summaryrefslogtreecommitdiff
path: root/macros/luatex/latex/piton/piton.dtx
diff options
context:
space:
mode:
Diffstat (limited to 'macros/luatex/latex/piton/piton.dtx')
-rw-r--r--macros/luatex/latex/piton/piton.dtx1712
1 files changed, 1187 insertions, 525 deletions
diff --git a/macros/luatex/latex/piton/piton.dtx b/macros/luatex/latex/piton/piton.dtx
index 7d7a94fa1d..c381b9eb27 100644
--- a/macros/luatex/latex/piton/piton.dtx
+++ b/macros/luatex/latex/piton/piton.dtx
@@ -15,8 +15,8 @@
%
% \fi
% \iffalse
-\def\myfileversion{0.8}
-\def\myfiledate{2022/10/23}
+\def\myfileversion{0.9}
+\def\myfiledate{2022/10/31}
%
%
%<*batchfile>
@@ -58,6 +58,7 @@ version 2005/12/01 or later.
\usepackage{ragged2e}
\usepackage[footnotehyper,escape-inside=$$]{piton} % $$
+
\def\interitem{\vskip 7mm plus 2 mm minus 3mm}
\parindent 0pt
\skip\footins = 2\bigskipamount
@@ -72,6 +73,9 @@ version 2005/12/01 or later.
%</driver>
% \fi
+%
+% \catcode`\" = 11
+%
% \title{The package \pkg{piton}\thanks{This document corresponds to the
% version~\myfileversion\space of \pkg{piton}, at the date of~\myfiledate.}}
% \author{F. Pantigny \\ \texttt{fpantigny@wanadoo.fr}}
@@ -84,6 +88,7 @@ version 2005/12/01 or later.
% \end{abstract}
%
%
+%
% \section{Presentation}
%
%
@@ -105,8 +110,10 @@ version 2005/12/01 or later.
% from math import pi
%
% def arctan(x,n=10):
-% '''Compute the value of arctan(x)
-% n is the number of terms of the sum'''
+% """Compute the mathematical value of arctan(x)
+%
+% n is the number of terms in the sum
+% """
% if x < 0:
% return -arctan(-x) # recursive call
% elif x > 1:
@@ -179,47 +186,62 @@ version 2005/12/01 or later.
% \begin{itemize}
% \item The key |gobble| takes in as value a positive integer~$n$: the first $n$
% characters are discarded (before the process of highlightning of the code) for
-% each line of the environment |{Piton}|.
+% each line of the environment |{Piton}|. These characters are not necessarily
+% spaces.
%
-% \item Then the key |auto-gobble| is in force, the extension \pkg{piton} computes
-% the minimal value $n$ of the number of consecutives space beginning each (non
+% \item When the key |auto-gobble| is in force, the extension \pkg{piton} computes
+% the minimal value $n$ of the number of consecutive spaces beginning each (non
% empty) line of the environment |{Piton}| and applies |gobble| with that value
% of~$n$.
%
-% \item When the key |env-gobble| is in force, \pkg{piton} applies |gobble| with
-% a value of $n$ equal to the number of spaces before |\end{Piton}| on the last
-% line (if that line contains only spaces). The name of that key comes from
-% \emph{environment gobble}: the effect of gobble is set by the position of the
-% commands |\begin{Piton}| and |\end{Piton}| which delimit the current environment.
-%
-% \item With the key |line-numbers|, the \emph{non empty} lines are numbered in
-% the environments |{Piton}| and in the listings resulting from the use of
-% |\PitonInputFile|.
+% \item When the key |env-gobble| is in force, \pkg{piton} analyzes the last
+% line of the environment |{Piton}|, that is to say the line which contains
+% |\end{Piton}| and determines whether that line contains only spaces followed
+% by the |\end{Piton}|. If we are in that situation, \pkg{piton} computes the
+% number~$n$ of spaces on that line and applies |gobble| with that value of~$n$.
+% The name of that key comes from \emph{environment gobble}: the effect of
+% gobble is set by the position of the commands |\begin{Piton}| and
+% |\end{Piton}| which delimit the current environment.
+%
+% \item With the key |line-numbers|, the \emph{non empty} lines are numbered (on
+% the left) in the environments |{Piton}| and in the listings resulting from the
+% use of |\PitonInputFile|.
%
% \item With the key |all-line-numbers|, \emph{all} the lines are numbered,
% including the empty ones.
%
-% \item By default, the counter of lines is set to zero at the beginning of each
-% environment |{Piton}| or use of |\PitonInputFile|. With the key |resume|, that
-% reinitialisation is not done.
+% \item With the key |resume| the counter of lines is not set to zero at the
+% beginning of each environment |{Piton}| or use |\PitonInputFile| as it is
+% otherwise. That's allows a numbering of the lines accross several
+% environments. See an example part \ref{example-numbering} on
+% page~\pageref{example-numbering}.
+%
+% \item The key |left-margin| corresponds to a margin on the left. That key may
+% be useful in conjonction with the key |line-numbers| or the key
+% |line-all-numbers| if one does not want the numbers in an overlapping position
+% on the left.
%
% \item The key |splittable| allows pages breaks within the environments
% |{Piton}| and the listings produced by
% |\PitonInputFile|.
%
+% \smallskip
+% \colorbox{yellow!50}{\textbf{New 0.9}}\enskip It's possible to give as value to
+% the key |splittable| a positive integer~$n$. With that value, the environments
+% |{Piton}| and the listings produced by |\PitonInputFile| are splittable but no
+% page break can occur within the first $n$ lines and within the last $n$ lines.
+% The default value of the key |splittable| is, in fact, $1$, which allows pages
+% breaks everywhere.
+%
% \item The key |background-color| sets the background color of the environments
% |{Piton}| and the listings produced by |\PitonInputFile| (that background has
% a width of |\linewidth|). Even with a background color, the pages breaks are
-% allowed, as soon as the key |splittable| is in force.\footnote{The
-% environments \texttt{\{Piton\}} are breakable, even within a breakable
-% environment of \pkg{tcolorbox}. Remind that an environment of \pkg{tcolorbox}
-% included in another environment of \pkg{tcolorbox} is \emph{not} breakable,
-% even when both environments use the key |breakable| of \pkg{tcolorbox}.}
-%
-% \item \colorbox{yellow!50}{\textbf{New 0.8}}\enskip The key |left-margin|
-% corresponds to a margin on the left. That key may be useful in conjonction
-% with the key |line-numbers| or the key |line-all-numbers| if one does not want
-% the numbers in an overlapping position on the left.
+% allowed, as soon as the key |splittable| is in force.\footnote{With the key
+% |splittable|, the environments \texttt{\{Piton\}} are breakable, even within a
+% (breakable) environment of \pkg{tcolorbox}. Remind that an environment of
+% \pkg{tcolorbox} included in another environment of \pkg{tcolorbox} is
+% \emph{not} breakable, even when both environments use the key |breakable| of
+% \pkg{tcolorbox}.}
% \end{itemize}
%
% \bigskip
@@ -230,10 +252,11 @@ version 2005/12/01 or later.
% ~emphase&\PitonOptions{line-numbers,auto-gobble,background-color = gray!15}@
% \begin{Piton}
% from math import pi
-%
% def arctan(x,n=10):
-% '''Compute the value of arctan(x)
-% n is the number of terms of the sum'''
+% """Compute the mathematical value of arctan(x)
+%
+% n is the number of terms in the sum
+% """
% if x < 0:
% return -arctan(-x) # recursive call
% elif x > 1:
@@ -254,8 +277,10 @@ version 2005/12/01 or later.
% from math import pi
%
% def arctan(x,n=10):
-% '''Compute the value of arctan(x)
-% n is the number of terms of the sum'''
+% """Compute the mathematical value of arctan(x)
+%
+% n is the number of terms in the sum
+% """
% if x < 0:
% return -arctan(-x) # recursive call
% elif x > 1:
@@ -323,7 +348,7 @@ version 2005/12/01 or later.
% The package \pkg{piton} provides the command |\SetPitonStyle| to customize the
% different styles used to format the syntactic elements of the Python listings.
% The customizations done by that command are limited to the current TeX
-% group.\footnote{We remind that an LaTeX environment is, in particuilar, a TeX group.}
+% group.\footnote{We remind that an LaTeX environment is, in particular, a TeX group.}
%
% \bigskip
% The command |\SetPitonStyle| takes in as argument a comma-separated list of
@@ -379,25 +404,27 @@ version 2005/12/01 or later.
% String.Long & the long strings (between \texttt{'''} or \verb|"""|) except the
% documentation strings \\
% String & that keys sets both |String.Short| and |String.Long| \\
-% String.Doc & the documentation strings \\
+% String.Doc & the documentation strings (only between \texttt{"""} following
+% PEP~257) \\
% String.Interpol & the syntactic elements of the fields of the f-strings (that
-% is to say the characters \texttt{\{}, \texttt{\}} and \texttt{:}) \\
+% is to say the characters \texttt{\{} and \texttt{\}}) \\
% Operator & the following operators : \texttt{!= == << >> - \~{} + / * \% = < > \& .} \verb+|+ \verb|@|\\
% Operator.Word & the following operators : \texttt{in}, \texttt{is},
% \texttt{and}, \texttt{or} and \texttt{not} \\
% Name.Builtin & the predefined functions of Python \\
% Name.Function & the name of the functions defined by the user, at the point of
% their definition (that is to say after the keyword |def|) \\
-% Name.Decorator & the decorators (instructions beginning by \verb|@| in the classes) \\
+% Name.Decorator & the decorators (instructions beginning by \verb|@|) \\
% Name.Namespace & the name of the modules (= external libraries) \\
-% Name.Class & the name of the classes at the point of their definition \\
+% Name.Class & the name of the classes at the point of their definition (that is
+% to say after the keyword |class|) \\
% Exception & the names of the exceptions (eg: \texttt{SyntaxError}) \\
% Comment & the comments beginning with \verb|#| \\
-% Comment.LaTeX & the comments beginning by \verb|##| which are composed in LaTeX by
-% \pkg{piton} (and called merely ``LaTeX comments'' in this document) \\
+% Comment.LaTeX & the comments beginning by \verb|##|, which are composed in LaTeX by
+% \pkg{piton} (and simply called ``LaTeX comments'' in this document) \\
% Keyword.Constant & |True|, |False| and |None| \\
% Keyword & the following keywords :
-% \ttfamily assert, break, case, continue, del,
+% \ttfamily as, assert, break, case, continue, def, del,
% elif, else, except, exec, finally, for, from,
% global, if, import, lambda, non local,
% pass, raise, return, try, while,
@@ -458,8 +485,8 @@ version 2005/12/01 or later.
% pair |\footnotemark|--|\footnotetext|.
%
% \smallskip
-% It's also possible to extract the footnotes with the help of the package
-% \pkg{footnote} or the package \pkg{footnotehyper}.
+% However, it's also possible to extract the footnotes with the help of the
+% package \pkg{footnote} or the package \pkg{footnotehyper}.
%
% \smallskip
% If \pkg{piton} is loaded with the option |footnote| (with
@@ -479,17 +506,30 @@ version 2005/12/01 or later.
% in particular: it must be loaded after the package \pkg{xcolor} and it is not
% perfectly compatible with \pkg{hyperref}.
%
-% \medskip
-% In this document, the package \pkg{piton} has been loaded with the option
-% |footnotehyper|: see the first page of this document for an example of a
-% footnote in an environment |{Piton}|.
+% \medskip
+% In this document, the package \pkg{piton} has been loaded with the
+% option |footnotehyper|. For examples of notes, cf. \ref{notes-examples},
+% p.~\pageref{notes-examples}.
%
+% \subsection{Tabulations}
%
-% \section{Examples}
+% \smallskip
+% \colorbox{yellow!50}{\textbf{New 0.9}}\enskip Even though it's recommended to
+% indent the Python listings with spaces (see PEP~8), \pkg{piton} accepts the
+% characters of tabulation (that is to say the characters U+0009) at the
+% beginning of the lines. Each character U+0009 is replaced by $n$~spaces. The
+% initial value of $n$ is $4$ but it's possible to change it with the key
+% |tab-size| of |\PitonOptions|.
