% This is the AMERICAN ENGLISH package documentation for "listofitems" \documentclass[american]{article} \usepackage[T1]{fontenc} \usepackage[utf8]{inputenc} \usepackage{listofitems,babel,xspace} \usepackage[bottom]{footmisc} \usepackage{libertine}\usepackage[italic]{mathastext} \usepackage[scaled=0.8]{luximono} \usepackage[margin=3cm]{geometry} \usepackage{fancybox} \usepackage{xcolor} \usepackage{listings,MnSymbol} \def\listofitems{\textsf{\loiname}\xspace} \def\eTeX{\hbox{$\varepsilon$-\TeX}} \makeatletter \def\code{\expandafter\code@i\string} \def\code@i#1{% \begingroup \par\nobreak\medskip\parindent0pt \leftskip.1\linewidth \catcode`\^^I 13 \begingroup\lccode`\~`\^^I \lowercase{\endgroup\def~{\leavevmode\space\space\space\space}}% \let\do\@makeother \dospecials \ttfamily\small\@noligs \obeylines\obeyspaces \def\code@ii##1#1{##1\par\medbreak\endgroup}% \code@ii } \long\def\grab@toks#1\relax{\gdef\right@content{#1}} \newcommand\disable@lig[1]{% \catcode`#1\active \begingroup \lccode`\~`#1\relax \lowercase{\endgroup\def~{\leavevmode\kern\z@\string#1}}% } \newcommand\exemple[1][65]{% \edef\par@@indent{\the\parindent}% \par\nobreak\vskip5pt \noindent \def\part@coeff{#1}% \relax\leavevmode\null \bgroup \let\do\@makeother\dospecials \catcode`\^^M\active \catcode`\ \active \begingroup \lccode`\~`\ \lowercase{\def~{ {}}}% \lccode`\~`\^^M\lowercase{\endgroup\def~{\par\noexpand\leavevmode}}% \@makeother:\disable@lig,\disable@lig-% \exemple@@ } \newcommand\exemple@@[1]{% \def\@tempa##1#1{% \xdef\left@content{##1}% \egroup \def\right@content{##1}% \begingroup \newlinechar`\^^M\everyeof{\relax}% \expandafter\grab@toks\scantokens\expandafter{\right@content}% \endgroup \exemple@@@ }% \@tempa } \begingroup \catcode`#=12 \gdef\exemple@@@{% \begingroup \fboxsep1pt\relax \edef\part@left{\the\dimexpr0.\part@coeff\linewidth-\fboxsep-\fboxrule}% \edef\part@right{\the\dimexpr\linewidth-0.\part@coeff\linewidth-\fboxsep-\fboxrule}% \fbox{% \parbox[c]\part@left{\vskip5pt\relax\ttfamily\footnotesize\left@content\vskip5pt}% \parbox[c]\part@right{\vskip5pt\relax\normalfont\footnotesize\right@content\vskip5pt}% }% \vskip7.5pt\relax \endgroup } \endgroup \begingroup \catcode`\<13 \catcode`\>13 \gdef\verb{\relax\ifmmode\hbox\else\leavevmode\null\fi \bgroup \verb@eol@error \let\do\@makeother \dospecials \verbatim@font\@noligs \catcode`\<13 \catcode`\>13 \def<{\begingroup$\langle$\itshape}\def>{$\rangle$\endgroup}% \@ifstar\@sverb\@verb} \endgroup \makeatother % FOR AMERICAN-ENGLISH VERSION \usepackage{setspace} \setstretch{1.01} \usepackage{enumitem} \newcommand\no[1]{number #1} \let\bsc\textsc \def\americanloidate{\expandafter\americanloidatei\loidate/} \def\americanloidatei#1/#2/#3/{\number#3\relax\space\ifcase #2 \or January\or February\or March\or April\or May\or June\or July\or August\or September\or October\or November\or December\fi{} #1} %\newcommand\flag[1]{\textcolor{red}{#1}}% DIAGNOSTIC FLAG FOR FURTHER TRANSLATION %\newcommand\cflag[1]{\textcolor{blue}{#1}}% DIAGNOSTIC FLAG FOR FURTHER TRANSLATION % \begin{document} \parindent=0pt \thispagestyle{empty} \begin{titlepage} \renewcommand\thefootnote{\fnsymbol{footnote}} \begingroup \centering \null\vskip.25\vsize {\large\bfseries Package\par \Huge \listofitems\par} \bigbreak v\loiver \smallbreak \americanloidate \vskip1.5cm {Christian \bsc{Tellechea}\footnote{\texttt{unbonpetit@netc.fr}}\par Steven B. \bsc{Segletes}\footnote{steven.b.segletes.civ@mail.mil}} \par \endgroup \vskip2cm \leftskip=.2\linewidth \rightskip=.2\linewidth \small This simple package is designed to read a list of items whose parsing separator may be selected by the user. Once the list is read, its items are stored in a structure that behaves as a dimensioned array. As such, it becomes very easy to access an item in the list by its number. For example, if the list is stored in the macro \verb|\foo|, the item \no3 is designated by \verb|\foo[3]|. A component may, in turn, be a list with a parsing delimiter different from the parent list, paving the way for nesting and employing a