|
|
@ -0,0 +1,426 @@ |
|
|
|
% mangletex (11 May 1992) run at 22:09 GMT Monday 08 November 1993 |
|
|
|
%%\title {\TeX\ macros for proof boxes} |
|
|
|
%%\author {Paul Taylor\\ |
|
|
|
%% Department of Computing,\\ |
|
|
|
%% Imperial College,\\ |
|
|
|
%% London SW7 2BZ\\ |
|
|
|
%% +44 71 589 5111 {\em ext.} 5057\\ |
|
|
|
%% {\tt<pt@doc.ic.ac.uk>}} |
|
|
|
%%\date{8 November 1993} |
|
|
|
%% % to get the provisional user documentation do |
|
|
|
%% % grep ^%% boxproof.tex | sed 's/^%*//' > boxproof-manual.tex |
|
|
|
%%\documentstyle{article} |
|
|
|
%%\input boxproof.tex |
|
|
|
%%\def\meta#1{\mbox{$\langle\hbox{#1}\rangle$}} |
|
|
|
%%\def\macrowitharg#1#2{{\tt\string#1\bra\meta{#2}\ket}} |
|
|
|
%%{\escapechar-1 \xdef\bra{\string\{}\xdef\ket{\string\}}} |
|
|
|
%%\let\subsection\section |
|
|
|
%%\begin{document} |
|
|
|
%%\maketitle |
|
|
|
%% |
|
|
|
%% \subsection{Introduction} |
|
|
|
%% The proof |
|
|
|
%%\begin{proofbox} |
|
|
|
%% \(\"1"\:\alpha\leftrightarrow\psi(x,\top)\=\\ |
|
|
|
%% \:\Some\beta.\psi(x,\beta)\=\mathsf{total}\\ |
|
|
|
%% \[\exists\beta\kern-1em\:\psi(x,\beta)\\ |
|
|
|
%% \(\:\beta\=\\ |
|
|
|
%% \:\beta=\top\=(*)\\ |
|
|
|
%% \:\psi(x,\!\top)\=\mathsf{subs}\\ |
|
|
|
%% \:\alpha\=\elim\leftrightarrow(\ref{1})\\ |
|
|
|
%% \*\:\alpha\=\\ |
|
|
|
%% \:\psi(x,\!\top)\=\elim\leftrightarrow(\ref{1})\\ |
|
|
|
%% \:\beta=\top\=\mathsf{func}\\ |
|
|
|
%% \:\beta\=(*)\\ |
|
|
|
%% \)\:\alpha=\beta\=\intro\leftrightarrow\\ |
|
|
|
%% \:\psi(x,\alpha)\=\mathsf{subs}\\ |
|
|
|
%% \] \:\psi(x,\alpha)\=\elim\exists\\ |
|
|
|
%% \* \:\psi(x,\alpha)\=\\ |
|
|
|
%% \(\:\alpha\=\\ |
|
|
|
%% \:\alpha=\top\=(*)\\ |
|
|
|
%% \:\phi(x,\!\top)\=\mathsf{subs}\\ |
|
|
|
%% \*\:\psi(x,\!\top)\=\\ |
|
|
|
%% \:\alpha=\top\=\mathsf{func}\\ |
|
|
|
%% \:\alpha\=(*)\\ |
|
|
|
%% \)\:\alpha\leftrightarrow\psi(x,\!\top)\=\intro\leftrightarrow\\ |
|
|
|
%%% \) \:\psi(x,\alpha)\leftrightarrow(\alpha\leftrightarrow\psi(x,\!\top)) |
|
|
|
%% \=\intro\leftrightarrow\\ |
|
|
|
%%\end{proofbox} |
|
|
|
%% is produced by |
|
|
|
%%\begin{verbatim} |
|
|
|
%%\begin{proofbox} |
|
|
|
%% \(\"1"\:\alpha\leftrightarrow\psi(x,\top)\\ |
|
|
|
%% \:\Some\beta.\psi(x,\beta) \=\mathsf{total}\\ |
|
|
|
%% \[\exists\beta\kern-1em\:\psi(x,\beta)\\ |
|
|
|
%% \(\:\beta\\ |
|
|
|
%% \:\beta=\top \=(*)\\ |
|
|
|
%% \:\psi(x,\!\top) \=\mathsf{subs}\\ |
|
|
|
%% \:\alpha \=\elim\leftrightarrow(\ref{1})\\ |
|
|
|
%% \*\:\alpha\\ |
|
|
|
%% \:\psi(x,\!\top) \=\elim\leftrightarrow(\ref{1})\\ |
|
|
|
%% \:\beta=\top \=\mathsf{func}\\ |
|
|
|
%% \:\beta\=(*)\\ |
|
|
|
%% \)\:\alpha=\beta \=\intro\leftrightarrow\\ |
|
|
|
%% \:\psi(x,\alpha) \=\mathsf{subs}\\ |
|
|
|
%% \] \:\psi(x,\alpha) \=\elim\exists\\ |
|
|
|
%% \* \:\psi(x,\alpha)\\ |
|
|
|
%% \(\:\alpha\\ |
|
|
|
%% \:\alpha=\top \=(*)\\ |
|
|
|
%% \:\phi(x,\!\top) \=\mathsf{subs}\\ |
|
|
|
%% \*\:\psi(x,\!\top)\\ |
|
|
|
%% \:\alpha=\top \=\mathsf{func}\\ |
|
|
|
%% \:\alpha \=(*)\\ |
|
|
|
