Having Fun With 31.521 Shell Scripts
Résumé
Statically parsing shell scripts is, due to various peculiarities of the shell language, a challenge. One of the difficulties is that the shell language is designed to be executed by intertwining reading chunks of syntax with semantic actions. We have analyzed a corpus of 31.521 POSIX shell scripts occurring as maintainer scripts in the Debian GNU/Linux distribution. Our parser, which makes use of recent developments in parser generation technology, succeeds on 99.9% of the corpus. The architecture of our tool allows us to easily plug in various statistical analyzers on the syntax trees constructed from the shell scripts. The statistics obtained by our tool are the basis for the definition of a model which we plan to use in the future for the formal verification of scripts.
Domaines
Logique en informatique [cs.LO]Origine | Fichiers produits par l'(les) auteur(s) |
---|
Loading...