syntax reminiscent of an array of several dimensions of the type \verb|\foo[3,2]| to access the item \no2 of the list contained within the item \no3 of the top-tier list. \end{titlepage} \def\listofitems{\textsf\loiname\xspace} \section{Preface} This package loads no external packages, must be used with the \eTeX{} engine, and must be called in (pdf)\hskip0pt(Xe)\hskip0pt(lua)\LaTeX{} with the invocation \code|\usepackage{listofitems}| and under (pdf)(Xe)(Lua)\TeX {} by way of \code|\input listofitems.tex| \section{Read a Simple List} \paragraph{Set the parsing separator} The default parsing separator is the comma and if we want change it, we must do so before reading a list of items, with the definition \verb|\setsepchar{}|. A \verb|| is a set of tokens which possess catcodes different from 1 and 2 (the opening and closing braces), 14 (usually \verb|%|) and 15. The token of catcode 6 (usually \verb|#|) is accepted only if it is followed by an integer, denoting the argument of a macro; In no case should this token be provided alone as the \verb||. Commands can be included in this set of tokens, including the \TeX{} primitive \verb|\par|.% \smallbreak The parsing-separator \verb|| ``\verb-/-'' is reserved by default for nested lists (see page 3). %\ref{nested.list.delimiters}). It is therefore not proper to write "\verb-\setsepchar{/}-" because the \listofitems{} package would misunderstand that you want to read a nested list. To set ``\verb-/-'' as the \verb|| for a simple list, it is necessary, using the optional argument, to choose a different parsing-separator \verb|| for nested lists, for example ``\verb|.|'', and write ``\verb+\setsepchar[.]{/}+''. \medbreak It is not possible to select \verb-|- as the \verb|| because it would conflict with the logical \textbf{OR}, denoted ``\verb-||-'' (see below). However, one can work around this limitation, at one's own peril, writing ``\verb-\setsepchar{{|}}-''. \paragraph{Read a list} To read the list of items, the \verb|\readlist{}| should be called. In so doing, the \verb|| is read and the items are stored in a macro, denoted \verb|| which therefore acts as a table with the items of the \verb||. If braces appear as part of a list item, they \emph{must} be balanced. Tokens possessing the catcodes 6, 14 and 15 are not allowed in the lists. For example, to set the \verb|| named \verb|\foo|, we can write \code|\setsepchar{,} \readlist\foo{12,abc, x y ,{\bfseries z}, ,\TeX,,!}| \setsepchar{,} \readlist\foo{12,abc, x y ,{\bfseries z}, ,\TeX,,!} If the \verb|| is contained in a macro, then this macro is expanded. Therefore, we can simply employ the syntax \verb|\readlist| as in \code|\setsepchar{,} \def\List{12,abc, x y ,{\bfseries z}, ,\TeX,,!} \readlist\foo\List| The macro \verb|\greadlist| makes \emph{global} assignments and therefore, enables the use of \verb|| outside of the group where \verb|\greadlist| has been executed. \paragraph{Access an item} The macro \verb|\foo| \emph{requires} a numeric argument in square brackets, which we symbolically denote as $i$, indicating the rank of the item you wish to access. So \verb|\foo[1]| is\footnote{\texttt{\textbackslash foo[}$i$\texttt] requires 2 expansions to give the item.} ``\verb|12|''. Similarly, \verb|\foo[4]| is ``\verb|{\bfseries z}|''. The number $i$ can also be negative in which case the counting is done from the end of the list: $-1$ represents the last item, $-2$ the penultimate, etc. If the number of items is $n$, then the argument $-n$ is the first item. \medbreak In general, if a \verb|| has a length $n$, then the index $i$ can be in the interval $[1\,;n]$ or $[-n\,;-1]$. Otherwise, a compilation error occurs. If the index is empty, \verb|\foo[]| produces the complete \verb||. The macro \verb|\foosep| is created. It is used with the syntax \verb|\foosep[]| and allows access to the parsing-separator that follows the item of rank \verb||. The last parsing-separator (the one following the last item) is empty. If the \verb|| is empty, \verb|\foosep[]| is empty. \paragraph{Select