The NLTK library's nltk.featstruct.FeatStructReader uses a recursive-descent parser to parse feature-structure strings without enforcing a maximum nesting depth. Deeply nested inputs cause the parser to exceed Python's recursion limit, raising an unhandled RecursionError instead of the library's usual catchable exceptions. This uncontrolled recursion leads to denial of service by crashing applications that parse user-supplied feature-structure or grammar text. Similar protections exist in other parts of NLTK (e.g., LogicParser and JSONTaggedDecoder) but are missing in FeatStructReader. The vulnerability affects both direct use of FeatStruct and grammar parsing via FeatureGrammar.fromstring(). A proof of concept demonstrates crashing the parser with a small (~700 bytes) crafted input at nesting depth 167.

Applications using NLTK's FeatStructReader to parse feature-structure or grammar strings can be crashed by unauthenticated attackers supplying deeply nested input. This results in denial of service due to an unhandled RecursionError. There is no impact on confidentiality or integrity, and no code execution or memory corruption is involved.

A patch is available to address this vulnerability by adding recursion depth limits similar to other NLTK parsers. Users should apply the official fix to prevent unhandled RecursionError crashes. Until patched, avoid parsing untrusted or deeply nested feature-structure inputs. No additional mitigations are required beyond applying the patch.