1 \documentclass{llncs}

     2 \usepackage{times}

     3 \usepackage{isabelle}

     4 \usepackage{isabellesym}

     5 \usepackage{amsmath}

     6 \usepackage{amssymb}

     7 

     8 

     9 \usepackage{pdfsetup}

    10 \urlstyle{rm}

    11 \isabellestyle{it}

    12 \renewcommand{\isastyle}{\isastyleminor}

    13 \renewcommand{\isacharunderscore}{\mbox{$\_\!\_$}}

    14 \renewcommand{\isasymbullet}{{\raisebox{-0.4mm}{\Large$\boldsymbol{\cdot}$}}}

    15 \def\dn{\,\stackrel{\mbox{\scriptsize def}}{=}\,}

    16 \renewcommand{\isasymequiv}{$\dn$}

    17 \renewcommand{\isasymiota}{}

    18 \renewcommand{\isasymrightleftharpoons}{}

    19 \renewcommand{\isasymemptyset}{$\varnothing$}

    20 

    21 \newcommand{\numbered}[1]{\refstepcounter{equation}{\rm(\arabic{equation})}\label{#1}}

    22 

    23 

    24 \begin{document}

    25 

    26 \title{Proof Pearl: A New Foundation for Nominal Isabelle}

    27 \author{Brian Huffman\inst{1} and Christian Urban\inst{2}}

    28 \institute{Portland State University \and Technical University of Munich}

    29 \maketitle

    30 

    31 \begin{abstract}

    32 Pitts et al introduced a beautiful theory about names and binding based on the

    33 notions of permutation and support. The engineering challenge is to smoothly

    34 adapt this theory to a theorem prover environment, in our case Isabelle/HOL.

    35 We present a formalisation of this work that differs from our earlier approach

    36 in two important respects: First, instead of representing permutations as

    37 lists of pairs of atoms, we now use a more abstract representation based on

    38 functions.  Second, whereas the earlier work modeled different sorts of atoms

    39 using different types, we now introduce a unified atom type that includes all

    40 sorts of atoms. Interestingly, we allow swappings, that is permutations build from

    41 two atoms, to be ill-sorted.  As a result of these design changes, we can iron

    42 out inconveniences for the user, considerably simplify proofs and also

    43 drastically reduce the amount of custom ML-code. Furthermore we can extend the

    44 capabilities of Nominal Isabelle to deal with variables that carry additional

    45 information. We end up with a pleasing and formalised theory of permutations

    46 and support, on which we can build an improved and more powerful version of

    47 Nominal Isabelle.

    48 \end{abstract}

    49 

    50 % generated text of all theories

    51 \input{session}

    52 

    53 % optional bibliography

    54 \bibliographystyle{abbrv}

    55 \bibliography{root}

    56 

    57 \end{document}

    58 

    59 %%% Local Variables:

    60 %%% mode: latex

    61 %%% TeX-master: t

    62 %%% End: