NLTK: WekaClassifier passes an unvalidated caller-controlled model path to the weka JVM, bypassing nltk.pathsec containment

影響: pip:nltk <= 3.10.3 → 3.10.4

WekaClassifier in NLTK passes a caller-controlled model file path to the weka Java VM without any nltk.pathsec validation. WekaClassifier.train puts the caller's model_filename verbatim into the weka argv ["-d", model_filename, ...] (weka writes the trained model there), and WekaClassifier._classify_many puts self._model into ["-l", self._model, ...] (weka reads it). Under nltk.pathsec.ENFORCE=True (the default on current HEAD), NLTK's stated invariant — established by advisory CVE-2026-81726 /…

Read on GitHub Advisory Database

On 10/27/2026, 11:50 PM, GitHub Advisory Database published "NLTK: WekaClassifier passes an unvalidated caller-controlled model path to the weka JVM, bypassing nltk.pathsec containment" (type: Advisory). Severity is "High" (high). Related technologies: Python.