%
+% \section{Examples}
%
% \subsection{Line numbering}
%
+% \label{example-numbering}
+%
+% We remind that it's possible to have an automatic numbering of the lines in
+% the Python listings by using the key |line-numbers| or the key |all-line-numbers|.
+%
% By default, the numbers of the lines are composed by \pkg{piton} in an
% overlapping position on the left (by using internally the command |\llap| of LaTeX).
%
@@ -625,8 +665,10 @@ version 2005/12/01 or later.
%
% \subsection{Notes in the listings}
%
+% \label{notes-examples}
+%
% In order to be able to extract the notes (which are typeset with the command
-% |\footnote|, the extension \pkg{piton} must be loaded with the key |footnote|
+% |\footnote|), the extension \pkg{piton} must be loaded with the key |footnote|
% or the key |footenotehyper| as explained in the section \ref{footnote}
% p.~\pageref{footnote}. In this document, the extension \pkg{piton} has been
% loaded with the key |footnotehyper|.
@@ -709,7 +751,7 @@ version 2005/12/01 or later.
%
%
% \vspace{1cm}
-% If one embed an environment |{Piton}| in an environment |{minipage}|
+% If we embed an environment |{Piton}| in an environment |{minipage}|
% (typically in order to limit the width of a colored background), it's
% necessary to embed the whole environment |{minipage}| in an environment
% |{savenotes}| (of \pkg{footnote} or \pkg{footnotehyper}) in order to have the
@@ -778,7 +820,7 @@ version 2005/12/01 or later.
% Name.Builtin = ,
% Name.Function = \bfseries \colorbox{gray!20} ,
% Comment = \color{gray} ,
-% Comment.LaTeX = \color{gray},
+% Comment.LaTeX = \normalfont \color{gray},
% Keyword = \bfseries ,
% Name.Namespace = ,
% Name.Class = ,
@@ -792,19 +834,19 @@ version 2005/12/01 or later.
%
% \setmonofont[Scale=0.85]{DejaVu Sans Mono}
%
-% \PitonOptions{splittable}
+% \PitonOptions{splittable}
%
% \SetPitonStyle
% {
% Number = ,
% String = \itshape ,
% String.Doc = \color{gray} \slshape ,
-% Operator = ,
% Operator.Word = \bfseries ,
+% Operator = ,
% Name.Builtin = ,
% Name.Function = \bfseries \colorbox{gray!20} ,
% Comment = \color{gray} ,
-% Comment.LaTeX = \color{gray} ,
+% Comment.LaTeX = \normalfont \color{gray} ,
% Keyword = \bfseries ,
% Name.Namespace = ,
% Name.Class = ,
@@ -819,8 +861,10 @@ version 2005/12/01 or later.
% from math import pi
%
% def arctan(x,n=10):
-% '''Compute the value of arctan(x)
-% n is the number of terms if the sum'''
+% """Compute the mathematical value of arctan(x)
+%
+% n is the number of terms in the sum
+% """
% if x < 0:
% return -arctan(-x) # appel récursif
% elif x > 1:
@@ -837,12 +881,137 @@ version 2005/12/01 or later.
%
%
% \bigskip
-% The previous example has been composed while the key |splittable| (of
-% |\PitonOptions|) is in force. That's why a page break may have occurred.
% \clearpage
%
% \section{Implementation}
%
+% \subsection{Introduction}
+%
+% The main job of the package \pkg{piton} is to take in as input a Python
+% listing and to send back to LaTeX as output that code \emph{with interlaced LaTeX
+% instructions of formatting}.
+%
+% In fact, all that job is done by a \textsc{lpeg} called |SyntaxPython|. That
+% \textsc{lpeg}, when matched against the string of a Python listing,
+% returns as capture a Lua table containing data to send to LaTeX.
+% The only thing to do after will be to apply |tex.tprint| to each element of
+% that table.\footnote{Recall that |tex.tprint| takes in as argument a Lua table whose
+% first component is a ``catcode table'' and the second element a string. The
+% string will be sent to LaTeX with the regime of catcodes specified by the
+% catcode table. If no catcode table is provided, the standard catcodes of LaTeX
+% will be used.}
+%
+% \bigskip
+% Consider, for example, the following Python code:
+%
+% \begin{Piton}
+% def parity(x):
+% return x%2
+% \end{Piton}
+%
+% The capture returned by the \pkg{lpeg} |SyntaxPython| against that code is the
+% Lua table containing the following elements :
+%
+% \bigskip
+% \begin{minipage}{\linewidth}
+% \color{gray}
+%
+% |{ "\\__piton_begin_line:" }|\footnote{Each line of the Python listings will
+% be encapsulated in a pair: \texttt{\textbackslash_@@_begin_line:} --
+% \texttt{\textbackslash@@_end_line:}. The token
+% \texttt{\textbackslash@@_end_line:} must be explicit because it will be used as
+% marker in order to delimit the argument of the command \texttt{\textbackslash
+% @@\_begin\_line:}. Both tokens \texttt{\textbackslash_@@_begin_line:} and
+% \texttt{\textbackslash@@_end_line:} will be nullified in the command
+% \texttt{\textbackslash piton} (since there can't be lines breaks in the
+% argument of a command \texttt{\textbackslash piton}).}
+%
+% \texttt{\{ "\{\textbackslash PitonStyle\{Keyword\}\{" \}}\footnote{The
+% lexical elements of Python for which we have a \pkg{piton} style will be
+% formatted via the use of the command \texttt{\textbackslash PitonStyle}.
+% Such an element is typeset in LaTeX via the syntax \texttt{\{\textbackslash
+% PitonStyle\{\textsl{style}\}\{...\}\}} because the instructions inside an \texttt{\textbackslash
+% PitonStyle} may be both semi-global declarations like
+% \texttt{\textbackslash bfseries} and commands with one argument like
+% \texttt{\textbackslash fbox}.}
+%
+% \texttt{\{
+% luatexbase.catcodetables.CatcodeTableOther\footnote{\texttt{luatexbase.catcodetables.CatcodeTableOther} is a mere number which corresponds to the ``catcode table'' whose all characters have the catcode ``other'' (which means that they will be typeset by LaTeX verbatim).}, "def" \} }
+%
+% |{ "}}" }|
+%
+% |{ luatexbase.catcodetables.CatcodeTableOther, " " }|
+%
+% |{ "{\PitonStyle{Name.Function}{" }|
+%
+% |{ luatexbase.catcodetables.CatcodeTableOther, "parity" }|
+%
+% |{ "}}" }|
+%
+% |{ luatexbase.catcodetables.CatcodeTableOther, "(" }|
+%
+% |{ luatexbase.catcodetables.CatcodeTableOther, "x" }|
+%
+% |{ luatexbase.catcodetables.CatcodeTableOther, ")" }|
+%
+% |{ luatexbase.catcodetables.CatcodeTableOther, ":" }|
+%
+% |{ "\\__piton_end_line: \\__piton_newline: \\__piton_begin_line:" }|
+%
+% |{ luatexbase.catcodetables.CatcodeTableOther, " " }|
+%
+% |{ "{\PitonStyle{Keyword}{" }|
+%
+% |{ luatexbase.catcodetables.CatcodeTableOther, "return" }|
+%
+% |{ "}}" }|
+%
+% |{ luatexbase.catcodetables.CatcodeTableOther, " " }|
+%
+% |{ luatexbase.catcodetables.CatcodeTableOther, "x" }|
+%
+% |{ "{\PitonStyle{Operator}{" }|
+%
+% |{ luatexbase.catcodetables.CatcodeTableOther, "&" }|
+%
+% |{ "}}" }|
+%
+% |{ "{\PitonStyle{Number}{" }|
+%
+% |{ luatexbase.catcodetables.CatcodeTableOther, "2" }|
+%
+% |{ "}}" }|
+%
+% |{ "\\__piton_end_line:" }|
+%
+% \end{minipage}
+%
+% \bigskip
+% We give now the LaTeX code which is sent back by Lua to TeX (we have written
+% on several lines for legibility but no character |\r| will be sent to LaTeX). The
+% characters which are greyed-out are sent to LaTeX with the catcode ``other''
+% (=12). All the others characters are sent with the regime of catcodes of L3
+% (as set by |\ExplSyntaxOn|)
+%
+%
+% \begingroup
+% \def\gbox#1{\colorbox{gray!20}{\strut #1}}
+% \setlength{\fboxsep}{1pt}
+%
+% \begin{Verbatim*}[formatcom = \color{black}]
+% \__piton_begin_line:{\PitonStyle{Keyword}{~gbox#def@}}
+% ~gbox# @{\PitonStyle{Name.Function}{~gbox#parity@}}~gbox#(x):@\__piton_end_line:\__piton_newline:
+% \__piton_begin_line:~gbox# @{\PitonStyle{Keyword}{~gbox#return@}}
+% ~gbox# x@{\PitonStyle{Operator}{~gbox#%@}}{\PitonStyle{Number}{~gbox#2@}}\__piton_end_line:
+% \end{Verbatim*}
+% \endgroup
+%
+%
+%
+%
+% \subsection{The L3 part of the implementation}
+%
+% \subsubsection{Declaration of the package}
% \begin{macrocode}
\NeedsTeXFormat{LaTeX2e}
\RequirePackage{l3keys2e}
@@ -1010,9 +1179,23 @@ version 2005/12/01 or later.
% |\c_@@_footnote_bool| in order to know if we have to insert an environment
% |{savenotes}|.
%
-% \medskip
-% \subsection{Parameters}
+% \bigskip
+% \subsubsection{Parameters and technical definitions}
%
+% We will compute (with Lua) the numbers of lines of the Python code and store
+% it in the following counter.
+% \begin{macrocode}
+\int_new:N \l_@@_nb_lines_int
+% \end{macrocode}
+%
+% The following counter will be used to count the lines during the composition.
+% It will count all the lines, empty or not empty. It won't be used to print the
+% numbers of the lines.
+% \begin{macrocode}
+\int_new:N \g_@@_line_int
+% \end{macrocode}
+%
+% \medskip
% The following token list will contains the (potential) informations to write
% on the |aux| (to be used in the next compilation).
% \begin{macrocode}
@@ -1020,9 +1203,18 @@ version 2005/12/01 or later.
% \end{macrocode}
%
% \medskip
-% The following flag corresponds to the key |splittable| of |\PitonOptions|.
+% The following counter corresponds to the key |splittable| of |\PitonOptions|.
+% If the value of |\l_@@_splittable_int| is equal to $n$, then no line break can
+% occur within the first $n$~lines or the last $n$~lines of the listings.
% \begin{macrocode}
-\bool_new:N \l_@@_splittable_bool
+\int_new:N \l_@@_splittable_int
+% \end{macrocode}
+%
+% \medskip
+% An initial value of |splittable| equal to 100 is equivalent to say that the
+% environments |{Piton}| are unbreakable.
+% \begin{macrocode}
+\int_set:Nn \l_@@_splittable_int { 100 }
% \end{macrocode}
%
% \medskip
@@ -1033,9 +1225,12 @@ version 2005/12/01 or later.
%
% \medskip
% We will compute the maximal width of the lines of an environment |{Piton}| in
-% |\l_@@_width_dim|.
+% |\g_@@_width_dim|. We need a global variable because when the key |footnote|
+% is in force, each line when be composed in an environment |{savenotes}| and
+% (when |slim| is in force) we need to exit |\g_@@_width_dim| from that
+% environment.
% \begin{macrocode}
-\dim_new:N \l_@@_width_dim
+\dim_new:N \g_@@_width_dim
% \end{macrocode}
% The value of that dimension as written on the |aux| file will be stored in
% |\l_@@_width_on_aux_dim|.
@@ -1061,24 +1256,52 @@ version 2005/12/01 or later.