several possible parsing separators} To specify several possible separators, use the \textbf{OR} operator, denoted ``\verb-||-''. One can use this feature, for example, to isolate the terms in an algebraic sum: \exemple/\setsepchar{+||-} \readlist\term{17-8+4-11} 1) \term[1] (parsing separator = \termsep[1])\par 2) \term[2] (parsing separator = \termsep[2])\par 3) \term[3] (parsing separator = \termsep[3])\par 4) \term[4] (parsing separator = \termsep[4])/ \paragraph{Number of items} If we write \verb|\readlist{}|, then the macro \verb|len| con\-tains\footnote{That is to say, it is purely expandable and grows into a number.} the number of the items in \verb||. In the example with \verb|\foo|, the macro \verb|\foolen| expands to \foolen. \paragraph{View all items} For purposes of debugging, the macro \verb|\showitems| includes all items from a list, while the star version displays these items ``detokenized.'' \footnote{The primitive \texttt{\string\detokenize}, conducting this decomposition, inserts a space after each control sequence.} \exemple/\showitems\foo\par \showitems*\foo/ The presentation of each list item is assigned to the macro \verb|\showitemsmacro| whose code is \code|\newcommand\showitemsmacro[1]{% \begingroup\fboxsep=0.25pt \fboxrule=0.5pt \fbox{\strut#1}\endgroup \hskip0.25em\relax}| It is therefore possible ---~and desirable~--- to redefine it if we desire a different presentation effect.\medbreak The macro \verb|\fbox| and associated dimensions \verb|\fboxsep| and \verb|\fboxrule|, are defined by \listofitems, when \textit{not} compiled under \LaTeX{}, to achieve the same result \textit{as if} performed under \LaTeX. \paragraph{Suppression of extreme (leading/trailing) spaces} By default, \listofitems reads and retains the spaces located at the beginning and end of an item. For these spaces to be ignored when reading the \verb||, execute the starred version \verb|\readlist*{}|: \exemple|\setsepchar{,} \readlist*\foo{12,abc, x y ,{\bfseries z}, ,\TeX,,!} \showitems\foo| \paragraph{Managing empty items} By default, the \listofitems package retains and accounts for empty items. Thus, in the previous ex\-am\-ple, the 2nd expansion of \verb|\foo[7]| is empty. For empty items of the list (i.e., those list items defined by two consecutive parsing delimiters) to be ignored, we must, before invoking \verb|\readlist|, execute the macro \verb|\ignoreemptyitems| . To return to the default package behavior, simply execute the macro \verb|\reademptyitems|. This option can be used alone or in combination with \verb|\readlist*|, in which case the suppression of extreme (leading/trailing) spaces occurs \textit{before} \textsf{listofitems} ignores the empty list items: \exemple|\setsepchar{,} \ignoreemptyitems \readlist\foo{12,abc, x y ,{\bfseries z}, ,\TeX,,!} a) number of items = \foolen\par \showitems\foo \readlist*\foo{12,abc, x y ,{\bfseries z}, ,\TeX,,!} b) number of items = \foolen\par \showitems\foo| \paragraph{Iterate over a list} Once a list read by \verb|\readlist| and stored in a \verb||, one may iterate over the list with the syntax \verb|\foreachitem \in {}|: The \verb|| is a macro chosen by the user that will loop over the value of each item in the list. The macro \verb|cnt| represents the sequence number of the item in \verb||. \exemple|\setsepchar{ }% parsing-separator = space \readlist\phrase{One phrase to test.} \foreachitem\word\in\phrase{List item number \wordcnt{}: \word\par}| \paragraph{Assign an item to a macro} The \verb|\itemtomacro[index]| assigns to the \verb|| the item designated by \verb|[index]|. The \verb|| thus defined is purely expandable provided that the tokens in the items are expandable. \exemple|\setsepchar{ }% parsing-separator = space \readlist\phrase{One phrase to test.