%% \)\:\alpha\leftrightarrow\psi(x,\!\top) \=\intro\leftrightarrow\\ |
|
|
|
%% \) \:\psi(x,\alpha)\leftrightarrow |
|
|
|
%% (\alpha\leftrightarrow\psi(x,\!\top)) \=\intro\leftrightarrow\\ |
|
|
|
%%\end{proofbox} |
|
|
|
%%\end{verbatim} |
|
|
|
%% |
|
|
|
%% Syntax as follows: |
|
|
|
%% each line is of the form |
|
|
|
%% \begin{center} |
|
|
|
%% \meta{variables} |
|
|
|
%% \meta{name} |
|
|
|
%% \verb/\:/ \meta{formula} |
|
|
|
%% \verb/\=/ \meta{reason} |
|
|
|
%% \verb/\-/ \meta{use} |
|
|
|
%% \verb/\\/ |
|
|
|
%% \end{center} |
|
|
|
%% where |
|
|
|
%% \begin{itemize} |
|
|
|
%% \item \meta{variables} is something like ``$x,y$'' --- |
|
|
|
%% it's for variables declared at the beginning of $\intro\forall$- and |
|
|
|
%% $\elim\exists$-boxes. |
|
|
|
%% \item \meta{name} is a command \verb/\label{fred}/ which |
|
|
|
%% defines \verb/fred/ to be the label text, which may be used anywhere |
|
|
|
%% as \verb/\ref{fred}/ --- see {\em The \LaTeX book.} |
|
|
|
%% Local labels are also available, using \macrowitharg\lbl{name} |
|
|
|
%% or \verb/\"/\meta{name}\verb/"/; these obey the scoping rules of |
|
|
|
%% the boxes. |
|
|
|
%% You may also refer to the previous line as \verb/\ref{-}/. |
|
|
|
%% \item \meta{formula} is the proposition being asserted. |
|
|
|
%% \item \meta{reason} is \verb/\intro\land(\ref{john},\ref{mary})/ |
|
|
|
%% or \verb/\elim\forall(\ref{jim})/. |
|
|
|
%% \item \meta{use} is provided for linear logic, to record the step |
|
|
|
%%% which uses this one. How this accords with theory I don't yet know. |
|
|
|
%% \end{itemize} |
|
|
|
%% Note that the parts are separated by \verb/\:/, \verb/\=/ and \verb/\\/; |
|
|
|
%% these correspond to |
|
|
|
%% \begin{center} |
|
|
|
%% {\bf let } \meta{name} = \meta{expression} : \meta{type} |
|
|
|
%% \end{center} |
|
|
|
%% in a declarative language. |
|
|
|
%% The \verb/\:/, \verb/\=/ and \verb/\-/ fields are optional and may occur |
|
|
|
%% in any order. If any of them is repeated the last is taken. |
|
|
|
%% If none of them is present the \meta{variables} field is also ignored. |
|
|
|
%% |
|
|
|
%% Proof {\em boxes\/} are ``wrapped up'' as follows: |
|
|
|
%% \begin{itemize} |
|
|
|
%%% \item the whole proof in \verb/\begin{proofbox}/...\verb/\end{proofbox}/; |
|
|
|
%%% \item single-column boxes ($\intro\forall$, $\intro\imp$, $\elim\exists$), |
|
|
|
%% in \verb/\[/...\verb/\]/. |
|
|
|
%% \item multiple-column boxes are of two kinds: |
|
|
|
%% \begin{itemize} |
|
|
|
%%% \item separate ($\intro\land$) boxes: \verb/\(/...\verb/\*/...\verb/\)/. |
|
|
|
%% \item stuck together ($\elim\lor$) boxes: |
|
|
|
%% \verb/\(/...\verb/\+/...\verb/\)/. |
|
|
|
%% \end{itemize} |
|
|
|
%% \end{itemize} |
|
|
|
%% You may put more than two columns in \verb/\(/...\verb/\)/ and even mix |
|
|
|
%% the \verb/\+/ and \verb/\*/ separators. |
|
|
|
%% |
|
|
|
%% The whole proof is enclosed in \verb/\proofbox/...\verb/\endproofbox/ |
|
|
|
%% or \verb/\begin{proofbox}/...\verb/\end{proofbox}/, but the \LaTeX\ |