% \begin{macrocode}
\dim_new:N \l_@@_left_margin_dim
% \end{macrocode}
-%
-% \subsection{The gobbling mechanism}
%
% \medskip
-% The following integer is the number of characters to gobble on the left side
-% of the Python listings. Of course, the initial value is $0$.
+% The tabulators will be replaced by the content of the following token list.
% \begin{macrocode}
-\int_new:N \l_@@_gobble_int
+\tl_new:N \l_@@_tab_tl
% \end{macrocode}
%
% \medskip
% \begin{macrocode}
-\cs_new_protected:Npn \@@_define_gobble_syntax:n #1
- { \lua_now:n { define_gobble_syntax(#1) } }
+\cs_new_protected:Npn \@@_set_tab_tl:n #1
+ {
+ \tl_clear:N \l_@@_tab_tl
+ \prg_replicate:nn { #1 }
+ { \tl_put_right:Nn \l_@@_tab_tl { ~ } }
+ }
+\@@_set_tab_tl:n { 4 }
% \end{macrocode}
-%
+%
+% \medskip
+% \begin{macrocode}
+\cs_new_protected:Npn \@@_newline:
+ {
+ \int_gincr:N \g_@@_line_int
+ \int_compare:nNnT \g_@@_line_int > { \l_@@_splittable_int - 1 }
+ {
+ \int_compare:nNnT
+ { \l_@@_nb_lines_int - \g_@@_line_int } > \l_@@_splittable_int
+ {
+ \egroup
+ \bool_if:NT \c_@@_footnote_bool { \end { savenotes } }
+ \newline
+ \bool_if:NT \c_@@_footnote_bool { \begin { savenotes } }
+ \vtop \bgroup
+ }
+ }
+ }
+% \end{macrocode}
+%
% \medskip
-% \subsection{Treatment of a line of code}
+% The following integer corresponds to the key |gobble|.
+% \begin{macrocode}
+\int_new:N \l_@@_gobble_int
+% \end{macrocode}
+%
+% \bigskip
+% \subsubsection{Treatment of a line of code}
%
% In the contents provided by Lua, each line of the Python code will be
% surrounded by |\@@_begin_line:| and |\@@_end_line:|.
@@ -1086,8 +1309,6 @@ version 2005/12/01 or later.
% \begin{macrocode}
\cs_set_protected:Npn \@@_begin_line: #1 \@@_end_line:
{
- \bool_lazy_and:nnT \l_@@_splittable_bool \c_@@_footnote_bool
- { \begin { savenotes } }
% \end{macrocode}
%
% Be careful: there is curryfication in the following lines.
@@ -1115,11 +1336,11 @@ version 2005/12/01 or later.
#1 \hfil
}
% \end{macrocode}
-% We compute in |\l_@@_width_dim| the maximal width of the lines of the
+% We compute in |\g_@@_width_dim| the maximal width of the lines of the
% environments.
% \begin{macrocode}
- \dim_compare:nNnT { \box_wd:N \l_tmpa_box } > \l_@@_width_dim
- { \dim_set:Nn \l_@@_width_dim { \box_wd:N \l_tmpa_box } }
+ \dim_compare:nNnT { \box_wd:N \l_tmpa_box } > \g_@@_width_dim
+ { \dim_gset:Nn \g_@@_width_dim { \box_wd:N \l_tmpa_box } }
\box_set_dp:Nn \l_tmpa_box { \box_dp:N \l_tmpa_box + 1.25 pt }
\box_set_ht:Nn \l_tmpa_box { \box_ht:N \l_tmpa_box + 1.25 pt }
\tl_if_empty:NTF \l_@@_background_color_str
@@ -1139,15 +1360,13 @@ version 2005/12/01 or later.
\box_use_drop:N \l_tmpa_box
}
}
- \bool_lazy_and:nnT \l_@@_splittable_bool \c_@@_footnote_bool
- { \end { savenotes } }
\vspace { - 2.5 pt }
}
% \end{macrocode}
%
%
-% \medskip
-% \subsection{PitonOptions}
+% \bigskip
+% \subsubsection{PitonOptions}
%
% \medskip
% The following parameters correspond to the keys |line-numbers| and
@@ -1184,19 +1403,22 @@ version 2005/12/01 or later.
all-line-numbers .value_forbidden:n = true ,
resume .bool_set:N = \l_@@_resume_bool ,
resume .value_forbidden:n = true ,
- splittable .bool_set:N = \l_@@_splittable_bool ,
- splittable .default:n = true ,
+ splittable .int_set:N = \l_@@_splittable_int ,
+ splittable .default:n = 1 ,
background-color .str_set:N = \l_@@_background_color_str ,
background-color .value_required:n = true ,
slim .bool_set:N = \l_@@_slim_bool ,
slim .default:n = true ,
left-margin .dim_set:N = \l_@@_left_margin_dim ,
left-margin .value_required:n = true ,
+ tab-size .code:n = \@@_set_tab_tl:n { #1 } ,
+ tab-size .value_required:n = true ,
unknown .code:n =
\msg_error:nn { piton } { Unknown~key~for~PitonOptions }
}
% \end{macrocode}
%
+%
% \bigskip
% \begin{macrocode}
\msg_new:nnn { piton } { Unknown~key~for~PitonOptions }
@@ -1204,7 +1426,7 @@ version 2005/12/01 or later.
Unknown~key. \\
The~key~'\l_keys_key_str'~is~unknown~for~\token_to_str:N \PitonOptions.~The~
available~keys~are:~all-line-numbers,~auto-gobble,~env-gobble,~gobble,~
- left-margin,~line-numbers,~resume,~slim~and~splittable.\\
+ left-margin,~line-numbers,~resume,~slim~splittable~and~tab-size.\\
If~you~go~on,~that~key~will~be~ignored.
}
% \end{macrocode}
@@ -1215,32 +1437,33 @@ version 2005/12/01 or later.
\NewDocumentCommand \PitonOptions { } { \keys_set:nn { PitonOptions } }
% \end{macrocode}
%
-% \medskip
-% \subsection{The numbers of the lines}
+% \bigskip
+% \subsubsection{The numbers of the lines}
%
% \medskip
% The following counter will be used to count the lines in the code when the
-% user requires the numbers of the lines to be printed.
+% user requires the numbers of the lines to be printed (with |line-numbers| or
+% |all-line-numbers|).
% \begin{macrocode}
-\int_new:N \g_@@_line_int
+\int_new:N \g_@@_visual_line_int
% \end{macrocode}
%
%
% \begin{macrocode}
\cs_new_protected:Npn \@@_print_number:
{
- \int_gincr:N \g_@@_line_int
+ \int_gincr:N \g_@@_visual_line_int
\hbox_overlap_left:n
{
- { \color { gray } \footnotesize \int_to_arabic:n \g_@@_line_int }
+ { \color { gray } \footnotesize \int_to_arabic:n \g_@@_visual_line_int }
\skip_horizontal:n { 0.4 em }
}
}
% \end{macrocode}
%
%
-% \medskip
-% \subsection{The command to write on the aux file}
+% \bigskip
+% \subsubsection{The command to write on the aux file}
%
%
% \begin{macrocode}
@@ -1270,15 +1493,15 @@ version 2005/12/01 or later.
\tl_gput_right:Nx \g_@@_aux_tl
{
\dim_set:Nn \l_@@_width_on_aux_dim
- { \dim_use:N \l_@@_width_dim + 0.5 em }
+ { \dim_eval:n { \g_@@_width_dim + 0.5 em } }
}
}
}
}
% \end{macrocode}
%
-% \medskip
-% \subsection{The main commands and environments for the final user}
+% \bigskip
+% \subsubsection{The main commands and environments for the final user}
%
% \begin{macrocode}
\NewDocumentCommand { \piton } { v }
@@ -1305,29 +1528,48 @@ version 2005/12/01 or later.
}
% \end{macrocode}
%
-%
% \bigskip
+% Despite its name, |\@@_pre_env:| will be used both in |\PitonInputFile| dans
+% in the environments such as |{Piton}|.
% \begin{macrocode}
-\NewDocumentCommand { \PitonInputFile } { m }
+\cs_new:Npn \@@_pre_env:
{
\int_gincr:N \g_@@_env_int
\tl_gclear:N \g_@@_aux_tl
\tl_if_exist:cT { c_@@ _ \int_use:N \g_@@_env_int _ tl }
{ \use:c { c_@@ _ \int_use:N \g_@@_env_int _ tl } }
- \bool_if:NF \l_@@_splittable_bool
- { \bool_if:NT \c_@@_footnote_bool { \begin { savenotes } } }
\dim_compare:nNnT \l_@@_width_on_aux_dim = \c_zero_dim
{ \dim_set_eq:NN \l_@@_width_on_aux_dim \linewidth }
- \bool_if:NF \l_@@_resume_bool { \int_gzero:N \g_@@_line_int }
+ \bool_if:NF \l_@@_resume_bool { \int_gzero:N \g_@@_visual_line_int }
+ \dim_gzero:N \g_@@_width_dim
+ \int_gzero:N \g_@@_line_int
+ \dim_zero:N \parindent
+ \dim_zero:N \lineskip
+ }
+% \end{macrocode}
+%
+% \bigskip
+% \begin{macrocode}
+\NewDocumentCommand { \PitonInputFile } { m }
+ {
\group_begin:
- \dim_set_eq:NN \parindent \c_zero_dim
+ \@@_pre_env:
+ \mode_if_vertical:TF \mode_leave_vertical: \newline
+% \end{macrocode}
+% We count with Lua the number of lines of the argument. The result will be
+% stored by Lua in |\l_@@_nb_lines_int|. That information will be used to allow
+% or disallow page breaks.
+% \begin{macrocode}
+ \lua_now:n { CountLinesFile(token.scan_argument()) } { #1 }
\ttfamily
- \lua_now:e { ParseFile(token.scan_argument()) } { #1 }
- \@@_width_to_aux:
+ \bool_if:NT \c_@@_footnote_bool { \begin { savenotes } }
+ \vtop \bgroup
+ \lua_now:n { ParseFile(token.scan_argument()) } { #1 }
+ \egroup
+ \bool_if:NT \c_@@_footnote_bool { \end { savenotes } }
+ \@@_width_to_aux:
\group_end:
\@@_write_aux:
- \bool_if:NF \l_@@_splittable_bool
- { \bool_if:NT \c_@@_footnote_bool { \end { savenotes } } }
}
% \end{macrocode}
%
@@ -1353,42 +1595,26 @@ version 2005/12/01 or later.
}
{
\group_end:
- \mode_if_vertical:TF
- { \mode_leave_vertical: }
- \newline
+ \mode_if_vertical:TF \mode_leave_vertical: \newline
% \end{macrocode}
-% Be careful: there is curryfication in the following code.
+% We count with Lua the number of lines of the argument. The result will be
+% stored by Lua in |\l_@@_nb_lines_int|. That information will be used to allow
+% or disallow page breaks.
% \begin{macrocode}
- \bool_if:NF \l_@@_splittable_bool
+ \lua_now:n { CountLines(token.scan_argument()) } { ##1 }
+ \ttfamily
+ \bool_if:NT \c_@@_footnote_bool { \begin { savenotes } }
+ \vtop \bgroup
+ \lua_now:e
{
- \bool_if:NT \c_@@_footnote_bool { \begin { savenotes } }
- \vbox_top:n
- }
- {
- \ttfamily
- \dim_zero:N \lineskip
- \int_case:nnF \l_@@_gobble_int
- {
- 0
-% \end{macrocode}
-% Be careful: the last argument is provided by curryfication.
-% \begin{macrocode}
- { \lua_now:n { Parse(token.scan_argument()) } }
- { -1 }
- { \lua_now:n { AutoGobbleParse(token.scan_argument()) } }
- { -2 }
- { \lua_now:n { EnvGobbleParse(token.scan_argument()) } }
- }
- {
- \exp_args:NV \@@_define_gobble_syntax:n \l_@@_gobble_int
- \lua_now:n { GobbleParse(token.scan_argument()) }
- }
- { ##1 }
- \vspace { 2.5 pt }
- \@@_width_to_aux:
- }
- \bool_if:NF \l_@@_splittable_bool
- { \bool_if:NT \c_@@_footnote_bool { \end { savenotes } } }
+ GobbleParse
+ ( \int_use:N \l_@@_gobble_int , token.scan_argument() )
+ }
+ { ##1 }
+ \vspace { 2.5 pt }
+ \egroup
+ \bool_if:NT \c_@@_footnote_bool { \end { savenotes } }
+ \@@_width_to_aux:
% \end{macrocode}
% The following |\end{#1}| is only for the groups and the stack of
% environments of LaTeX.
@@ -1407,17 +1633,10 @@ version 2005/12/01 or later.
\NewDocumentEnvironment { #1 } { #2 }
{
#3
- \int_gincr:N \g_@@_env_int
- \tl_gclear:N \g_@@_aux_tl
- \tl_if_exist:cT { c_@@ _ \int_use:N \g_@@_env_int _ tl }
- { \use:c { c_@@ _ \int_use:N \g_@@_env_int _ tl } }
- \dim_compare:nNnT \l_@@_width_on_aux_dim = \c_zero_dim
- { \dim_set_eq:NN \l_@@_width_on_aux_dim \linewidth }
- \bool_if:NF \l_@@_resume_bool { \int_gzero:N \g_@@_line_int }
+ \@@_pre_env:
\group_begin:
- \box_clear:N \l_tmpa_box
\tl_map_function:nN
- { \ \\ \{ \} \$ \& \# \^ \_ \% \~ }
+ { \ \\ \{ \} \$ \& \# \^ \_ \% \~ \^^I }
\char_set_catcode_other:N
\use:c { _@@_collect_ #1 :w }
}
@@ -1443,7 +1662,7 @@ version 2005/12/01 or later.
% \end{macrocode}
%
% \bigskip
-% \subsection{The styles}
+% \subsubsection{The styles}
%
% \medskip
% The following command is fundamental: it will be used by the Lua code.
@@ -1553,8 +1772,8 @@ version 2005/12/01 or later.
}
% \end{macrocode}
%
-%
-% \subsection{The initial style}
+% \bigskip
+% \subsubsection{The initial style}
%
% The initial style is inspired by the style ``manni'' of Pygments.
%
@@ -1581,12 +1800,15 @@ version 2005/12/01 or later.
Name.Type = \color[HTML]{336666} ,
InitialValues = \@@_piton:n ,
Dict.Value = \@@_piton:n ,
- Post.Function = \@@_piton:n ,
Interpol.Inside = \color{black}\@@_piton:n ,
+ Post.Function = \@@_piton:n ,
}
% \end{macrocode}
+% The last style |Post.Function| should be considered as an ``internal style''
+% (not available for the final user).
%
-% \subsection{Security}
+% \bigskip
+% \subsubsection{Security}
%
% \begin{macrocode}
\AddToHook { env / piton / begin }
@@ -1603,371 +1825,739 @@ version 2005/12/01 or later.
%
%
% \bigskip
-% \subsection{The Lua code}
+% \subsection{The Lua part of the implementation}
%
-% \medskip
+% \bigskip
% \begin{macrocode}
\ExplSyntaxOff
\RequirePackage{luacode}
% \end{macrocode}
%
-% \vspace{1cm}
+% \medskip
% \begin{macrocode}
\begin{luacode*}
-local P, S, V , C , Ct, Cc, Cf = lpeg.P, lpeg.S, lpeg.V, lpeg.C, lpeg.Ct, lpeg.Cc, lpeg.Cf
-
-
---[[ By convention, a capture which provides as value a table (and not a string), provides, in fact,
-a string (the first element of the table) which is a formatting LaTeX instruction (it will be
-thrown back to TeX with normal catcodes (ant not ``other'' catcode for everybody).]]
-
-local function L(string)
- return Cc ( { string } )
+% \end{macrocode}
+%
+% \bigskip
+% \subsubsection{Special functions dealing with LPEG}
+%
+% \medskip
+% We will use the Lua library \pkg{lpeg} which is built in LuaTeX. That's why we
+% define first aliases for several functions of that library.
+% \begin{macrocode}
+local P, S, V, C, Ct, Cc, Cf = lpeg.P, lpeg.S, lpeg.V, lpeg.C, lpeg.Ct, lpeg.Cc, lpeg.Cf
+% \end{macrocode}
+%
+%
+% \bigskip
+% The function |Q| takes in as argument a pattern and returns a \textsc{lpeg}
+% \emph{which does a capture} of the pattern. That capture will be sent to LaTeX
+% with the catcode ``other'' for all the characters: it's suitable for elements
+% of the Python listings that \pkg{piton} will typeset verbatim (thanks to the
+% catcode ``other''). That function will be widely used.
+% \begin{macrocode}
+local function Q(pattern)
+ return Ct ( Cc ( luatexbase.catcodetables.CatcodeTableOther ) * C ( pattern ) )
end
-
+% \end{macrocode}
+%
+% \bigskip
+% The function |L| takes in as argument a pattern and returns a \textsc{lpeg}
+% \emph{which does a capture} of the pattern. That capture will be sent to LaTeX
+% with standard LaTeX catcodes for all the characters: the elements captured
+% will be formatted as normal LaTeX codes. It's suitable for the ``comment
+% LateX'' in the environments |{Piton}| and the elements beetween
+% ``|escape-inside|''. That function won't be much used.
+% \begin{macrocode}
+local function L(pattern)
+ return Ct ( C ( pattern ) )
+end
+% \end{macrocode}
+%
+% \bigskip
+% The function |Lc| (the c is for \emph{constant}) takes in as argument a string
+% and returns a \textsc{lpeg} \emph{with does a constant capture} which returns
+% that string. The elements captured will be formatted as L3 code. It will be
+% used to send to LaTeX all the formatting LaTeX instructions we have to insert
+% in order to do the syntactic highlighting (that's the main job of
+% \pkg{piton}). That function will be widely used.
+% \begin{macrocode}
+local function Lc(string)
+ return Cc ( { luatexbase.catcodetables.expl , string } )
+end
+% \end{macrocode}
+%
+% \bigskip
+% The function |K| function creates a \textsc{lpeg} which will return as capture
+% the whole LaTeX code corresponding to a Python chunk (that is to say with the
+% LaTeX formatting instructions corresponding to the syntactic nature of that
+% Python chunk). The first argument is a pattern (that is to say a \textsc{lpeg}
+% without capture) and the second element is a Lua string corresponding to the
+% name of a \pkg{piton} style.
+% \begin{macrocode}
local function K(pattern, style)
return
- L ( "{\\PitonStyle{" .. style .. "}{" )
- * C(pattern)
- * L ( "}}" )
+ Lc ( "{\\PitonStyle{" .. style .. "}{" )
+ * Q ( pattern )
+ * Lc ( "}}" )
end
-
-
---[[ The text in "escape" (between begin_escape and end_escape) is captured
-and put in a table (with only one component). Indeed, we have decided that a capture
-which is encapsulated in a table must be transmitted to TeX with the normal TeX catcodes.]]
-
-local Escape = P(begin_escape)
- * Ct ( C ( ( 1 - P(end_escape) ) ^ 1 ) )
- * P(end_escape)
-
-
-lpeg.locale(lpeg) -- mandatory
-
-local alpha , digit , space , punct = lpeg.alpha , lpeg.digit , lpeg.space , lpeg.punct
-
--- Remember that à, â, ç, etc. are strings of length 2 (2 bytes)
+% \end{macrocode}
+% The formatting commands in a given \pkg{piton} style (eg. the style |Keyword|)
+% may be semi-global declarations (such as |\bfseries| or |\slshape|) or LaTeX
+% macros with an argument (such as |\fbox| or |\colorbox{yellow}|). In order to
+% deal with both syntaxes, we have used two pairs of braces:
+% |{\PitonStyle{Keyword}{|\texttt{\slshape text to format}|}}|.
+%
+% \bigskip
+% The following \textsc{lpeg} catches the Python chunks which are in LaTeX
+% escapes (and that chunks will be considered as normal LaTeX constructions). We
+% recall that |begin_espace| and |end_escape| are Lua strings corresponding to
+% the key |escape-inside|\footnote{The \pkg{piton} key |escape-inside| is
+% available at load-time only.}. Since the elements that will be catched must be
+% sent to LaTeX with standard LaTeX catcodes, we put the capture (done by the
+% function |C|) in a table (by using |Ct|, which is an alias for |lpeg.Ct|)
+% without number of catcode table at the first component of the table.
+% \begin{macrocode}
+local Escape =
+ P(begin_escape)
+ * L ( ( 1 - P(end_escape) ) ^ 1 )
+ * P(end_escape)
+% \end{macrocode}
+%
+% \vspace{1cm}
+% The following line is mandatory.
+% \begin{macrocode}
+lpeg.locale(lpeg)
+% \end{macrocode}
+%
+% \bigskip
+% \subsubsection{The LPEG SyntaxPython}
+%
+% \begin{macrocode}
+local alpha, digit, space = lpeg.alpha, lpeg.digit, lpeg.space
+% \end{macrocode}
+%
+% Remember that, for \textsc{lpeg}, the Unicode characters such as |à|, |â|,
+% |ç|, etc. are in fact strings of length 2 (2 bytes) because \pkg{lpeg} is not
+% Unicode-aware.
+% \begin{macrocode}
local letter = alpha + P "_"
- + P "â" + P "à" + P "ç" + P "é" + P "è" + P "ê" + P "ë" + P "ï" + P "î" + P "ô" + P "û" + P "ü" +
- P "Â" + P "À" + P "Ç" + P "É" + P "È" + P "Ê" + P "Ë" + P "Ï" + P "Î" + P "Ô" + P "Û" + P "Ü"
+ + P "â" + P "à" + P "ç" + P "é" + P "è" + P "ê" + P "ë" + P "ï" + P "î"
+ + P "ô" + P "û" + P "ü" + P "Â" + P "À" + P "Ç" + P "É" + P "È" + P "Ê"
+ + P "Ë" + P "Ï" + P "Î" + P "Ô" + P "Û" + P "Ü"
local alphanum = letter + digit
-
+% \end{macrocode}
+%
+% \bigskip
+% The following \textsc{lpeg} |identifier| is a mere pattern (that is to say
+% more or less a regular expression) which matches the Python identifiers (hence
+% the name).
+% \begin{macrocode}
local identifier = letter * alphanum ^ 0
-
-local Identifier = C ( identifier )
-
-local Space = C ( ( space - P "\r" ) ^ 1 )
-
-local SkipSpace = C ( ( space - P "\r" ) ^ 0 )
-
-local Punct = C ( punct )
-
-
-local EOL = ( P "\r" )
- *
- (
- ( space^0 * -1 )
- +
- Cc (
- {
- luatexbase.catcodetables.expl ,
- '\\__piton_end_line: \\bool_if:NT \\l__piton_splittable_bool \\newline \\__piton_begin_line:'
- }
- )
- )
-
-
+% \end{macrocode}
+%
+% \medskip
+% On the other hand, the \textsc{lpeg} |Identifier| (with a capital) also return
+% a \emph{capture}. Since no special LaTeX formatting will be applied to the
+% Python identifiers, we use the function~|Q| and not the function~|K|. For
+% elements which require formatting, we will usually use our function~|K|
+% instead of the function~|C|. See just below for an example of use of the
+% function~|K|.
+% \begin{macrocode}
+local Identifier = Q ( identifier )
+% \end{macrocode}
+%
+% \bigskip
+% By convention, we will use names with an initial capital for \textsc{lpeg}
+% which return captures.
+%
+%
+% \bigskip
+% Here is the first use of our function~|K|. That function will be used to
+% construct \textsc{lpeg} which capture Python chunks for which we have a
+% dedicated \pkg{piton} style. For example, for the numbers, \pkg{piton}
+% provides a style which is called |Number|. The name of the style is provided
+% as a Lua string in the second argument of the function~|K|. By convention, we
+% use single quotes for delimiting the Lua strings which are names of
+% \pkg{piton} styles (but this is only a convention).