} \itemtomacro\phrase[2]\aword \meaning\aword\par \itemtomacro\phrase[-1]\wordattheend \meaning\wordattheend| \section{Nested Lists} We speak of a list being ``nested'' when asking \listofitems to read a list where the items are, in turn, understood as being a list (implying a parsing separator different from the top-tier list). The nesting depth is not limited, but in practice, a depth of 2 or 3 will usually suffice. \paragraph{Defining the parsing separators}\label{nested.list.delimiters} To indicate that a list will be nested, so that the list parsing will be performed recursively, one must specify multiple parsing separators, each corresponding to the particular tier of nesting. This list of parsing separators is itself given as a delimited list to the macro \verb|\setsepchar|, with the syntax \verb|\setsepchar[]{}|. By default, the \verb|| is ``\texttt{/}''. Thus, writing \code|\setsepchar{\\/,/ }| indicates a recursive depth of 3, with the parsing-separator list delimiter defaulting to ``\texttt{/}'': \begin{itemize}[itemsep=0pt,topsep=0pt,partopsep=0pt,parsep=0pt,label=---] \item Tier 1 items are parsed between ``\verb|\\|'' delimiters; \item Tier 2 items are found within Tier 1 items, parsed between ``\verb|,|'' delimiters; \item finally, the Tier 3 items are found within Tier 2 items, parsed between the ``\verb*| |'' delimiters. \end{itemize} The \verb|| of nesting is contained in the purely expandable macro \verb|\nestdepth|. \paragraph{Read and access list items} For nested lists, the use of indices obey the following rules: \begin{itemize}[itemsep=0pt,topsep=0pt,partopsep=0pt,parsep=0pt,label=---] \item \verb|[]| is the main list, i.e., the argument of \verb|\readlist|; \item \verb|[]| means the item \no\verb|| of the main list; \item \verb|[,]| means the item \no\verb|| of the list mentioned in the previous point (a subitem); \item \verb|[,,]| means the item \no\verb|| of the list mentioned in the previous point (a sub-subitem); \item etc. \end{itemize} As in the case of a non-nested list, the index may be negative. To read items, the syntax of \verb|\readlist| is exactly the same as that for simple (non-nested) lists: \exemple|\setsepchar{\\/,/ } \readlist\baz{1,2 a b,3 c\\4 d e f,5,6\\7,,8, ,9 xy z} a) \string\baz[1] is \baz[1]\par b) \string\baz[1,1] is \baz[1,1]\par c) \string\baz[1,1,1] is \baz[1,1,1]\par b) \string\bar[1,2] is \baz[1,2]\par e) \string\baz[1,2,3] is \baz[1,2,3]\par f) \string\baz[-2,1,-1] is \baz[-2,1,-1]| \paragraph{The operator ``||''} This operator may be employed at any level of nesting. \exemple=\setsepchar[,]{+||-,*||/} \readlist\numbers{1+2*3-4/5*6} Term 1: \numbers[1]\par Term 2: \numbers[2] (factors: \numbers[2,1] and \numbers[2,2])\par Term 3: \numbers[3] (factors: \numbers[3,1], \numbers[3,2] and \numbers[3,3])= \paragraph{Number of list items} The macro \verb|\listlen[]| requires 2 expansions in order to give the number of items in the list specified by the \verb||. The \verb|| of the \verb|| must be strictly less than that of the list. For the case where the \verb|| is empty, \verb|\listlen[]|, with 2 expansions, yields the identical result as \verb|len| with 1 expansion. \exemple|\setsepchar{\\/,/ } \readlist\baz{1,2 a b,3 c\\4 d e f,5,6\\7,,8, ,9 xy z} a) \bazlen\ or \listlen\baz[]\par b) \listlen\baz[1]\par c) \listlen\baz[2]\par d) \listlen\baz[3]\par e) \listlen\baz[3,1]\par f) \listlen\baz[3,4]\par% 2 empty items g) \listlen\baz[3,5]| \paragraph{Displaying list items} The macro \verb|\showitems[]| displays items from the list specified by \verb||, in the same manner as \verb|\listlen|. The \verb|| of the \verb|| must be strictly less than that of the \verb||. \exemple|\setsepchar{\\/,/ } \readlist\baz{1,2 a b,3 c\\4 d e f,5,6\\7,,8, ,9 xy z} a) \showitems\baz[]\par b) \showitems\baz[1]\par c) \showitems\baz[2]\par d) \showitems\baz[3]\par e) \showitems\baz[3,1]\par f) \showitems\baz[3,4]\par% 2 empty items g) \showitems\baz[3,5]| \paragraph{Empty items and extreme (leading/trailing) spaces} The removal of empty items and/or leading/\allowbreak trailing spaces will occur in \emph{all} the items, regardless of the degree of nesting. It is clear that a space, ``\verb*| |'', is useless as a parsing separator if you want to use \verb|\readlist*|. Therefore, in the following example, ``\verb|*|'' is