When multiple prefixes are defined for a single namespace URI, a _SaxParserTarget can't know which prefix was originally used for a given element

[Leonard Richardson]

This report comes from a bug reported against my project, Beautiful Soup: https://bugs.launchpad.net/beautifulsoup/+bug/1915583

Here's the output of running the attached script:

===
Python : sys.version_info(major=3, minor=8, micro=5, releaselevel='final', serial=0)
lxml.etree : (4, 6, 2, 0)
libxml used : (2, 9, 10)
libxml compiled : (2, 9, 10)
libxslt used : (1, 1, 34)
libxslt compiled : (1, 1, 34)

lxml
<package xmlns:opf="http://www.idpf.org/2007/opf" xmlns:opf2="http://www.idpf.org/2007/opf" xmlns:dc="http://purl.org/dc/elements/1.1/">
  <metadata>
    <dc:identifier opf:scheme="URI">a</dc:identifier>
    <dc:identifier opf2:scheme="URI">b</dc:identifier>
  </metadata>
</package>

--------------------------------------------------------------------------------
Beautiful Soup
<?xml version="1.0" encoding="utf-8"?>
<package xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:opf="http://www.idpf.org/2007/opf" xmlns:opf2="http://www.idpf.org/2007/opf">
<metadata>
<dc:identifier opf2:scheme="URI">a</dc:identifier>
<dc:identifier opf2:scheme="URI">b</dc:identifier>
</metadata>
</package>
===

The markup in the attached script defines two different prefixes for the namespace URI "http://www.idpf.org/2007/opf": "opf" and "opf2". When the markup is parsed through lxml, the original prefix of each tag and attribute is preserved. When the markup is parsed through Beautiful Soup, using lxml's _SaxParserTarget interface, this isn't always possible.

The "_SaxParserTarget._handleSaxStart" method is passed a namespace-qualified tag name, a dictionary mapping namespace-qualified attribute names to their values, and a dictionary mapping namespace prefixes to the corresponding URIs. Given this information, when multiple prefixes are defined for a given URI, it's not possible to determine which prefix was used in the original markup for a given tag or attribute name.

In this case, lxml gives me the attribute name '{http://www.idpf.org/2007/opf}scheme' and I have no way of knowing whether the original prefix was "opf" or "opf2"; I just have to pick one.

I don't consider this a serious problem, but I wanted to bring it to your attention; maybe I've missed something in the _SaxParserTarget interface that would make it an easy fix.

[scoder (scoder)]

Thanks for the report. I don't think there is an easy way to improve this. The interface uses ElementTree's qualified tag names. It's intentional to resolve the prefixes here, in order to make it _easier_ for users to deal with namespaces. There is obviously a loss of parser available data in doing this, but since prefixes are not part of the XML information set, it's not a loss in the document information. Only round-trips suffer from this issue.

So, I do admit that it's an issue. But it seems a very rare issue and it isn't easy to do something about.
Sounds like a "won't fix".