Skip to content

xml.dom.minidom: DOMBuilderFilter accept/reject is inverted for notation declarations #152142

Description

@tonghuaroot

ExpatBuilder.notation_decl_handler in Lib/xml/dom/expatbuilder.py drops a notation when the DOMBuilderFilter accepts it and keeps it when the filter rejects it, the opposite of every other filtered handler in that file and of the DOMBuilderFilter contract (FILTER_REJECT drops the node, FILTER_ACCEPT keeps it).

import io
import xml.dom.expatbuilder as eb
from xml.dom.xmlbuilder import DOMBuilderFilter, Options

src = b'<?xml version="1.0"?><!DOCTYPE root [<!NOTATION gif PUBLIC "image/gif">]><root/>'

class RejectNotations(DOMBuilderFilter):
    def acceptNode(self, node):
        if node.nodeType == node.NOTATION_NODE:
            return self.FILTER_REJECT
        return self.FILTER_ACCEPT
    def startContainer(self, node):
        return self.FILTER_ACCEPT

opts = Options()
opts.filter = RejectNotations()
doc = eb.makeBuilder(opts).parseFile(io.BytesIO(src))
print(doc.doctype.notations.length)   # 1 -- the filter rejected it but it was kept

The handler compares against FILTER_ACCEPT:

self.document.doctype.notations._seq.append(node)
if self._filter and self._filter.acceptNode(node) == FILTER_ACCEPT:
    del self.document.doctype.notations._seq[-1]

while the sibling entity_decl_handler (and the element, comment and processing-instruction handlers) compare against FILTER_REJECT. So an accepted notation is dropped and a rejected one is kept.

Reproduces on main and on released 3.x. Lib/test/test_minidom.py has no DOMBuilderFilter coverage for notations.

Linked PRs

Metadata

Metadata

Assignees

No one assigned

    Labels

    stdlibStandard Library Python modules in the Lib/ directorytopic-XMLtype-bugAn unexpected behavior, bug, or error
    No fields configured for issues without a type.

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions