Parser change tag name

[Xiaomeng Yi]

from bs4 import BeautifulSoup

content = """
<?xml version="1.0" encoding="UTF-8"?>
<cj-api>
    <links total-matched="32" records-returned="10" page-number="1">
        <link>
            <advertiser-id>4076189</advertiser-id>
        </link>
    </links>
</cj-api>
"""

root = BeautifulSoup(content, "html.parser")
print root

output:
------------------------------------------
<?xml version="1.0" encoding="UTF-8"?>
<cj-api>
<links page-number="1" records-returned="10" total-matched="32">
<link/>
<advertiser-id>4076189</advertiser-id>
</links>
</cj-api>
------------------------------------------

why <link> is changed to <link/> and </link> disappear?
I test both on mac and ubuntu with html.parser and lxml.
The results are same.

[Leonard Richardson (leonardr)]

All three of the parsers you tried are HTML parsers, and they all treated the <link> tag as an HTML link tag which is not allowed to contain subtags.

The markup you have is XML. Parsing it as XML preserves the <advertiser-id> tag within the <link> tag.

Script:

content = """<?xml version="1.0" encoding="UTF-8"?>
<cj-api>
    <links total-matched="32" records-returned="10" page-number="1">
        <link>
            <advertiser-id>4076189</advertiser-id>
        </link>
    </links>
</cj-api>
"""
from bs4 import BeautifulSoup
print BeautifulSoup(content, "xml")

output:

<?xml version="1.0" encoding="utf-8"?>
<cj-api>
<links page-number="1" records-returned="10" total-matched="32">
<link>
<advertiser-id>4076189</advertiser-id>
</link>
</links>
</cj-api>