Python处理XML时NS1命名空间前缀残留问题求助

Hey Lucas, let’s fix that pesky NS1 namespace prefix that’s sticking around even after you removed NS0 with ET.register_namespace. I’ve dealt with this exact headache when working with NFe XMLs (they’re notoriously finicky with namespaces), so here’s how to get rid of it:

Step 1: Identify what NS1 actually maps to #

ElementTree assigns NS0, NS1, etc. to any namespaces you haven’t explicitly registered. First, let’s find out which URI is tied to NS1. Run this quick snippet on your original XML:

import xml.etree.ElementTree as ET

tree = ET.parse("your_input_nfe.xml")
root = tree.getroot()
print(root.nsmap)

This will print a dictionary of all namespaces in your XML. Look for an entry like 'ns1': 'http://some-namespace-uri.com' — that’s the URI we need to handle. For NFe files, this is often a secondary namespace like http://www.portalfiscal.inf.br/nfe/xsd or a related schema URI.

Step 2: Register all required namespaces #

You already registered the default NFe namespace, but you need to register the NS1 URI too. If the URI should use a specific prefix (like xsd for schema-related tags), register it with that prefix. If it’s supposed to be part of the default namespace (unlikely, since default can only have one), double-check your XML structure.

Example registration:

# Default NFe namespace (you already have this)
ET.register_namespace('', "http://www.portalfiscal.inf.br/nfe")
# Register the NS1 namespace with a meaningful prefix (adjust the URI to match your root.nsmap output)
ET.register_namespace('xsd', "http://www.portalfiscal.inf.br/nfe/xsd")

If you don’t want a prefix for this namespace (and it’s allowed by your XML schema), you can’t set two defaults — so you’ll need to ensure all elements using this URI are properly qualified with the namespace when you create/modify them.

Step 3: Fix element creation/modification #

The most common reason NS1 pops up is when you create new elements without specifying their correct namespace. Instead of creating elements like ET.Element("new_tag"), you need to qualify them with the namespace URI.

For example:

# Get the namespace URIs you need
ns_default = "http://www.portalfiscal.inf.br/nfe"
ns_xsd = "http://www.portalfiscal.inf.br/nfe/xsd"

# Create an element in the default namespace
new_default_elem = ET.Element(f"{{{ns_default}}}new_tag")
root.append(new_default_elem)

# Create an element in the previously NS1 namespace
new_xsd_elem = ET.Element(f"{{{ns_xsd}}}schema_tag")
root.append(new_xsd_elem)

This tells ElementTree exactly which namespace each element belongs to, so it won’t assign random NS1 prefixes.

Step 4: Write the XML correctly #

When saving your modified XML, use these parameters to avoid unexpected issues:

tree.write("output_nfe.xml", encoding="utf-8", xml_declaration=True)

Avoid adding the default_namespace parameter here — since you’ve already registered the default namespace, ElementTree will handle it properly.

If NS1 still won’t go away... #

Check for elements in your XML that might have leftover namespace declarations. Sometimes, old xmlns:ns1="..." attributes stick around in child elements. You can clean these up by iterating through all elements and removing any unwanted namespace attributes:

for elem in root.iter():
    # Remove any ns1 declarations (adjust the attribute name to match your case)
    if 'xmlns:ns1' in elem.attrib:
        del elem.attrib['xmlns:ns1']

That should do it! NFe XMLs are strict with namespaces, but once you register all URIs and qualify your elements correctly, those random NS prefixes will disappear.

内容的提问来源于stack exchange，提问作者Lucas Monteiro