+% \begin{macrocode}
local Number =
- K (
- ( digit^1 * P "." * digit^0 + digit^0 * P "." * digit^1 + digit^1 )
- * ( S "eE" * S "+-" ^ -1 * digit^1 ) ^ -1
- + digit^1
- , 'Number' )
-
-
-local Word = C ( ( ( 1 - space ) - S "'\"\r[()]" - digit ) ^ 1 )
-
+ K (
+ ( digit^1 * P "." * digit^0 + digit^0 * P "." * digit^1 + digit^1 )
+ * ( S "eE" * S "+-" ^ -1 * digit^1 ) ^ -1
+ + digit^1 ,
+ 'Number'
+ )
+% \end{macrocode}
+%
+% \bigskip
+% We recall that |begin_espace| and |end_escape| are Lua strings
+% corresponding to the key |escape-inside|\footnote{The \pkg{piton} key
+% |escape-inside| is available at load-time only.}. Of course, if the final user
+% has not used the key |escape-inside|, these strings are empty.
+% \begin{macrocode}
if begin_escape ~= ''
-then Word = C ( ( ( 1 - space - P(begin_escape) - P(end_escape) ) - S "'\"\r[()]" - digit ) ^ 1 )
+then Word = Q ( ( ( 1 - space - P(begin_escape) - P(end_escape) )
+ - S "'\"\r[()]" - digit ) ^ 1 )
+else Word = Q ( ( ( 1 - space ) - S "'\"\r[()]" - digit ) ^ 1 )
end
+% \end{macrocode}
+%
+% \bigskip
+% \begin{macrocode}
+local Space = Q ( ( space - P "\r" ) ^ 1 )
+
+local SkipSpace = Q ( ( space - P "\r" ) ^ 0 )
+
+local Punct = Q ( S ".,:;!" )
+% \end{macrocode}
+%
+% \bigskip
+% \begin{macrocode}
+local Tab = P "\t" * Lc ( '\\l_@@_tab_tl' )
+% \end{macrocode}
+%
+% \bigskip
+% The following \textsc{lpeg} |EOL| is for the end of lines.
+% \begin{macrocode}
+local EOL =
+ P "\r"
+ *
+ (
+ ( space^0 * -1 )
+ +
+% \end{macrocode}
+% We recall that each line in the Python code we have to parse will be sent
+% back to LaTeX between a pair |\@@_begin_line:| --
+% |\@@_end_line:|\footnote{Remember that the \texttt{\textbackslash
+% @@\_end\_line:} must be explicit because it will be used as marker in order to
+% delimit the argument of the command \texttt{\textbackslash @@\_begin\_line:}}.
+% \begin{macrocode}
+ Lc ( '\\@@_end_line: \\@@_newline: \\@@_begin_line:' )
+ )
+% \end{macrocode}
+%
+% \bigskip
+% \begin{macrocode}
+local Delim = Q ( S "[()]" )
-local Delim = C ( S "[()]" )
+local Operator =
+ K ( P "!=" + P "==" + P "<<" + P ">>" + S "-~+/*%=<>&.@|" , 'Operator')
+local OperatorWord =
+ K ( P "in" + P "is" + P "and" + P "or" + P "not" , 'Operator.Word')
local Keyword =
- K ( P "assert" + P "break" + P "case" + P "continue" + P "del"
- + P "elif" + P "else" + P "except" + P "exec" + P "finally" + P "for" + P "from"
- + P "global" + P "if" + P "import" + P "lambda" + P "non local"
- + P "pass" + P "return" + P "try" + P "while"
- + P "with" + P "yield" + P "yield from" ,
- 'Keyword' )
- + K ( P "True" + P "False" + P "None" , 'Keyword.Constant' )
-
+ K ( P "as" + P "assert" + P "break" + P "case" + P "class" + P "continue"
+ + P "def" + P "del" + P "elif" + P "else" + P "except" + P "exec"
+ + P "finally" + P "for" + P "from" + P "global" + P "if" + P "import"
+ + P "lambda" + P "non local" + P "pass" + P "return" + P "try"
+ + P "while" + P "with" + P "yield" + P "yield from" ,
+ 'Keyword' )
+ + K ( P "True" + P "False" + P "None" , 'Keyword.Constant' )
local Builtin =
- K ( P "__import__" + P "abs" + P "all" + P "any" + P "bin" + P "bool" + P "bytearray"
- + P "bytes" + P "chr" + P "classmethod" + P "compile" + P "complex" + P "delattr"
- + P "dict" + P "dir" + P "divmod" + P "enumerate" + P "eval" + P "filter"
- + P "float" + P "format" + P "frozenset" + P "getattr" + P "globals" + P "hasattr"
- + P "hash" + P "hex" + P "id" + P "input" + P "int" + P "isinstance" + P "issubclass"
- + P "iter" + P "len" + P "list" + P "locals" + P "map" + P "max" + P "memoryview" + P "min"
- + P "next" + P "object" + P "oct" + P "open" + P "ord" + P "pow" + P "print" + P "property"
- + P "range" + P "repr" + P "reversed" + P "round" + P "set" + P "setattr" + P "slice"
- + P "sorted" + P "staticmethod" + P "str" + P "sum" + P "super" + P "tuple" + P "type"
- + P "vars" + P "zip" ,
- 'Name.Builtin' )
-
+ K ( P "__import__" + P "abs" + P "all" + P "any" + P "bin" + P "bool"
+ + P "bytearray" + P "bytes" + P "chr" + P "classmethod" + P "compile"
+ + P "complex" + P "delattr" + P "dict" + P "dir" + P "divmod"
+ + P "enumerate" + P "eval" + P "filter" + P "float" + P "format"
+ + P "frozenset" + P "getattr" + P "globals" + P "hasattr" + P "hash"
+ + P "hex" + P "id" + P "input" + P "int" + P "isinstance" + P "issubclass"
+ + P "iter" + P "len" + P "list" + P "locals" + P "map" + P "max"
+ + P "memoryview" + P "min" + P "next" + P "object" + P "oct" + P "open"
+ + P "ord" + P "pow" + P "print" + P "property" + P "range" + P "repr"
+ + P "reversed" + P "round" + P "set" + P "setattr" + P "slice" + P "sorted"
+ + P "staticmethod" + P "str" + P "sum" + P "super" + P "tuple" + P "type"
+ + P "vars" + P "zip" ,
+ 'Name.Builtin' )
local Exception =
- K ( "ArithmeticError" + P "AssertionError" + P "AttributeError"
- + P "BaseException" + P "BufferError" + P "BytesWarning" + P "DeprecationWarning"
- + P "EOFError" + P "EnvironmentError" + P "Exception" + P "FloatingPointError"
- + P "FutureWarning" + P "GeneratorExit" + P "IOError" + P "ImportError"
- + P "ImportWarning" + P "IndentationError" + P "IndexError" + P "KeyError"
- + P "KeyboardInterrupt" + P "LookupError" + P "MemoryError" + P "NameError"
- + P "NotImplementedError" + P "OSError" + P "OverflowError"
- + P "PendingDeprecationWarning" + P "ReferenceError" + P "ResourceWarning"
- + P "RuntimeError" + P "RuntimeWarning" + P "StopIteration"
- + P "SyntaxError" + P "SyntaxWarning" + P "SystemError" + P "SystemExit"
- + P "TabError" + P "TypeError" + P "UnboundLocalError" + P "UnicodeDecodeError"
- + P "UnicodeEncodeError" + P "UnicodeError" + P "UnicodeTranslateError"
- + P "UnicodeWarning" + P "UserWarning" + P "ValueError" + P "VMSError"
- + P "Warning" + P "WindowsError" + P "ZeroDivisionError"
- + P "BlockingIOError" + P "ChildProcessError" + P "ConnectionError"
- + P "BrokenPipeError" + P "ConnectionAbortedError" + P "ConnectionRefusedError"
- + P "ConnectionResetError" + P "FileExistsError" + P "FileNotFoundError"
- + P "InterruptedError" + P "IsADirectoryError" + P "NotADirectoryError"
- + P "PermissionError" + P "ProcessLookupError" + P "TimeoutError"
- + P "StopAsyncIteration" + P "ModuleNotFoundError" + P "RecursionError" ,
- 'Exception' )
-
-local RaiseException = K ( P "raise" , 'Keyword' ) * SkipSpace * Exception * C ( P "(" )
-
-local ExceptionInConsole = Exception * C ( ( 1 - P "\r" ) ^ 0 ) * EOL
-
-
-local Namespace =
- K ( P "from" , 'Keyword' ) * Space * K ( alphanum^1 , 'Name.Namespace' )
- * ( Space * K ( P "import" , 'Keyword' ) ) ^ -1
-
-
-local ImportAs = K ( P "import" , 'Keyword' )
- * Space
- * K ( identifier , 'Name.Namespace' )
- * ( SkipSpace * C ( P "," ) * SkipSpace * K ( identifier , 'Name.Namespace' ) ) ^ 0
- * (
- Space * K ( P "as" , 'Keyword' ) * Space * K ( identifier , 'Name.Namespace' )
- ) ^ 0
-
-local Class = K ( P "class" , 'Keyword' )
- * ( Space * K ( identifier , 'Name.Class' ) ) ^ -1
-
+ K ( "ArithmeticError" + P "AssertionError" + P "AttributeError"
+ + P "BaseException" + P "BufferError" + P "BytesWarning" + P "DeprecationWarning"
+ + P "EOFError" + P "EnvironmentError" + P "Exception" + P "FloatingPointError"
+ + P "FutureWarning" + P "GeneratorExit" + P "IOError" + P "ImportError"
+ + P "ImportWarning" + P "IndentationError" + P "IndexError" + P "KeyError"
+ + P "KeyboardInterrupt" + P "LookupError" + P "MemoryError" + P "NameError"
+ + P "NotImplementedError" + P "OSError" + P "OverflowError"
+ + P "PendingDeprecationWarning" + P "ReferenceError" + P "ResourceWarning"
+ + P "RuntimeError" + P "RuntimeWarning" + P "StopIteration"
+ + P "SyntaxError" + P "SyntaxWarning" + P "SystemError" + P "SystemExit"
+ + P "TabError" + P "TypeError" + P "UnboundLocalError" + P "UnicodeDecodeError"
+ + P "UnicodeEncodeError" + P "UnicodeError" + P "UnicodeTranslateError"
+ + P "UnicodeWarning" + P "UserWarning" + P "ValueError" + P "VMSError"
+ + P "Warning" + P "WindowsError" + P "ZeroDivisionError"
+ + P "BlockingIOError" + P "ChildProcessError" + P "ConnectionError"
+ + P "BrokenPipeError" + P "ConnectionAbortedError" + P "ConnectionRefusedError"
+ + P "ConnectionResetError" + P "FileExistsError" + P "FileNotFoundError"
+ + P "InterruptedError" + P "IsADirectoryError" + P "NotADirectoryError"
+ + P "PermissionError" + P "ProcessLookupError" + P "TimeoutError"
+ + P "StopAsyncIteration" + P "ModuleNotFoundError" + P "RecursionError" ,
+ 'Exception' )
+
+local RaiseException = K ( P "raise" , 'Keyword' ) * SkipSpace * Exception * Q ( P "(" )
+
+local ExceptionInConsole = Exception * Q ( ( 1 - P "\r" ) ^ 0 ) * EOL
+% \end{macrocode}
+%
+% \bigskip
+% In Python, a ``decorator'' is a statement whose begins by |@| which patches
+% the function defined in the following statement.
+% \begin{macrocode}
local Decorator = K ( P "@" * letter^1 , 'Name.Decorator' )
-
+% \end{macrocode}
+%
+% \bigskip
+% The following \textsc{lpeg} |DefClass| will be used to detect the definition of a
+% new class (the name of that new class will be formatted with the \pkg{piton}
+% style |Name.Class|).