|
|
|
%% environment form {\em must not\/} be used for nested boxes. |
|
|
|
%% |
|
|
|
%% If the proof occurs in paragraph mode (ie in vertical or |
|
|
|
%% unrestricted horizontal mode) then it is set as a display, using the |
|
|
|
%% full width of the page. Otherwise it uses only the required width. |
|
|
|
%% |
|
|
|
%% A lot of the internals are potentially configurable, but there is not |
|
|
|
%% yet a user interface suitable for doing this. This will be provided |
|
|
|
%% in the next version. |
|
|
|
%% |
|
|
|
%% This is a {\em prototype\/} proof-box macro package; |
|
|
|
%% the syntax and implementation may change. |
|
|
|
%% It is {\tt boxproof.tex} in the \TeX\ system |
|
|
|
%% (so just \verb/\input boxproof/). |
|
|
|
%% |
|
|
|
%% There are some examples in |
|
|
|
%% \verb+~pt/utilities/proofs/proofboxeg.tex+ and |
|
|
|
%% \verb+proofboxeg1.tex+. |
|
|
|
%% |
|
|
|
%% The prototype version was begun on 2 August 1991, the new one on |
|
|
|
%% 4 June 1993. |
|
|
|
%% |
|
|
|
%% |
|
|
|
\message{}\edef\proofboxrestoreat |
|
|
|
{\noexpand\catcode`\noexpand\@\the\catcode`\@}\catcode`\@=11 \count@\year |
|
|
|
\multiply\count@12 \advance\count@\month\ifnum\count@> 23999 \message{}% |
|
|
|
\message{}\fi\let\pb@A\+\let\+\relax\let\pb@B\(\let\pb@C\)\def\pb@D{\pb@B |
|
|
|
\let\)\pb@C}\def\pb@E{\pb@B\displaystyle\let\]\pb@C}%% |
|
|
|
%% The following are user-settable dimensions. |
|
|
|
%% |
|
|
|
\newdimen\prooflinenowidth\prooflinenowidth1em \newdimen\proofboxfullwidth |
|
|
|
\newdimen\proofboxleftmargin\proofboxleftmargin3em \newdimen |
|
|
|
\proofboxrightmargin\newdimen\proofboxrulebreadth\proofboxrulebreadth.4pt |
|
|
|
\newdimen\proofboxseparation\proofboxseparation.2em \newdimen |
|
|
|
\proofboxbaselineskip\proofboxbaselineskip4ex %% |
|
|
|
%% |
|
|
|
%% The following are user-accessible numbers. |
|
|
|
%% |
|
|
|
\newcount\proofcolumns\newcount\prooflinetotal\newcount\prooflineinbox%% |
|
|
|
%% |
|
|
|
%% These boxes receive the fields of the proof steps. |
|
|
|
%% They are used in \verb/\proofboxmakeleftcolumn/ {\em etc}. |
|
|
|
%% |
|
|
|
\newbox\proofboxproposition\newbox\proofboxreason\newbox\proofboxuse\newbox |
|
|
|
\proofboxvariables%% |
|
|
|
%% The left-most switch is used by |
|
|
|
%%% \verb/\proofboxmakelabel/ to suppress the label in all but the first column. |
|
|
|
%% On the outside it is true. |
|
|
|
\newif\ifleftmostproofbox\leftmostproofboxtrue%% |
|
|
|
%% |
|
|
|
%% The following are obsolete: they are only included for reverse |
|
|
|
%% compatibility with the prototype version. |
|
|
|
\newdimen\proofboxintercol\newdimen\proofboxsurround\newdimen |
|
|
|
\proofboxformulawidth\newdimen\proofboxmargin\newskip\proofboxlefttabskip |
|
|
|
\newskip\proofboxrighttabskip%% |
|
|
|
%% |
|
|
|
%% \subsection{Redefinable macros} |
|
|
|
%% WARNING: most of these commands will be hidden and replaced with |
|
|
|
%% optional arguments to \verb/\proofbox/ in a future version. |
|
|
|
%% Do not rely on them. |
|
|
|
%% |
|
|
|
%% We provide three different ways of numbering the lines of the proof: |
|
|
|
%% \begin{itemize} |
|
|
|
%% \item \verb/\runningproofline/: a global running sequence (default), |
|
|
|
%% |
|
|
|
\def\runningproofline{\number\prooflinetotal}%% |
|
|
|
%% \item \verb/\nestedproofline/: a hierarchical system with dots, |
|
|
|
%% |
|
|
|
\def\nestedproofline{\relax\ifx\enclosingproofline\empty\else |
|
|
|
\enclosingproofline{.}\fi\number\prooflineinbox}\def\enclosingproofline{}%% |
|
|
|
%% \item \verb/\nestedproofline/: a fully hierarchical system which |
|
|
|
%% also includes the column number |
|
|
|
%% (\verb/\proof@columns/) as a letter (ASCII quote plus number). |
|
|
|
%% |
|
|
|
\def\proofboxcolumn{\ifnum\proofcolumns>\z@{\count@96\advance\count@ |
|
|
|
\proofcolumns\char\count@}\fi}\def\fullynestedproofline{\relax\ifx |
|
|
|
\enclosingproofline\empty\enclosingproofline\proofboxcolumn.\else |
|
|
|
\proofboxcolumn\fi\number\prooflineinbox}%% |
|
|
|
%% \end{itemize} |
|
|
|
%% \verb/\theproofline/ is the default. |
|
|
|
%% |
|
|
|
\let\theproofline\runningproofline%% |
|
|
|
%% The macro \verb/\proofboxmakelabel#1/ is used to print the line label. |
|
|
|
%% We only put it in the leftmost box. |
|
|
|
%% It is printed in small non-ranging Arabic numerals |
|
|
|
%% ({\the\scriptfont1 0123456789}). |
|
|
|
%% Right-justify it in \verb/\prooflinenowidth/ if it will fit, |
|
|
|
%% otherwise let it stick out on the right, {\em i.e.}~left-justify it. |
|
|
|
%% |
|
|
|
\def\proofboxmakelabel#1{{\relax\ifleftmostproofbox\setbox\z@\hbox{\hss\the |
|
|
|
\scriptfont\@ne#1}\ifdim\wd\z@<\prooflinenowidth\setbox\z@\hbox to% |
|
|
|
\prooflinenowidth{\unhbox\z@}\fi\box\z@\quad\fi}}%% |
|
|
|
%% Kill the numbers altogether with \verb/\proofboxnonumbers/. |
|
|
|
%% |
|
|
|
\def\proofboxnonumbers{\def\proofboxmakelabel##1{}}%% |
|
|
|
%% How to make the left column of the proof box: |
|
|
|
%% use the variables field, a space if necessary and the line label. |
|
|
|
%% |
|
|
|
\def\proofboxmakeleftcolumn{\kern\proofboxseparation\hfil\dimen@\wd |
|
|
|
\proofboxvariables\unhbox\proofboxvariables\ifdim\dimen@>\z@\quad\fi |
|
|
|
\proofboxmakelabel\@currentlabel}%% |
|
|
|
%% How to make the middle column of the proof box: |
|
|
|
%% left justify the formula field. |
|
|
|
%% |
|
|
|
\def\proofboxmakemiddlecolumn{\unhbox\proofboxproposition\hfil\quad}%% |
|
|
|
%% How to make the right column of the proof box: |
|
|
|
%% use the reason and use fields. |
|
|
|
%% |
|
|
|
\def\proofboxmakerightcolumn{\box\proofboxreason\box\proofboxuse\hfil\kern |
|
|
|
\proofboxseparation}%% |
|
|
|
%% |
|
|
|
%% Make the four edges of a rectangular box and the separator |
|
|
|
%% between \verb/\+/ columns. |
|
|
|
%% |
|
|
|
\def\proofboxleftside{\leaders\hrule width\proofboxrulebreadth\vfill}\def |
|
|
|
\proofboxrightside{\leaders\hrule width\proofboxrulebreadth\vfill}\def |
|
|
|
\proofboxorseparator{\leaders\hrule width\proofboxrulebreadth\vfill}\def |
