GHSA-2m96-52r3-2f3g

fugit parse and parse_nat stall on lengthy input

En bref

Publiée
Mise à jour
CVE
CVE-2024-43380
CVSS_V3
CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:L
CVSS_V4
CVSS:4.0/AV:N/AC:L/AT:N/PR:N/UI:P/VC:N/VI:N/VA:L/SC:N/SI:N/SA:N

Description

### Impact The fugit "natural" parser, that turns "every wednesday at 5pm" into "0 17 * * 3", accepted any length of input and went on attempting to parse it, not returning promptly, as expected. The parse call could hold the thread with no end in sight. Fugit dependents that do not check (user) input length for plausability are impacted. ### Patches Problem was reported in #104 and the fix was released in [fugit 1.11.1](https://rubygems.org/gems/fugit/versions/1.11.1) ### Workarounds By making sure that `Fugit.parse(s)`, `Fugit.do_parse(s)`, `Fugit.parse_nat(s)`, `Fugit.do_parse_nat(s)`, `Fugit::Nat.parse(s)`, and `Fugit::Nat.do_parse(s)` are not fed strings too long. 1000 chars feels ok, while 10_000 chars makes it stall. In fewer words, making sure those fugit methods are not fed unvetted input strings. ### References gh-104

Paquets affectés

PaquetVersions
RubyGems · fugitintroduced 0, fixed 1.11.1

Références

Sources

SourceLicence
OSV.devCC-BY 4.0 — GitHub Advisory Database