Skip to content

xml.dom.pulldom.DOMEventStream leaks file handles opened by parse() #148428

Description

@WYSIATI

pulldom.parse() opens a file when called with a filename string, but the resulting DOMEventStream never closes it. The class has no close() method, no context manager support, and clear() just sets self.stream = None without closing.

This produces ResourceWarning: unclosed file and can lead to file descriptor exhaustion in long-running processes.

The documentation example at https://docs.python.org/3/library/xml.dom.pulldom.html#module-xml.dom.pulldom shows:

doc = pulldom.parse('sales_items.xml')
for event, node in doc:
    ...

— with no cleanup, so users following the docs will hit this.

Reproduction

import warnings, tempfile, os, gc
from xml.dom import pulldom

warnings.simplefilter('always', ResourceWarning)

fd, path = tempfile.mkstemp(suffix='.xml')
os.write(fd, b'<root><item>test</item></root>')
os.close(fd)

events = pulldom.parse(path)
for event, node in events:
    pass

# stream is still open after full iteration
print(f'stream.closed: {events.stream.closed}')  # False

# clear() doesn't close it either
events.clear()

del events
gc.collect()  # ResourceWarning: unclosed file ...

os.unlink(path)

The existing test in test_pulldom.py already works around this with self.addCleanup(handler.stream.close) (line 36), which further confirms the leak.

Suggested fix

Track whether parse() opened the file (vs. receiving a user-provided stream), then add close(), __enter__/__exit__, and __del__ with ResourceWarning to DOMEventStream. Update clear() to close owned streams.

This follows the same approach used for ElementTree.iterparse() (gh-140601).

Linked PRs

Activity

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    stdlibStandard Library Python modules in the Lib/ directorytopic-XMLtype-bugAn unexpected behavior, bug, or error

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions