CVE-2022-39353

[david-nikolai-mueller]

Textract currently utilizes an outdated version – 0.1.27 – of xmldom (npm), which is susceptible to the security vulnerability outlined in CVE-2022-39353.

Impact

xmldom parses XML that is not well-formed because it contains multiple top level elements, and adds all root nodes to the childNodes collection of the Document, without reporting any error or throwing.
This breaks the assumption that there is only a single root node in the tree, which led to https://nvd.nist.gov/vuln/detail/CVE-2022-39299 and is a potential issue for dependents.

Patches

Update to @xmldom/xmldom@~0.7.7, @xmldom/xmldom@~0.8.4 (dist-tag latest) or @xmldom/xmldom@>=0.9.0-beta.4 (dist-tag next).

Workarounds

One of the following approaches might help, depending on your use case:

• Instead of searching for elements in the whole DOM, only search in the documentElement.
• Reject a document with a document that has more then 1 childNode.

References

• https://nvd.nist.gov/vuln/detail/CVE-2022-39299
• Missing error for XML documents with multiple root element nodes jindw/xmldom#150