|
|
|
\proofboxtopside{\leaders\vrule height\proofboxrulebreadth depth\z@\hfill}% |
|
|
|
\def\proofboxbottomside{\leaders\vrule height\proofboxrulebreadth depth\z@ |
|
|
|
\hfill}%% |
|
|
|
%% |
|
|
|
%% Use dotted lines: \verb/\dottedproofbox/. |
|
|
|
%% |
|
|
|
\def\dottedproofbox{\def\proofboxleftside{\leaders\pb@F\vfill} \def |
|
|
|
\proofboxrightside{\leaders\pb@F\vfill} \def\proofboxorseparator{\leaders |
|
|
|
\pb@F\vfill} \def\proofboxtopside{\leaders\pb@F\hfill}\def\proofboxbottomside |
|
|
|
{\leaders\pb@F\hfill}%% |
|
|
|
}\def\pb@F{\vbox to 5pt{\vss\hbox to 5pt{\hss.\hss}\vss}}%% |
|
|
|
%% Leave the boxes open at the bottom: \verb/\openproofbox/ |
|
|
|
%% |
|
|
|
\def\openproofbox{\def\proofboxbottomside{\hfill}} %% |
|
|
|
%% \subsection{Miscellaneous logical notations} |
|
|
|
%% |
|
|
|
%% Print the names of the introduction and elimination rules, for example: |
|
|
|
%% \begin{quote} |
|
|
|
%% \verb/\elim\forall/ $\elim\forall$ \qquad |
|
|
|
%% \verb/\intro\land/ $\intro\land$ \qquad |
|
|
|
%% \end{quote} |
|
|
|
\def\intro#1{{{#1}{\cal I}}}\def\elim#1{{{#1}{\cal E}}}%% |
|
|
|
%% Recall that in \TeX\ the logical connectives and quantifiers are called |
|
|
|
%% \begin{center} |
|
|
|
%% \verb/\lor/ $\lor$\quad |
|
|
|
%% \verb/\land/ $\land$\quad |
|
|
|
%% \verb/\lnot/ $\lnot$\quad |
|
|
|
%% \verb/\forall/ $\forall$\quad |
|
|
|
%% \verb/\exists/ $\exists$ |
|
|
|
%% \end{center} |
|
|
|
%% The following provide macros for the \verb/\implies/ {\em relation\/} |
|
|
|
%% and for the binary {\em operation\/} which yields the abstract |
|
|
|
%% \verb/\implic/ation between formulae. |
|
|
|
%% The point is that \TeX\ spaces them and breaks the lines differently: |
|
|
|
%% \begin{center} |
|
|
|
%% \verb/A\implies B/ $A\implies B$\quad{\em versus\/}\quad |
|
|
|
%% \verb/A\implic B/ $A\implic B$ |
|
|
|
%% \end{center} |
|
|
|
%% There are forward and reverse, single and Double versions. |
|
|
|
%% |
|
|
|
\mathchardef\implies="3221 \mathchardef\impliedby="3220 \mathchardef\Implies=% |
|
|
|
"3229 \mathchardef\Impliedby="3228 \mathchardef\implic="2221 \mathchardef |
|
|
|
\implicby="2220 \mathchardef\Implic="2229 \mathchardef\Implicby="2228 \let |
|
|
|
\imp\to%% |
|
|
|
%% Handle the spacing after a variable (and optionally its type) |
|
|
|
%% bound by a quantifier symbol. For example |
|
|
|
%% \begin{quote} |
|
|
|
%% \verb/\All x:X. \phi(x)/\quad prints as\quad $\All x:X.\phi(x)$\quad |
|
|
|
%% instead of\quad $\forall x:X.\phi(x)$ |
|
|
|
%% \end{quote} |
|
|
|
\def\Quantifier#1#2.{\pb@G{#1}#2::.}\def\pb@G#1#2:#3:#4.{{{#1}{#2}\def\next{#% |
|
|
|
3}\ifx\next\empty{.}\mkern4mu \else\mkern1mu{\colon}\mkern1mu{#3}\mkern.5mu{.% |
|
|
|
}\mkern6mu \fi}}% |
|
|
|
%% We provide some commonly used forms; \verb/\iota/ ($\iota$) is Russell's |
|
|
|
%% description operator and should really be inverted. |
|
|
|
\def\All{\Quantifier\forall}\def\Some{\Quantifier\exists}\def\Function{% |
|
|
|
