The flaw in pypdf is triggered when the parse_bfrange routine processes unusually large source‑code or destination‑string tokens in a font /ToUnicode CMap while extracting text. A crafted PDF can cause pypdf to allocate a large amount of memory, leading to process instability or termination. This represents a classic CWE‑400 problem and a CWE‑770 vulnerability that results in a denial‑of‑service condition for the application reading the PDF.

The issue affects the pypdf library provided by py‑pdf. Versions prior to 6.15.0 are vulnerable; upgrading to 6.15.0 or later resolves the problem.

The CVSS score of 4.8 indicates moderate severity, and no EPSS data is available, so the historical exploitation likelihood is unclear. Based on the description, it is inferred that any PDF that is parsed by the affected library could trigger the vulnerability, implying a local or user‑controlled attack vector. Although not listed in KEV, the absence of a known public exploit does not reduce the risk of an attacker sending a malicious PDF to a service that uses pypdf.