+%
+% \smallskip
+% Example:\enskip \piton|class myclass:|
+% \begin{macrocode}
+local DefClass =
+ K ( P "class" , 'Keyword' ) * Space * K ( identifier , 'Name.Class' )
+% \end{macrocode}
+%
+% If the word |class| is not followed by a identifier, it will be catched as
+% keyword by the \textsc{lpeg} |Keyword| (useful if we want to type a
+% list of keywords).
+%
+%
+% \bigskip
+% The following \textsc{lpeg} |ImportAs| is used for the lines beginning by |import|.
+% % We have to detect the potential keyword |as| because both the name of the
+% module and its alias must be formatted with the \pkg{piton} style |Name.Namespace|.
+%
+% \smallskip
+% Example:\enskip \piton|import numpy as np|
+%
+% \smallskip
+% Moreover, after the keyword |import|, it's possible to have a comma-separated
+% list of modules (if the keyword |as| is not used).
+%
+% \smallskip
+% Example:\enskip \piton|import math, numpy|
+% \begin{macrocode}
+local ImportAs =
+ K ( P "import" , 'Keyword' )
+ * Space
+ * K ( identifier * ( P "." * identifier ) ^ 0 ,
+ 'Name.Namespace'
+ )
+ * (
+ ( Space * K ( P "as" , 'Keyword' ) * Space * K ( identifier , 'Name.Namespace' ) )
+ +
+ ( SkipSpace * Q ( P "," ) * SkipSpace * K ( identifier , 'Name.Namespace' ) ) ^ 0
+ )
+% \end{macrocode}
+% Be careful: there is no commutativity of |+| in the previous expression.
+%
+% \bigskip
+% The \textsc{lpeg} |FromImport| is used for the lines beginning by |from|. We
+% need a special treatment because the identifier following the keyword |from|
+% must be formatted with the \pkg{piton} style |Name.Namespace| and the
+% following keyword |import| must be formatted with the \pkg{piton} style
+% |Keyword| and must \emph{not} be catched by the \textsc{lpeg} |ImportAs|.
+%
+% \smallskip
+% Example:\enskip \piton|from math import pi|
+%
+% \smallskip
+% \begin{macrocode}
+local FromImport =
+ K ( P "from" , 'Keyword' )
+ * Space * K ( identifier , 'Name.Namespace' )
+ * Space * K ( P "import" , 'Keyword' )
+% \end{macrocode}
+%
+% \bigskip
+% \paragraph{The strings of Python}
+%
+% For the strings in Python, there are four categories of delimiters (without
+% counting the prefixes for f-strings and raw strings). We will use, in the
+% names of our \textsc{lpeg}, prefixes to distinguish the \textsc{lpeg} dealing
+% with that categories of strings, as presented in the following tabular.
+% \begin{center}
+% \begin{tabular}{ccc}
+% \toprule
+% & |Single| & |Double| \\
+% \midrule
+% |Short| & |'text'| & |"text"| \\
+% |Long| & |'''test'''| & |"""text"""| \\
+% \bottomrule
+% \end{tabular}
+% \end{center}
+%
+%
+% \bigskip
+% First, we define \textsc{lpeg} for the interpolations in the f-strings. Here
+% is an example of a f-string with an interpolation and a format
+% instruction\footnote{There is no special \pkg{piton} style for the formatting
+% instruction (after the comma): the style which will be applied will be the
+% style of the encompassing string, that is to say |String.Short| or
+% |String.Long|.} in that interpolation:
+%
+% \piton|f'Total price: {total+1:.2f} €'|
+%
+% \medskip
+% The following \textsc{lpeg} |SingleShortInterpol| (and the three variants)
+% will catch the whole interpolation, included the braces, that is to say, in
+% the previous example:\enskip |{total+1:.2f}|
+%
+% \medskip
+% \begin{macrocode}
local SingleShortInterpol =
K ( P "{" , 'String.Interpol')
* K ( ( 1 - S "}':" ) ^ 0 , 'Interpol.Inside' )
- * C ( P ":" * (1 - S "}:'") ^ 0 ) ^ -1
+ * Q ( P ":" * (1 - S "}:'") ^ 0 ) ^ -1
* K ( P "}" , 'String.Interpol' )
local DoubleShortInterpol =
K ( P "{" , 'String.Interpol' )
* K ( ( 1 - S "}\":" ) ^ 0 , 'Interpol.Inside' )
- * ( K ( P ":" , 'String.Interpol' ) * C ( (1 - S "}:\"") ^ 0 ) ) ^ -1
+ * ( K ( P ":" , 'String.Interpol' ) * Q ( (1 - S "}:\"") ^ 0 ) ) ^ -1
* K ( P "}" , 'String.Interpol' )
local SingleLongInterpol =
K ( P "{" , 'String.Interpol' )
* K ( ( 1 - S "}:\r" - P "'''" ) ^ 0 , 'Interpol.Inside' )
- * C ( P ":" * (1 - S "}:\r" - P "'''" ) ^ 0 ) ^ -1
+ * Q ( P ":" * (1 - S "}:\r" - P "'''" ) ^ 0 ) ^ -1
* K ( P "}" , 'String.Interpol' )
local DoubleLongInterpol =
K ( P "{" , 'String.Interpol' )
* K ( ( 1 - S "}:\r" - P "\"\"\"" ) ^ 0 , 'Interpol.Inside' )
- * C ( P ":" * (1 - S "}:\r" - P "\"\"\"" ) ^ 0 ) ^ -1
+ * Q ( P ":" * (1 - S "}:\r" - P "\"\"\"" ) ^ 0 ) ^ -1
* K ( P "}" , 'String.Interpol' )
+% \end{macrocode}
+%
+% \bigskip
+% Now, we define \textsc{lpeg} for the parts of the strings which are \emph{not}
+% in the interpolations.
+% \begin{macrocode}
+local SingleShortPureString =
+ Q ( ( P "\\'" + P "{{" + P "}}" + 1 - S "{}'" ) ^ 1 )
-local SingleShortPureString = C ( ( P "\\'" + P "{{" + P "}}" + 1 - S "{}'" ) ^ 1 )
-
-local DoubleShortPureString = C ( ( P "\\\"" + P "{{" + P "}}" + 1 - S "{}\"" ) ^ 1 )
-
-local SingleLongPureString = C ( ( 1 - P "'''" - S "{}'\r" ) ^ 1 )
+local DoubleShortPureString =
+ Q ( ( P "\\\"" + P "{{" + P "}}" + 1 - S "{}\"" ) ^ 1 )
-local DoubleLongPureString = C ( ( 1 - P "\"\"\"" - S "{}\"\r" ) ^ 1 )
+local SingleLongPureString =
+ Q ( ( 1 - P "'''" - S "{}'\r" ) ^ 1 )
+local DoubleLongPureString =
+ Q ( ( 1 - P "\"\"\"" - S "{}\"\r" ) ^ 1 )
+% \end{macrocode}
+%
+% \bigskip
+% We can now define the \textsc{lpeg} for the four kinds of strings. It's not
+% possible to use our function~|K| because of the interpolations which must be
+% formatted with another \pkg{piton} style that the rest of the
+% string.\footnote{The interpolations are formatted with the \pkg{piton} style
+% |Interpol.Inside|. The initial value of that style is \texttt{\textbackslash
+% @@\_piton:n} wich means that the interpolations are parsed once again by \pkg{piton}.}
+% \begin{macrocode}
local SingleShortString =
- L ( "{\\PitonStyle{String.Short}{" )
- * (
- C ( P "f'" + P "F'" )
- * ( SingleShortInterpol + SingleShortPureString ) ^ 0
- * C ( P "'" )
- +
- C ( ( P "'" + P "r'" + P "R'" ) * ( P "\\'" + 1 - S "'\r" ) ^ 0 * P "'" )
- )
- * L ( "}}" )
+ Lc ( "{\\PitonStyle{String.Short}{" )
+ * (
+% \end{macrocode}
+% First, we deal with the f-strings of Python, which are prefixed by |f| or |F|.
+% \begin{macrocode}
+ Q ( P "f'" + P "F'" )
+ * ( SingleShortInterpol + SingleShortPureString ) ^ 0
+ * Q ( P "'" )
+ +
+% \end{macrocode}
+% Now, we deal with the standard strings of Python, but also the ``raw strings''.
+% \begin{macrocode}
+ Q ( ( P "'" + P "r'" + P "R'" )
+ * ( P "\\'" + 1 - S "'\r" ) ^ 0 * P "'" )
+ )
+ * Lc ( "}}" )
local DoubleShortString =
- L ( "{\\PitonStyle{String.Short}{" )
- * (
- C ( P "f\"" + P "F\"" )
- * ( DoubleShortInterpol + DoubleShortPureString ) ^ 0
- * C ( P "\"" )
- +
- C ( ( P "\"" + P "r\"" + P "R\"" ) * ( P "\\\"" + 1 - S "\"\r" ) ^ 0 * P "\"" )
- )
- * L ( "}}" )
+ Lc ( "{\\PitonStyle{String.Short}{" )
+ * (
+ Q ( P "f\"" + P "F\"" )
+ * ( DoubleShortInterpol + DoubleShortPureString ) ^ 0
+ * Q ( P "\"" )
+ +
+ Q ( ( P "\"" + P "r\"" + P "R\"" )
+ * ( P "\\\"" + 1 - S "\"\r" ) ^ 0 * P "\"" )
+ )
+ * Lc ( "}}" )
local ShortString = SingleShortString + DoubleShortString
-
-
+% \end{macrocode}
+%
+% \bigskip
+% Of course, it's more complicated for ``longs strings'' because, by definition,
+% in Python, those strings may be broken by an end on line (which is catched by
+% the \textsc{lpeg} |EOL|).