\Quantifier\lambda}\def\Product{\Quantifier\Pi}\def\Sum{\Quantifier\Sigma}% |
|
|
|
\def\TheOne{\Quantifier\iota}\def\Least{\Quantifier\mu}\def\Greatest{% |
|
|
|
\Quantifier\nu}%% |
|
|
|
%% There are several notations for substitution. |
|
|
|
%% After writing $a[x:=b]$ throughout my book I~thought I~might change to |
|
|
|
%% $[b/x]^*a$. |
|
|
|
%% This macro reads the source in the first form and prints in the second. |
|
|
|
%% If you use it you can, like me, defer the decision about |
|
|
|
%% which notation to use until the final stages, doing |
|
|
|
%% \begin{quote} |
|
|
|
%% \verb/\renewcommand{\Subst}{\plainsubstitution}/ |
|
|
|
%% \end{quote} |
|
|
|
%% if you finally decide on making substitution act on the right. |
|
|
|
%% This is already an improvement on the literal text, because it |
|
|
|
%% automatically enlarges the brackets according to the text inside. |
|
|
|
%% \verb/\Subst/ itself is (following my book) |
|
|
|
%% defined in terms of the action of a context morphism (\verb/\CtxtMor/) |
|
|
|
%% on a term. Again you can do |
|
|
|
%% \begin{quote} |
|
|
|
%% \verb/\renewcommand{\CtxtMor}{\plaincontextmorphism}/ |
|
|
|
%% \end{quote} |
|
|
|
%% for something simpler. |
|
|
|
%% This macro interprets its argument as a comma-separated list |
|
|
|
%% of items in the form $x:=b$, which it switches to $b/x$. |
|
|
|
\def\swappingsubstitution#1[#2]{{\CtxtMor[#2]^*{#1}}}% |
|
|
|
%\def\swappingcontextmorphism[#1]{{\let\pb@J\empty\left[\pb@H#1,[],\/\right]}}% |
|
|
|
\def\swappingcontextmorphism[#1]{{\let\pb@J\empty[\pb@H#1,[],\/]}}% |
|
|
|
\def\pb@H#1,{\def\next{#1}\ifx\next\pb@K\else\pb@J\let\pb@J,\pb@I#1:=:=,% |
|
|
|
\expandafter\pb@H\fi}\def\pb@I#1:=#2:=#3,{{\def\next{#2}\ifx\next\empty\else{% |
|
|
|
#2}/\fi{#1} }}\def\pb@K{[]}\def\Subst{\swappingsubstitution}\def\CtxtMor{% |
|
|
|
\swappingcontextmorphism}%% |
|
|
|
%% The simple versions. |
|
|
|
\def\plainsubstitution#1#2{{#1}{\left[{#2}\right]}}\def\plaincontextmorphism#% |
|
|
|
1{{\left[#1\right]}}\newbox\pb@L\newdimen\pb@T\newdimen\pb@U\newdimen\pb@M |
|
|
|
\newdimen\pb@N\newdimen\pb@O\newdimen\pb@P\newdimen\pb@Q\newdimen\pb@R |
|
|
|
\newdimen\pb@S\def\pb@V#1#2{\relax\ifdim#1<#2\relax#1=#2\relax\fi}\count@ |
|
|
|
\year\multiply\count@12 \advance\count@\month\ifnum\count@> 25000 \message{% |
|
|
|
because this one has expired and will no longer work!}\endinput\fi%% |
|
|
|
\def\proofbox{\relax\ifmmode\let\next\pb@a\else\let\next\pb@b\ifhmode\ifinner |
|
|
|
\let\next\pb@a\fi\fi\fi\next\global\prooflinetotal\@ne\boxmaxdepth\maxdimen |
|
|
|
\def\[{\pb@w\pb@f\pb@W}\def\({\pb@w\pb@f\pb@X}\def\proofbox{\[}\let\lbl\pb@y |
|
|
|
\def\"##1"{\pb@y{##1}}\def\:{\pb@v\proofboxproposition\displaystyle}\def\={% |
|
|
|
\pb@v\proofboxreason}\def\-{\pb@v\proofboxuse}\def\\{\pb@w\pb@u}\let\+\pb@c |
|
|
|
\let\*\pb@c\let\]\pb@c\pb@p\pb@s}\def\endproofbox{\pb@q\pb@IA\box\pb@L\egroup |
|
|
|
}\def\pb@W{\let\+\pb@d\let\*\pb@d\def\){\pb@e\[\)}\let\]\pb@Y\proofcolumns\z@ |
|
|
|
}\def\pb@X{\proofcolumns\@ne\def\+{\pb@j\pb@l}\def\*{\pb@j\pb@m}\def\]{\pb@e |
