使用BeautifulSoup配合寬松解析器可有效處理不良XML，1. 選用'html.parser'或'lxml'等容錯性強的解析器，避免使用嚴格的'xml'解析器；2. 解析后通過.find()、.find_all()等方法提取數(shù)據(jù)，即使標簽未閉合或結(jié)構(gòu)混亂也能恢復大部分層級；3. 對自閉合或非法標簽如能自然處理；4. 建議結(jié)合真實樣本測試并預處理編碼問題，確保解析穩(wěn)定性，最終實現(xiàn)可靠的數(shù)據(jù)提取。

When dealing with poorly formed XML in Python, BeautifulSoup is a solid choice because it’s designed to handle messy or malformed markup—unlike strict parsers like xml.etree.ElementTree or lxml (in default mode), which often fail on invalid input.

Here’s how to use BeautifulSoup effectively for parsing broken or loosely structured XML:

1. Use a Lenient Parser Backend

BeautifulSoup itself doesn’t parse the raw text; it relies on an external parser. For malformed XML, your best bet is the html.parser (built-in) or lxml (if installed), even though they're typically used for HTML. They’re more forgiving than pure XML parsers.

from bs4 import BeautifulSoup

# Example of malformed XML
malformed_xml = """
<root>
  <item id="1">
    <name>Item One</name>
  <item id="2">
    <name>Item Two</name>
  </item>
</root>
"""

# Parse using html.parser (no extra dependencies)
soup = BeautifulSoup(malformed_xml, 'html.parser')

Even though the first <item> tag isn’t closed properly, BeautifulSoup will infer the structure and build a usable tree.

2. Avoid Strict XML Parsers

Don’t use xml as the parser if the input is malformed:

# This may raise an error on bad XML
soup = BeautifulSoup(malformed_xml, 'xml')  # Avoid for broken XML

The xml parser expects well-formed input and will fail on missing closes, unescaped characters, or overlapping tags.

Stick with:

• 'html.parser' – built-in, decent tolerance
• 'lxml' – faster and more robust (requires pip install lxml)
• 'html5lib' – most forgiving, builds HTML5-compliant tree (slower, requires pip install html5lib)

soup = BeautifulSoup(malformed_xml, 'lxml')  # Recommended if lxml is available

3. Navigating and Extracting Data

Once parsed, treat the result like any BeautifulSoup object. You can search using .find(), .find_all(), or CSS selectors.

items = soup.find_all('item')
for item in items:
    print(f"ID: {item.get('id')}, Name: {item.find('name').get_text()}")

Output:

ID: 1, Name: Item One
ID: 2, Name: Item Two

Even with incorrect nesting or missing tags, BeautifulSoup usually reconstructs the hierarchy well enough for practical use.

4. Handle Self-Closing or Invalid Tags Gracefully

If your XML includes tags like <image src="pic.jpg"/> or even <br> in non-XML contexts, BeautifulSoup with lxml or html5lib handles them naturally.

broken_xml = '<data><value>10<br><value>20</value></data>'
soup = BeautifulSoup(broken_xml, 'html.parser')
values = soup.find_all('value')
# Works: extracts both values despite the <br> in between

Key Tips:

• ? Use 'lxml' or 'html.parser' for malformed XML
• ? Avoid 'xml' parser unless input is guaranteed valid
• ? Always test on real-world samples—results depend on how broken the input is
• ? Preprocess if needed (e.g., fix encoding, remove control chars)
• ? Combine with logging or validation to catch unexpected structures

Basically, if you’re stuck with real-world XML that’s not well-formed, BeautifulSoup with a tolerant parser backend is a pragmatic solution. It won’t give you a perfect DOM, but it’ll get the data out reliably in most cases.

? ??? ??? ???? ?? XML? ?? ???? ?? Python? BeautifulSoup? ?????? ?? ?????. ??? ??? PHP ??? ????? ?? ?? ??? ?????!