General Description
POULE is a prover
for linear nested sequent calculi
[Lm15]
for classical modal
logics. It implements several systems specified in
[LP15].
Syntax of Formulas
Formulas in POULE can be written as follows
(Object level formulas) F ::= A | F cand F | F cor F | F cimp F | box F | hearts F.
(Labelled formulas) L ::= r X F | l X F.
(Linear logic formulas) M := L | M @ M | M , M | M & M | pi X\ M | sigma X\ M | top.
References
[Lm15]
|
Björn Lellmann. Linear Nested Sequents, 2-Sequents and
Hypersequents, accepted for publication, TABLEAUX 2015,
pdf.
|
[LP15]
|
Björn Lellmann and Elaine Pimentel. Proof search in Nested Sequents,
submitted to LPAR2015, pdf.
|
Dependencies
POULE requires the
Teyjus system, an implementation of
Lambda Prolog, a higher-order logic programming language.
For runnig POULE, you have to:
- Download the files llf.mod and
llf.sig
- give the command tjcc llf && tjlink llf && tjsim llf
- The quest has to have the form
llf N TH M.
where N is the maximum depth
of the proof (100 is usually enough),
TH is the theory (for now either g3cp, k, kts4 or ec) and M is the
linear logic clause correspondent to the object
level formula to be proven. Observe that the object level formula has to be
labelled with label "x".
Here there are some simple examples of queries:
System G3cp
- llf 100 g3cp (r x (a cimp a)).
- llf 100 g3cp (r x ((a cand b) cimp (a cand b))).
- llf 100 g3cp (r x ((a cand b) cimp a)).
- llf 100 g3cp (r x ((a cand b) cimp (a cor b))).
- llf 100 g3cp (r x ((a cor b) cimp (a cand b))).
- llf 100 g3cp (r x ((a cand (a cimp b)) cimp b)).
System K
- llf 100 k (r x ((box a) cimp (box a))).
- llf 100 k (r x (a cimp (box a))).
- llf 100 k (r x ((box a) cimp a)).
- llf 100 k (r x ((box (a cimp b)) cimp ((box a) cimp (box b)))).
System KTS4
- llf 100 kts4 (r x ((box a) cimp (box a))).
- llf 100 kts4 (r x ((hearts a) cimp (hearts a))).
- llf 100 kts4 (r x ((hearts a) cimp (box a))).
- llf 100 kts4 (r x ((box a) cimp (hearts a))).
And here there are some examples of theories, already included in llf's code (that is, you don't have to download them):
Available soon.
We are currently working on a parser for generating automatically LL
specifications out of inference figures.
This is a joint work with
Carlos Olarte.