<comment>", child.text, flags=re.MULTILINE)
re.sub("(-->)|(--&gr;)", "</comment>", child.text, flags=re.MULTILINE)
# The HTML comments should have been replaced but nothing changed.
print (soup.prettify(formatter=None))
这是我第一次使用BeautifulSoup。如何使用BeautifulSoup查找所有HTML注释并将其替换为<comment>
标记?
我可以通过pickle
将其转换为字节流,对其进行序列化,应用regex,然后将其取消关联为BeautifulSoup
对象吗?这会奏效吗?还是只会引发更多问题?
我尝试在子标记对象上使用pickle,但TypeError: __new__() missing 1 required positional argument: 'name'
反序列化失败。
然后,我尝试通过child.text
只酸洗标记的文本,但由于AttributeError: can't set attribute
,反序列化失败。基本上,child.text
是只读的,这解释了regex不起作用的原因。所以,我不知道如何修改文本。
您有几个问题:
-
您不能修改
child.text
。它是一个只读属性,只在后台调用get_text()
,其结果是一个与文档无关的全新字符串。 -
re.sub()
不会修改任何内容。您的线路re.sub("(<!--)|(<!--)", "<comment>", child.text, flags=re.MULTILINE)
必须是
child.text = re.sub("(<!--)|(<!--)", "<comment>", child.text, flags=re.MULTILINE)
但由于第1点的原因,这无论如何都不起作用。
-
试图通过用正则表达式替换文档中的文本块来修改文档是使用BeautifulSoup的错误方法。相反,您需要找到节点并用其他节点替换它们。
这里有一个有效的解决方案:
import bs4
with open("example.html") as f:
soup = bs4.BeautifulSoup(f)
for comment in soup.find_all(text=lambda e: isinstance(e, bs4.Comment)):
tag = bs4.Tag(name="comment")
tag.string = comment.strip()
comment.replace_with(tag)
这段代码首先迭代调用find_all()
的结果,利用我们可以将函数作为text
参数传递的事实。在BeautifulSoup中,Comment
是NavigableString
的一个子类,所以我们把它当作字符串来搜索,而lambda ...
只是的缩写
def is_comment(e):
return isinstance(e, bs4.Comment)
soup.find_all(text=is_comment)
然后,我们用适当的名称创建一个新的Tag
,将其内容设置为原始注释的剥离内容,并用我们刚刚创建的标签替换注释。
结果如下:
<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN"
"http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
<comment>this is an HTML comment</comment>
<comment>this is another HTML comment</comment>
<html xmlns="http://www.w3.org/1999/xhtml">
<head>
...
<comment>here is a comment inside the head tag</comment>
</head>
<body>
...
<comment>Comment inside body tag</comment>
<comment>Another comment inside body tag</comment>
<comment>There could be many comments in each file and scattered, not just 1 in the head and three in the body. This is just a sample.</comment>
</body>
</html>
<comment>This comment is the last line of the file</comment>