instead selected as the (3rd-tier) parsing separator. Further, we remove only the extreme spaces, but retain empty items. \exemple|\setsepchar{\\/,/*} \readlist*\baz{1, 2*a*b ,3*c\\4*d*e*f,5,6\\7,,8, ,9* xy *z} a) \showitems\baz[]\par b) \showitems\baz[1]\par c) \showitems\baz[2]\par d) \showitems\baz[3]\par e) \showitems\baz[3,1]\par f) \showitems\baz[3,4]\par g) \showitems\baz[3,5]% "xy" without extreme spaces| \paragraph{Iterate over a list} The syntax \verb|\foreachitem \in []{}| remains valid where now the \verb|| specifies the item (understood as a list) on which to iterate. The \verb|| of the \verb|| must be strictly less than that of the \verb||. \paragraph{Assign an item to a macro} The syntax \verb|\itemtomacro[]| remains valid to assign to \verb|| the item specified by \verb|[]|. \exemple[55]|\setsepchar[,]{\\, } \readlist\poem{There once was a runner named Dwight\\% Who could speed even faster than light.\\% He set out one day\\% In a relative way\\% And returned on the previous night.} \itemtomacro\poem[2]\verse 2nd verse = \verse \itemtomacro\poem[2,-4]\word A word = \word| The macro \verb|\gitemtomacro| makes a global assignment. \section{Balanced Tokens} For the parsing of items, it is possible, with version 1.6, to take into account the presence of \emph{balanced tokens}. Thus, if a list of paired tokens is defined, then each parsed item in the list will extend to the first \verb||, while assuring that any paired tokens are balanced (i.e., occur in matched pairs within the item). To define a list of balanced-token pairs, we use \code|\defpair{...}| where the token list is read in pairs to form each matched-token pair. A \verb|| that serves within a matched pair must consist of a single character---% macros, primitives, spaces, braces, the token "\verb|#|", as well as sets of several-tokens-between-% braces are all forbidden. The two tokens which form a pair \emph{must} be different from each other. \exemple/\setsepchar{+||-} \defpair{()[]} \readlist\terms{1+2*[3+4*(5+6-7)+8]-9+10} \showitems\terms/ To return to the package's default behavior, that is, without paired tokens, you must execute \code|\defpair{}| In an expression, in order to store in a macro that which is between two matched tokens, we can call on \code|\insidepair{}\macro| which will put in the \verb|\macro| that which lies between the pair \verb|| in the \verb||. \exemple/\setsepchar{+||-} \defpair{()} \readlist\terms{1+2*(3+4*(5+6-7)+8)-9+10} \showitems\terms \itemtomacro\terms[2]\parenterm In the outer parenthesis: \insidepair()\parenterm\inbigparen "\inbigparen" In the inner parenthesis: \insidepair()\inbigparen\insmallparen "\insmallparen"/ \section{The Code} Any suggestion, bug report, remark, request, addition or modification of functionality is welcome; in this case, I invite users of \listofitems to send me an email to \verb|unbonpetit@netc.fr|.\medbreak The code below is the exact verbatim of the file \verb|listofitems.tex|. I hope that the few comments scattered throughout it will be enough for the user or the curious to understand the internal machinery of this package: \lstinputlisting[ language=TeX, inputencoding=latin1, moretexcs={unless,ifcsname,ifdefined,detokenize,numexpr,dimexpr,glueexpr,unexpanded}, basicstyle=\small\ttfamily\color{black!25}, identifierstyle=\bfseries\color{white},% backgroundcolor=\color{black!85}, keywordstyle=\bfseries\color{orange}, commentstyle=\color{cyan!66}, columns=fixed, alsoletter={\_}, tabsize=2, extendedchars=true, showspaces=false, showstringspaces=false, numbers=left, numberstyle=\tiny\ttfamily\color{black}, breaklines=true, prebreak={\hbox{$\rhookswarrow$}}, breakindent=3em, breakautoindent=true, xleftmargin=1em, xrightmargin=0pt, lineskip=0pt, numbersep=1em, classoffset=1, alsoletter={\_}, morekeywords={defpair,insidepair,setsepchar,greadlist,readlist,listlen,foreachitem,% showitems,showitemsmacro,itemtomacro,gitemtomacro,ignoreemptyitems,reademptyitems}, keywordstyle=\bfseries\color{red!85!black}, classoffset=0, ]{listofitems.tex} \end{document}