summaryrefslogtreecommitdiff
path: root/academic/abella/README
blob: 7954d1e75158794317925898e12f97c125a76a73 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
Abella is an interactive theorem prover based on lambda-tree syntax.

This means that Abella is well-suited for reasoning about the
meta-theory of programming languages and other logical systems
which manipulate objects with binding. For example, the following
applications are included in the distribution of Abella.

* Various results on the lambda calculus involving big-step
  evaluation, small-step evaluation, and typing judgments
* Cut-admissibility for a sequent calculus
* Part 1a and Part 2a of the POPLmark challenge
* Takahashi's proof of the Church-Rosser theorem
* Tait's logical relations argument for weak normalization of the
  simply-typed lambda calculus
* Girard's proof of strong normalization of the simply-typed lambda
  calculus
* Some ?-calculus meta-theory
* Relation between ?-reduction and paths in A-calculus

For Full List:
http://abella-prover.org/examples/index.html

Abella uses a two-level logic approach to reasoning. Specifications
are made in the logic of second-order hereditary Harrop formulas using
lambda-tree syntax. This logic is executable and is a subset of the
AProlog language (see the Teyjus system for an implementation of this
language).

The reasoning logic of Abella is the culmination of a series
of extensions to proof theory for the treatment of definitions,
lambda-tree syntax, and generic judgments. The reasoning logic of
Abella is able to encode the semantics of our specification logic as a
definition and thereby reason over specifications in that logic.