+% \begin{macrocode}
local SingleLongString =
- L "{\\PitonStyle{String.Long}{"
- * (
- C ( S "fF" * P "'''" )
- * ( SingleLongInterpol + SingleLongPureString ) ^ 0
- * L "}}"
- * (
- EOL
- +
- L "{\\PitonStyle{String.Long}{"
- * ( SingleLongInterpol + SingleLongPureString ) ^ 0
- * L "}}"
- * EOL
- ) ^ 0
- * L "{\\PitonStyle{String.Long}{"
+ Lc "{\\PitonStyle{String.Long}{"
+ * (
+ Q ( S "fF" * P "'''" )
+ * ( SingleLongInterpol + SingleLongPureString ) ^ 0
+ * Lc "}}"
+ * (
+ EOL
+ +
+ Lc "{\\PitonStyle{String.Long}{"
* ( SingleLongInterpol + SingleLongPureString ) ^ 0
- +
- C ( ( S "rR" ) ^ -1 * P "'''" * ( 1 - P "'''" - P "\r" ) ^ 0 )
- * L "}}"
- * (
- L "{\\PitonStyle{String.Long}{"
- * C ( ( 1 - P "'''" - P "\r" ) ^ 0 )
- * L "}}"
- * EOL
- ) ^ 0
- * L "{\\PitonStyle{String.Long}{"
- * C ( ( 1 - P "'''" - P "\r" ) ^ 0 )
- )
- * C ( P "'''" )
- * L "}}"
+ * Lc "}}"
+ * EOL
+ ) ^ 0
+ * Lc "{\\PitonStyle{String.Long}{"
+ * ( SingleLongInterpol + SingleLongPureString ) ^ 0
+ +
+ Q ( ( S "rR" ) ^ -1 * P "'''"
+ * ( 1 - P "'''" - P "\r" ) ^ 0 )
+ * Lc "}}"
+ * (
+ Lc "{\\PitonStyle{String.Long}{"
+ * Q ( ( 1 - P "'''" - P "\r" ) ^ 0 )
+ * Lc "}}"
+ * EOL
+ ) ^ 0
+ * Lc "{\\PitonStyle{String.Long}{"
+ * Q ( ( 1 - P "'''" - P "\r" ) ^ 0 )
+ )
+ * Q ( P "'''" )
+ * Lc "}}"
local DoubleLongString =
- L "{\\PitonStyle{String.Long}{"
- * (
- C ( S "fF" * P "\"\"\"" )
- * ( DoubleLongInterpol + DoubleLongPureString ) ^ 0
- * L "}}"
- * (
- EOL
- +
- L "{\\PitonStyle{String.Long}{"
- * ( DoubleLongInterpol + DoubleLongPureString ) ^ 0
- * L "}}"
- * EOL
- ) ^ 0
- * L "{\\PitonStyle{String.Long}{"
+ Lc "{\\PitonStyle{String.Long}{"
+ * (
+ Q ( S "fF" * P "\"\"\"" )
+ * ( DoubleLongInterpol + DoubleLongPureString ) ^ 0
+ * Lc "}}"
+ * (
+ EOL
+ +
+ Lc "{\\PitonStyle{String.Long}{"
* ( DoubleLongInterpol + DoubleLongPureString ) ^ 0
- +
- C ( ( S "rR" ) ^ -1 * P "\"\"\"" * ( 1 - P "\"\"\"" - P "\r" ) ^ 0 )
- * L "}}"
- * (
- L "{\\PitonStyle{String.Long}{"
- * C ( ( 1 - P "\"\"\"" - P "\r" ) ^ 0 )
- * L "}}"
- * EOL
- ) ^ 0
- * L "{\\PitonStyle{String.Long}{"
- * C ( ( 1 - P "\"\"\"" - P "\r" ) ^ 0 )
- )
- * C ( P "\"\"\"" )
- * L "}}"
-
-
-
+ * Lc "}}"
+ * EOL
+ ) ^ 0
+ * Lc "{\\PitonStyle{String.Long}{"
+ * ( DoubleLongInterpol + DoubleLongPureString ) ^ 0
+ +
+ Q ( ( S "rR" ) ^ -1 * P "\"\"\""
+ * ( 1 - P "\"\"\"" - P "\r" ) ^ 0 )
+ * Lc "}}"
+ * (
+ Lc "{\\PitonStyle{String.Long}{"
+ * Q ( ( 1 - P "\"\"\"" - P "\r" ) ^ 0 )
+ * Lc "}}"
+ * EOL
+ ) ^ 0
+ * Lc "{\\PitonStyle{String.Long}{"
+ * Q ( ( 1 - P "\"\"\"" - P "\r" ) ^ 0 )
+ )
+ * Q ( P "\"\"\"" )
+ * Lc "}}"
+% \end{macrocode}
+%
+% \begin{macrocode}
local LongString = SingleLongString + DoubleLongString
+% \end{macrocode}
+%
+% \bigskip
+% We have a \textsc{lpeg} for the Python docstrings. That \textsc{lpeg} will
+% be used in the \textsc{lpeg} |DefFunction| which deals with the whole preamble
+% of a function definition (which begins with |def|).
+% \begin{macrocode}
+local StringDoc =
+ K ( P "\"\"\"" , 'String.Doc' )
+ * ( K ( (1 - P "\"\"\"" - P "\r" ) ^ 0 , 'String.Doc' ) * EOL * Tab ^0 ) ^ 0
+ * K ( ( 1 - P "\"\"\"" - P "\r" ) ^ 0 * P "\"\"\"" , 'String.Doc' )
+% \end{macrocode}
+%
+% \bigskip
+% \paragraph{The comments in the Python listings}
+%
+% We define different \textsc{lpeg} dealing with comments in the Python
+% listings.
+% \begin{macrocode}
+local CommentMath =
+ P "$" * K ( ( 1 - S "$\r" ) ^ 1 , 'Comment.Math' ) * P "$"
+
+local Comment =
+ Lc ( "{\\PitonStyle{Comment}{" )
+ * Q ( P "#" )
+ * ( CommentMath + Q ( ( 1 - S "$\r" ) ^ 1 ) ) ^ 0
+ * Lc ( "}}" )
+ * ( EOL + -1 )
+% \end{macrocode}
+%
-local Expression =
- P { "E" ,
- E = ( 1 - S "{}()[]\r," ) ^ 0
- * (
- ( P "{" * V "F" * P "}"
- + P "(" * V "F" * P ")"
- + P "[" * V "F" * P "]" ) * ( 1 - S "{}()[]\r," ) ^ 0
- ) ^ 0 ,
- F = ( 1 - S "{}()[]\r\"'" ) ^ 0
- * ( (
- P "'" * (P "\\'" + 1 - S"'\r" )^0 * P "'"
- + P "\"" * (P "\\\"" + 1 - S"\"\r" )^0 * P "\""
- + P "{" * V "F" * P "}"
- + P "(" * V "F" * P ")"
- + P "[" * V "F" * P "]"
- ) * ( 1 - S "{}()[]\r\"'" ) ^ 0 ) ^ 0 ,
- }
-
-local Param = SkipSpace * K ( identifier , '' ) * SkipSpace
- * ( K ( P "=" * Expression , 'InitialValues' )
- + K ( P ":" , '' ) * SkipSpace * K ( letter^1 , 'Name.Type' ))
- + SkipSpace * K ( alphanum ^ 1 , '' ) * SkipSpace
-
-local Params = Param * ( K ( P "," , '' ) * Param ) ^ 0
-
-local StringDoc = K ( P "\"\"\"" , 'String.Doc' )
- * ( K ( (1 - P "\"\"\"" - P "\r" ) ^ 0 , 'String.Doc' ) * EOL ) ^ 0
- * K ( ( 1 - P "\"\"\"" - P "\r" ) ^ 0 * P "\"\"\"" , 'String.Doc' )
- + K ( P "'''" , 'String.Doc' )
- * ( K ( (1 - P "'''" - P "\r")^0 , 'String.Doc' ) * EOL ) ^ 0
- * K ( ( 1 - P "'''" - P "\r")^0 * P "'''" , 'String.Doc' )
-
-local CommentMath = P "$" * K ( ( 1 - S "$\r" ) ^ 1 , 'Comment.Math' ) * P "$"
-
-
-local Comment = L ( "{\\PitonStyle{Comment}{" )
- * C ( P "#" ) * ( CommentMath + C ( ( 1 - S "$\r" ) ^ 1 ) ) ^ 0
- * L ( "}}" )
- * ( EOL + -1 )
-
+% \bigskip
+% The following \textsc{lpeg} |CommentLaTeX| is for what is called in that
+% document the ``LaTeX comments''. Since the elements that will be catched must
+% be sent to LaTeX with standard LaTeX catcodes, we put the capture (done by
+% the function~|C|) in a table (by using~|Ct|, which is an alias for |lpeg.Ct|).
+% \begin{macrocode}
local CommentLaTeX =
- P "##"
- * L "{\\PitonStyle{Comment.LaTeX}{\\ignorespaces"
- * Ct ( C ( ( 1 - P "\r" ) ^ 0 ) )
- * L "}}"
- * ( EOL + -1 )
-
+ P "##"
+ * Lc "{\\PitonStyle{Comment.LaTeX}{\\ignorespaces"
+ * L ( ( 1 - P "\r" ) ^ 0 )
+ * Lc "}}"
+ * ( EOL + -1 )
+% \end{macrocode}
+%
+% \bigskip
+% \paragraph{DefFunction}
+%
+% The following \textsc{lpeg} |Expression| will be used for the parameters in
+% the \emph{argspec} of a Python function. It's necessary to use a \emph{grammar}
+% because that pattern mainly checks the correct nesting of the delimiters
+% (and it's known in the theory of formal languages that this can't be done with
+% regular expressions \emph{stricto sensu} only).
+% \begin{macrocode}
+local Expression =
+ P { "E" ,
+ E = ( 1 - S "{}()[]\r," ) ^ 0
+ * (
+ ( P "{" * V "F" * P "}"
+ + P "(" * V "F" * P ")"
+ + P "[" * V "F" * P "]" ) * ( 1 - S "{}()[]\r," ) ^ 0
+ ) ^ 0 ,
+ F = ( 1 - S "{}()[]\r\"'" ) ^ 0
+ * ( (
+ P "'" * (P "\\'" + 1 - S"'\r" )^0 * P "'"
+ + P "\"" * (P "\\\"" + 1 - S"\"\r" )^0 * P "\""
+ + P "{" * V "F" * P "}"
+ + P "(" * V "F" * P ")"
+ + P "[" * V "F" * P "]"
+ ) * ( 1 - S "{}()[]\r\"'" ) ^ 0 ) ^ 0 ,
+ }
+% \end{macrocode}
+%
+% \bigskip
+% We will now define a \textsc{lpeg} |Params| that will catch the list of
+% parameters (that is to say the \emph{argspec}) in the definition of a Python
+% function. For example, in the line of code
+% \begin{center}
+% \piton|def MyFunction(a,b,x=10,n:int): return n|
+% \end{center}
+% the \textsc{lpeg} |Params| will be used to catch the chunk\enskip |a,b,x=10,n:int|.
+%
+% Or course, a |Params| is simply a comma-separated list of |Param|, and that's
+% why we define first the \textsc{lpeg} |Param|.
+%
+% \medskip
+% \begin{macrocode}
+local Param =
+ SkipSpace * Identifier * SkipSpace
+ * (
+ K ( P "=" * Expression , 'InitialValues' )
+ + Q ( P ":" ) * SkipSpace * K ( letter^1 , 'Name.Type' )
+ ) ^ -1
+% \end{macrocode}
+%
+% \medskip
+% \begin{macrocode}
+local Params = ( Param * ( Q "," * Param ) ^ 0 ) ^ -1
+% \end{macrocode}
+%
+% \bigskip
+% The following \textsc{lpeg} |DefFunction| catches a keyword |def| and the
+% following name of function \emph{but also everything else until a potential
+% docstring}. That's why this definition of \textsc{lpeg} must occur (in the file
+% |piton.sty|) after the definition of several other \textsc{lpeg} such as
+% |Comment|, |CommentLaTeX|, |Params|, |StringDoc|...
+% \begin{macrocode}
local DefFunction =
- K ( P "def" , 'Keyword' )
- * ( Space
- * K ( identifier , 'Name.Function' )
- * ( SkipSpace * K ( P "(" , '' ) * Params * K ( P ")" , '' ) ) ^ -1
- * ( SkipSpace
- * K ( ( 1 - S ":\r" )^0 , 'Post.Function' )
- * K ( P ":" , 'Keyword' )
- * SkipSpace
- * ( EOL + CommentLaTeX + Comment )
- * SkipSpace
- * StringDoc ) ^ -1
- ) ^ -1
-
-local ItemDict = ShortString * SkipSpace * C ( P ":" ) * K ( Expression , 'Dict.Value' )
+ K ( P "def" , 'Keyword' )
+ * Space
+ * K ( identifier , 'Name.Function' )
+ * SkipSpace
+ * Q ( P "(" ) * Params * Q ( P ")" )
+ * SkipSpace
+% \end{macrocode}
+% Here, we need a \pkg{piton} style |Post.Function| which will be linked to
+% |\@@_piton:n| (that means that the capture will be parsed once again by
+% \pkg{piton}). We could avoid that kind of trick by using a non-terminal of a
+% grammar but we have probably here a better legibility.
+% \begin{macrocode}
+ * K ( ( 1 - S ":\r" )^0 , 'Post.Function' )
+ * Q ( P ":" )
+ * ( SkipSpace
+ * ( EOL + CommentLaTeX + Comment ) -- in all cases, that contains an EOL
+ * Tab ^ 0
+ * SkipSpace
+ * StringDoc ^ 0 -- there may be additionnal docstrings
+ ) ^ -1
+% \end{macrocode}
+% Remark that, in the previous code, |CommentLaTeX| \emph{must} appear
+% before |Comment|: there is no commutativity of the addition for the
+% \emph{parsing expression grammars} (\textsc{peg}).
+%
+% \smallskip
+% If the word |def| is not followed by a identifier and parenthesis, it will be
+% catched as keyword by the \textsc{lpeg} |Keyword| (useful if, for example, the
+% final user wants to speak of the keyword \piton{def}).