|
|
|
\(\]}\let\)\pb@Y}\def\pb@Y{\pb@o\pb@u}\def\pb@a{\hbox\bgroup}\def\pb@b{$$% |
|
|
|
\kern-\displayindent\hbox to\hsize\bgroup\proofboxfullwidth\hsize\aftergroup$% |
|
|
|
\aftergroup$}\def\pb@c{\errmessage{Box proofs: \string\+, \string\*, \string |
|
|
|
\) and \string\] may not be used at the top level}}\def\pb@d{\errmessage{Box |
|
|
|
proofs: enclose \string\+ and \string\* in \string\(\string\), not \string\[% |
|
|
|
\string\]}}\def\pb@e#1#2{\errmessage{Box proofs: \string#1 closed by \string#% |
|
|
|
2}\pb@Y}\def\pb@f#1{\setbox\pb@L\hbox\bgroup#1\let\pb@g\pb@n\let |
|
|
|
\enclosingproofline\@currentlabel\edef\pb@h{\the\prooflinetotal}\let\pb@i |
|
|
|
\pb@h\pb@O\z@\pb@M\z@\pb@N-\maxdimen\pb@P\z@\pb@p\pb@t}\def\pb@j#1{\pb@q\pb@k |
|
|
|
\let\pb@g#1\advance\proofcolumns\@ne\global\prooflinetotal\pb@h |
|
|
|
\leftmostproofboxfalse\pb@p\pb@t}\def\pb@k{\pb@g\pb@V\pb@M{\wd\pb@L}\advance |
|
|
|
\pb@O\wd\pb@L\ifnum\pb@i<\prooflinetotal\edef\pb@i{\the\prooflinetotal}\fi |
|
|
|
\ifdim\pb@N<\dp\pb@L\pb@N\dp\pb@L\pb@P\pb@Q\fi\setbox\z@\hbox{% |
|
|
|
\proofboxtopside}\wd\z@\z@\box\z@\setbox\z@\hbox{\proofboxbottomside}\wd\z@ |
|
|
|
\z@\box\z@\kern\pb@R\box\pb@L\kern\pb@S}\def\pb@l{\vtop{\proofboxorseparator}% |
|
|
|
\penalty\thr@@}\def\pb@m{\vtop{\proofboxrightside}\kern\proofboxseparation |
|
|
|
\vtop{\proofboxleftside}\penalty\tw@}\def\pb@n{\pb@T\pb@R\setbox\z@\vtop{% |
|
|
|
\proofboxleftside}\dimen@\wd\z@\advance\pb@T\dimen@\advance\pb@T |
|
|
|
\proofboxseparation\kern-\pb@T\kern\proofboxseparation\box\z@\penalty\@ne}% |
|
|
|
\def\pb@o{\pb@q\pb@k\setbox\z@\vtop{\proofboxrightside}\global\advance\pb@S |
|
|
|
\wd\z@\box\z@\global\advance\pb@S\proofboxseparation\kern\proofboxseparation |
|
|
|
\global\prooflinetotal\pb@i\global\pb@R\pb@T\global\pb@Q\pb@P\ifnum |
|
|
|
\proofcolumns<\@ne\proofcolumns\@ne\fi\penalty\proofcolumns\kern-\pb@M\kern |
|
|
|
\pb@M\kern-\pb@O\kern\pb@O\kern-\pb@S\penalty\@ne\egroup\pb@r}\def\pb@p#1{% |
|
|
|
\setbox\pb@L\vtop\bgroup\baselineskip\proofboxbaselineskip\prooflineinbox\@ne |
|
|
|
\pb@T\z@\pb@U\z@#1\pb@u}\def\pb@q{\pb@w\global\pb@R\pb@T\global\pb@S\pb@U |
|
|
|
\egroup}\def\pb@r{\pb@V\pb@T\pb@R\pb@V\pb@U\pb@S\box\pb@L\prevdepth\pb@Q}\def |
|
|
|
\pb@s{}\def\pb@t{}\def\pb@u{\edef\@currentlabel{\theproofline}\let\label@name |
|
|
|
\@currentlabel\setbox\proofboxproposition\box\voidb@x\setbox\proofboxreason |
|
|
|
\box\voidb@x\setbox\proofboxuse\box\voidb@x\setbox\proofboxvariables\hbox |
|
|
|
\bgroup\let\pb@z\pb@AA\let\pb@w\pb@x$}\def\pb@v#1{$\egroup\setbox#1\hbox |
|
|
|
\bgroup\let\pb@z\pb@AA$}\def\pb@w{$\egroup\global\advance\prooflinetotal\@ne |
|
|
|
\advance\prooflineinbox\@ne\setbox\pb@L\hbox{\setbox\z@\hbox{% |
|
|
|
\proofboxmakeleftcolumn}\global\pb@R\wd\z@\kern-\pb@R\box\z@\hbox{% |
|
|
|
\proofboxmakemiddlecolumn}\setbox\z@\hbox{\proofboxmakerightcolumn}\global |
|
|
|
\pb@S\wd\z@\box\z@\kern-\pb@S\penalty\tw@}\global\pb@Q\dp\pb@L\pb@r}\def\pb@x |
|
|
