Advisory

PYSEC-2026-3833

Also known as: CVE-2026-55619, GHSA-m66c-fw79-6359

Summary

eml_parser has parser DoS via deeply nested parentheses in e-mail headers

Details

Summary

eml_parser uses the email.utils.getaddresses() function from the CPython standard library to parse e-mail headers that contain e-mail addresses (such as To, Cc, Bcc, From, Reply-To, Sender, ...). When the input header contains a deeply nested CFWS (comment / folding white space) construct, the recursive descent parser in the standard library exhausts the call stack. The resulting RecursionError is not caught by eml_parser, so the exception propagates and aborts parsing of the whole message.

Impact

SOC pipelines use eml_parser to process untrusted e-mails. An attacker can easily create an eml file that will trigger the RecursionError during parsing.

The impact is mitigated by the fact that there are various other situations in which eml_parser will raise an exception when attempting to parse a malformed or pathological eml file. In particular, very deeply nested multipart e-mails also result in a RecursionError being raised by the library voluntarily. Therefore, systems relying on eml_parser already need to detect and handle errors emanating from the library in an appropriate way.

Workarounds

The issue can be avoided by wrapping the call to eml_parser.decode_email or eml_parser.decode_email_bytes in a try/except construct.

Patches

Since version 3.0.2, eml_parser will catch the error in the standard library parser and fall back to a simpler parser based on a regular expression.

Severity

Medium Basis: CVSS v3 score 5.3, vector CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:L

Dates

Published
2026-09-10 09:44 UTC
Modified
2026-09-10 12:15 UTC

Affected packages and fixes

eml-parser

PyPI · All systems

Affected

  • introduced 0, fixed 3.0.2 (ecosystem)

Fixed versions

3.0.2

Fix

  • Fixed version 3.0.2
    pip install --upgrade eml-parser==3.0.2

Fix commands are the export's own generic commands. Ojuju never runs them. Check them against your setup first.

Check your own computer: guard packages check

View this record on OSV.dev

Source

Source: PyPA Advisory Database · CC BY 4.0

Sources & licences