+%
+% \bigskip
+% \paragraph{The dictionaries of Python}
+%
+% We have \textsc{lpeg} dealings with dictionaries of Python because, in
+% typesettings of explicit Python dictionnaries, one may prefer to have all the
+% values formatted in black (in order to see more clearly the keys which are
+% usually Python strings). That's why we have a \pkg{piton} style |Dict.Value|.
+%
+% The initial value of that \pkg{piton} style is |\@@_piton:n|, which means that
+% the value of the entry of the dictionary is parsed once again by \pkg{piton}
+% (and nothing special is done for the dictionary). In the following example, we
+% have set the \pkg{piton} style |Dict.Value| to |\color{black}|:
+%
+% \medskip
+% \begingroup
+% \SetPitonStyle{Dict.Value = \color{black}}
+% \piton|mydict = { 'name' : 'Paul', 'sex' : 'male', 'age' : 31 } |
+% \endgroup
+%
+% \medskip
+% At this time, this mechanism works only for explicit dictionaries on a single line!
+%
+% \begin{macrocode}
+local ItemDict =
+ ShortString * SkipSpace * Q ( P ":" ) * K ( Expression , 'Dict.Value' )
local ItemOfSet = SkipSpace * ( ItemDict + ShortString ) * SkipSpace
-local Set = C ( P "{" )
- * ItemOfSet * ( C ( P "," ) * ItemOfSet ) ^ 0
- * C ( P "}" )
-
-local Operator = K ( P "!=" + P "==" + P "<<" + P ">>" + S "-~+/*%=<>&.@|" , 'Operator')
+local Set =
+ Q ( P "{" )
+ * ItemOfSet * ( Q ( P "," ) * ItemOfSet ) ^ 0
+ * Q ( P "}" )
-local OperatorWord = K ( P "in" + P "is" + P "and" + P "or" + P "not" , 'Operator.Word')
-
-local SyntaxPython =
+% \end{macrocode}
+%
+%
+% \bigskip
+% \paragraph{The main LPEG}
+%
+% |SyntaxPython| is the main \textsc{lpeg} of the package \pkg{piton}. We have
+% written an auxiliary \textsc{lpeg} |SyntaxPythonAux| only for legibility.
+% \begin{macrocode}
+local SyntaxPythonAux =
+% \end{macrocode}
+% We recall that each line in the Python code to parse will be sent back to
+% LaTeX between a pair |\@@_begin_line:| -- |\@@_end_line:|\footnote{Remember
+% that the \texttt{\textbackslash @@\_end\_line:} must be explicit because it
+% will be used as marker in order to delimit the argument of the command
+% \texttt{\textbackslash @@\_begin\_line:}}.
+% \begin{macrocode}
+ Lc ( '\\@@_begin_line:' ) *
( ( space - P "\r" ) ^0 * P "\r" ) ^ -1 *
( ( space^1 * -1 )
+ EOL
+ + Tab
+ Space
+ Escape
+ CommentLaTeX
@@ -1976,117 +2566,183 @@ local SyntaxPython =
+ ExceptionInConsole
+ Set
+ Delim
- + Class * ( Space + Punct + EOL )
- + Namespace * ( Space + Punct + EOL )
+% \end{macrocode}
+% |Operator| must be before |Punct|.
+% \begin{macrocode}
+ + Operator
+ + ShortString
+ + Punct
+ + FromImport
+ ImportAs
+ RaiseException
- + Keyword * ( Space + Punct + EOL )
+ DefFunction
- + ShortString
- + Decorator * ( Space + Punct + EOL )
- + Operator
- + OperatorWord * ( Space + Punct + EOL )
- + Builtin * ( Space + Punct + EOL )
+ + DefClass
+ + Keyword * ( Space + Punct + Delim + EOL + -1)
+ + Decorator
+ + OperatorWord
+ + Builtin * ( Space + Punct + Delim + EOL + -1)
+ Identifier
+ Number
+ Word
- ) ^0 * -1
-
-
-local MinimalSyntax =
- P { "S" ;
- S = K ( (1 - P "\r" ) ^ 0 , '') + EOL * S
- }
-
-
+ ) ^0 * -1 * Lc ( '\\@@_end_line:' )
+% \end{macrocode}
+%
+% We have written a auxiliary \textsc{lpeg} |SyntaxPythonAux| for legibility only.
+% \begin{macrocode}
+local SyntaxPython = Ct ( SyntaxPythonAux )
+% \end{macrocode}
+%
+% \bigskip
+% \subsubsection{The function Parse}
+%
+% The function |Parse| is the main function of the package \pkg{piton}.
+% It parses its argument and sends back to LaTeX the code with interlaced
+% formatting LaTeX instructions. In fact, everything is done by the
+% \textsc{lpeg} |SyntaxPython| which returns as capture a Lua table containing
+% data to send to LaTeX.
+%
+% \bigskip
+% \begin{macrocode}
function Parse(code)
- local t = Ct(SyntaxPython) : match(code)
- tex.sprint( luatexbase.catcodetables.expl , '\\__piton_begin_line:' )
- if t then else t = Ct(MinimalSyntax) : match(code) end
- for i = 1 , #t do
- if type(t[i]) == 'string'
- then
- tex.sprint(luatexbase.catcodetables.CatcodeTableOther, t[i])
- else
- tex.tprint(t[i])
- end
- end
- tex.sprint( luatexbase.catcodetables.expl , '\\__piton_end_line:' )
+ local t = SyntaxPython : match ( code )
+ for _ , s in ipairs(t) do tex.tprint(s) end
end
-
+% \end{macrocode}
+%
+% \bigskip
+% The function |ParseFile| will be used by the LaTeX command |\PitonInputFile|.
+% That function merely reads the whole file (that is to say all its lines) and
+% then apply the function~|Parse| to the resulting Lua string.
+% \begin{macrocode}
function ParseFile(name)
s = ''
for line in io.lines(name) do s = s .. '\r' .. line end
Parse(s)
end
-
-
-function define_gobble_syntax(n)
- GobbleSyntax = ( 1 - P "\r" ) ^ (-n) * C ( ( 1 - P "\r" ) ^0 )
- * ( C ( P "\r" )
- * ( 1 - P "\r" ) ^ (-n)
- * C ( ( 1 - P "\r" ) ^0 )
- ) ^ 0
-end
-
-function GobbleParse(code)
- local t = Ct(GobbleSyntax):match(code)
- local new_code = ""
- for i = 1 , #t do
- new_code = new_code .. t[i]
+% \end{macrocode}
+%
+% \bigskip
+% \subsubsection{The preprocessors of the function Parse}
+%
+% We deal now with preprocessors of the function |Parse| which are needed when
+% the ``gobble mechanism'' is used.
+%
+%
+% \bigskip
+% The function |gobble| gobbles $n$ characters on the left of the code. It uses
+% a \textsc{lpeg} that we have to compute dynamically because if depends on the
+% value of~$n$.
+% \begin{macrocode}
+function gobble(n,code)
+ function concat(acc,new_value)
+ return acc .. new_value
+ end
+ if n==0
+ then return code
+ else
+ return Cf (
+ Cc ( "" ) *
+ ( 1 - P "\r" ) ^ (-n) * C ( ( 1 - P "\r" ) ^ 0 )
+ * ( C ( P "\r" )
+ * ( 1 - P "\r" ) ^ (-n)
+ * C ( ( 1 - P "\r" ) ^ 0 )
+ ) ^ 0 ,
+ concat
+ ) : match ( code )
end
- Parse(new_code)
end
-
+% \end{macrocode}
+%
+% \bigskip
+% \begin{macrocode}
+function GobbleParse(n,code)
+ if n==-1
+ then n = AutoGobbleLPEG : match(code)
+ else if n==-2
+ then n = EnvGobbleLPEG : match(code)
+ end
+ end
+ Parse(gobble(n,code))
+end
+% \end{macrocode}
+%
+%
+% \bigskip
+% The following function |add| will be used in the following \textsc{lpeg}
+% |AutoGobbleLPEG| and |EnvGobbleLPEG|.
+% \begin{macrocode}
function add(acc,new_value)
return acc + new_value
end
-
-
-
-
---[[ The following LPEG returns as capture the minimal number of spaces at
-the beginning of the lines of code]]
-AutoGobbleSyntax =
+% \end{macrocode}
+%
+% \bigskip
+% The following \textsc{lpeg} returns as capture the minimal number of spaces at
+% the beginning of the lines of code. The main work is done by two \emph{fold
+% captures} (|lpeg.Cf|), one using |add| and the other (encompassing the
+% previous one) using |math.min| as folding operator.
+% \begin{macrocode}
+AutoGobbleLPEG =
( space ^ 0 * P "\r" ) ^ -1
* Cf (
(
+% \end{macrocode}
+% We don't take into account the empty lines (with only spaces).
+% \begin{macrocode}
( P " " ) ^ 0 * P "\r"
+
Cf ( Cc(0) * ( P " " * Cc(1) ) ^ 0 , add )
* ( 1 - P " " ) * ( 1 - P "\r" ) ^ 0 * P "\r"
) ^ 0
+% \end{macrocode}
+% Now for the last line of the Python code...
+% \begin{macrocode}
*
( Cf ( Cc(0) * ( P " " * Cc(1) ) ^ 0 , add )
* ( 1 - P " " ) * ( 1 - P "\r" ) ^ 0 ) ^ -1 ,
math.min
)
-
-function AutoGobbleParse(code)
- local n = AutoGobbleSyntax:match(code)
- if n==0
- then Parse(code)
- else define_gobble_syntax(n)
- GobbleParse(code)
- end
-end
-
-
---[[ The following LPEG returns as capture the number of spaces at the last line,
-that is to say begin the \end{Piton} ]]
-
-EnvGobbleSyntax =
- ( ( 1 - P "\r" ) ^ 0 * P "\r" ) ^ 0
- * Cf ( Cc(0) * ( P " " * Cc(1) ) ^ 0 , add ) * -1
-
-function EnvGobbleParse(code)
- local n = EnvGobbleSyntax:match(code)
- if n==0
- then Parse(code)
- else define_gobble_syntax(n)
- GobbleParse(code)
- end
-end
+% \end{macrocode}
+%
+% \bigskip
+% The following \textsc{lpeg} returns as capture the number of spaces at the
+% last line, that is to say before the |\end{Piton}| (and usually it's also the
+% number of spaces before the corresponding |\begin{Piton}| because that's the
+% traditionnal way to indent in LaTeX). The main work is done by a \emph{fold
+% capture} (|lpeg.Cf|) using the function |add| as folding operator.
+% \begin{macrocode}
+EnvGobbleLPEG =
+ ( ( 1 - P "\r" ) ^ 0 * P "\r" ) ^ 0
+ * Cf ( Cc(0) * ( P " " * Cc(1) ) ^ 0 , add ) * -1
+% \end{macrocode}
+%
+% \bigskip
+% \subsubsection{To count the number of lines}
+%
+% \medskip
+% \begin{macrocode}
+function CountLines(code)
+ local count = 0
+ for i in code:gmatch("\r") do count = count + 1 end
+ tex.sprint(
+ luatexbase.catcodetables.expl ,
+ '\\int_set:Nn \\l_@@_nb_lines_int {' .. count .. '}' )
+end
+% \end{macrocode}
+%
+% \bigskip
+% \begin{macrocode}
+function CountLinesFile(name)
+ local count = 0
+ for line in io.lines(name) do count = count + 1 end
+ tex.sprint(
+ luatexbase.catcodetables.expl ,
+ '\\int_set:Nn \\l_@@_nb_lines_int {' .. count .. '}' )
+end
+% \end{macrocode}
+%
+% \bigskip
+% \begin{macrocode}
\end{luacode*}
% \end{macrocode}
%
@@ -2107,6 +2763,12 @@ end
%
% New key |left-margin|.
%
+% \subsection*{Changes between versions 0.8 and 0.9}
+%
+% New key |tab-size|.
+%
+% Integer value for the key |splittable|.
+%
% \end{document}
%
%