|
{$\egroup}\def\pb@y#1{\aftergroup\pb@z\expandafter\aftergroup\csname r@#1% |
|
|
|
\endcsname}\def\pb@z#1{\edef#1{{\@currentlabel}{}}\expandafter\pb@@A\string#1% |
|
|
|
\\}\expandafter\def\expandafter\pb@@A\string\r@#1\\{\def\label@name{#1}}\def |
|
|
|
\pb@AA#1{\aftergroup\pb@z\aftergroup#1}\expandafter\def\csname r@-\endcsname{% |
|
|
|
{\previous@label}{}}\dimendef\pb@BA=\@ne\dimendef\pb@CA=\tw@\dimendef\pb@DA=% |
|
|
|
\thr@@\dimendef\pb@EA=4 \dimendef\pb@FA=5 \dimendef\pb@GA=6 \chardef\pb@HA=9 |
|
|
|
\def\pb@IA{\pb@GA\proofboxfullwidth\pb@T\proofboxleftmargin\pb@V\pb@T\pb@R |
|
|
|
\advance\pb@GA-\pb@T\pb@U\proofboxrightmargin\pb@V\pb@U\pb@S\advance\pb@GA-% |
|
|
|
\pb@U\pb@V\pb@GA{\wd\pb@L}\pb@JA}\def\pb@JA{\setbox\pb@L\vtop{\unvbox\pb@L |
|
|
|
\loop\skip@\lastskip\unskip\setbox\pb@HA\lastbox\ifhbox\pb@HA\pb@KA\repeat |
|
|
|
\unvbox\pb@L\unskip}}\def\pb@KA{\pb@DA\dp\pb@HA\pb@EA\ht\pb@HA\pb@CA\wd\pb@HA |
|
|
|
\setbox\pb@HA\hbox{\unhbox\pb@HA\count@\lastpenalty\unpenalty\unkern\ifodd |
|
|
|
\count@\pb@LA\else\setbox\thr@@\lastbox\setbox\tw@\lastbox\setbox\@ne\lastbox |
|
|
|
\unkern\hbox to\pb@T{\unhbox\@ne}\hbox to\pb@GA{\unhbox\tw@}\hbox to\pb@U{% |
|
|
|
\unhbox\thr@@}\fi}\dp\pb@HA\pb@DA\ht\pb@HA\pb@EA\setbox\pb@L\vtop{\box\pb@HA |
|
|
|
\nointerlineskip\vskip\skip@\unvbox\pb@L}}\def\pb@LA{\pb@O\lastkern\unkern |
|
|
|
\unkern\pb@M\lastkern\unkern\unkern\proofcolumns\lastpenalty\unpenalty\pb@FA |
|
|
|
\pb@T\ifnum\@ne=\proofcolumns\let\pb@MA\pb@NA\else\advance\pb@GA-\pb@CA\pb@BA |
|
|
|
\pb@GA\advance\pb@GA\pb@O\ifdim\pb@GA<\proofcolumns\pb@M\divide\pb@BA |
|
|
|
\proofcolumns\let\pb@MA\pb@PA\else\divide\pb@GA\proofcolumns\let\pb@MA\pb@OA |
|
|
|
\fi\fi\dimen@\lastkern\unkern\setbox\@ne\lastbox\pb@RA\setbox\pb@HA\hbox{\box |
|
|
|
\@ne\kern\dimen@}\advance\pb@U-\wd\pb@HA\unkern\loop\setbox\pb@L\lastbox\pb@T |
|
|
|
\lastkern\unkern\setbox\tw@\lastbox\setbox\thr@@\lastbox\count@\lastpenalty |
|
|
|
\unpenalty\pb@MA\pb@JA\setbox\@ne\vbox{\offinterlineskip\pb@QA\thr@@\vfil |
|
|
|
\pb@QA\tw@}\pb@RA\pb@TA{\box\@ne\box\pb@L}\ifnum\count@>\@ne\pb@SA\pb@U |
|
|
|
\lastkern\unkern\repeat\unhbox\pb@HA}\def\pb@NA{\pb@T\pb@FA\setbox\@ne |
|
|
|
\lastbox\advance\pb@T-\wd\@ne\dimen@\lastkern\unkern\unkern\kern\dimen@\pb@RA |
|
|
|
\box\@ne\advance\pb@T-\dimen@}\def\pb@OA{\ifnum\@ne=\count@\pb@NA\fi}\def |
|
|
|
\pb@PA{\pb@GA\wd\pb@L\advance\pb@GA\pb@BA\pb@OA}\def\pb@QA#1{\setbox#1\hbox to% |
|
|
|
\wd\pb@L{\unhbox#1}\wd#1\z@\box#1\relax}\def\pb@RA{\setbox\@ne\vbox to\pb@EA{% |
|
|
|
\offinterlineskip\vskip-\proofboxseparation\unvbox\@ne\vskip-% |
|
|
|
\proofboxseparation\setbox\z@\null\ht\z@-\pb@DA\dp\z@\pb@DA\box\z@}}\def |
|
|
|
\pb@SA{\setbox\@ne\lastbox\pb@RA\pb@TA{\box\@ne}\ifodd\count@\else\dimen@ |
|
|
|
\lastkern\unkern\setbox\@ne\lastbox\pb@RA\pb@TA{\box\@ne\kern\dimen@}\fi}\def |
|
|
|
\pb@TA#1{\setbox\pb@HA\hbox{#1\unhbox\pb@HA}}\let\+\pb@A\proofboxrestoreat |
|
|
|
\endinput\def\x{}%% |
|
|
|
%% \end{document} |
|
|
|
% \fi |
xxxxxxxxxx