source: doc/papers/general/Paper.tex @ 41fcd94

aaron-thesisarm-ehcleanup-dtorsdeferred_resndemanglerjacob/cs343-translationjenkins-sandboxnew-astnew-ast-unique-exprnew-envno_listpersistent-indexerresolv-newwith_gc
Last change on this file since 41fcd94 was 41fcd94, checked in by Peter A. Buhr <pabuhr@…>, 4 years ago

Merge branch 'master' of plg2:software/cfa/cfa-cc

  • Property mode set to 100644
File size: 115.5 KB
Line 
1\documentclass{article}
2
3\usepackage{fullpage}
4\usepackage{epic,eepic}
5\usepackage{xspace,calc,comment}
6\usepackage{upquote}                                                                    % switch curled `'" to straight
7\usepackage{listings}                                                                   % format program code
8\usepackage{rotating}
9\usepackage[usenames]{color}
10\usepackage{pslatex}                                    % reduce size of san serif font
11\usepackage[plainpages=false,pdfpagelabels,pdfpagemode=UseNone,pagebackref=true,breaklinks=true,colorlinks=true,linkcolor=blue,citecolor=blue,urlcolor=blue]{hyperref}
12
13\setlength{\textheight}{9in}
14%\oddsidemargin 0.0in
15\renewcommand{\topfraction}{0.8}                % float must be greater than X of the page before it is forced onto its own page
16\renewcommand{\bottomfraction}{0.8}             % float must be greater than X of the page before it is forced onto its own page
17\renewcommand{\floatpagefraction}{0.8}  % float must be greater than X of the page before it is forced onto its own page
18\renewcommand{\textfraction}{0.0}               % the entire page maybe devoted to floats with no text on the page at all
19
20\lefthyphenmin=4                                                % hyphen only after 4 characters
21\righthyphenmin=4
22
23% Names used in the document.
24
25\newcommand{\CFAIcon}{\textsf{C}\raisebox{\depth}{\rotatebox{180}{\textsf{A}}}\xspace} % Cforall symbolic name
26\newcommand{\CFA}{\protect\CFAIcon} % safe for section/caption
27\newcommand{\CFL}{\textrm{Cforall}\xspace} % Cforall symbolic name
28\newcommand{\Celeven}{\textrm{C11}\xspace} % C11 symbolic name
29\newcommand{\CC}{\textrm{C}\kern-.1em\hbox{+\kern-.25em+}\xspace} % C++ symbolic name
30\newcommand{\CCeleven}{\textrm{C}\kern-.1em\hbox{+\kern-.25em+}11\xspace} % C++11 symbolic name
31\newcommand{\CCfourteen}{\textrm{C}\kern-.1em\hbox{+\kern-.25em+}14\xspace} % C++14 symbolic name
32\newcommand{\CCseventeen}{\textrm{C}\kern-.1em\hbox{+\kern-.25em+}17\xspace} % C++17 symbolic name
33\newcommand{\CCtwenty}{\textrm{C}\kern-.1em\hbox{+\kern-.25em+}20\xspace} % C++20 symbolic name
34\newcommand{\CCV}{\rm C\kern-.1em\hbox{+\kern-.25em+}obj\xspace} % C++ virtual symbolic name
35\newcommand{\Csharp}{C\raisebox{-0.7ex}{\Large$^\sharp$}\xspace} % C# symbolic name
36
37%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
38
39\newcommand{\Textbf}[2][red]{{\color{#1}{\textbf{#2}}}}
40\newcommand{\TODO}[1]{\textbf{TODO}: {\itshape #1}} % TODO included
41%\newcommand{\TODO}[1]{} % TODO elided
42
43% Default underscore is too low and wide. Cannot use lstlisting "literate" as replacing underscore
44% removes it as a variable-name character so keywords in variables are highlighted. MUST APPEAR
45% AFTER HYPERREF.
46%\DeclareTextCommandDefault{\textunderscore}{\leavevmode\makebox[1.2ex][c]{\rule{1ex}{0.1ex}}}
47\renewcommand{\textunderscore}{\leavevmode\makebox[1.2ex][c]{\rule{1ex}{0.075ex}}}
48
49\makeatletter
50% parindent is relative, i.e., toggled on/off in environments like itemize, so store the value for
51% use rather than use \parident directly.
52\newlength{\parindentlnth}
53\setlength{\parindentlnth}{\parindent}
54
55\newcommand{\LstKeywordStyle}[1]{{\lst@basicstyle{\lst@keywordstyle{#1}}}}
56\newcommand{\LstCommentStyle}[1]{{\lst@basicstyle{\lst@commentstyle{#1}}}}
57
58\newlength{\gcolumnposn}                                % temporary hack because lstlisting does not handle tabs correctly
59\newlength{\columnposn}
60\setlength{\gcolumnposn}{2.75in}
61\setlength{\columnposn}{\gcolumnposn}
62\newcommand{\C}[2][\@empty]{\ifx#1\@empty\else\global\setlength{\columnposn}{#1}\global\columnposn=\columnposn\fi\hfill\makebox[\textwidth-\columnposn][l]{\lst@basicstyle{\LstCommentStyle{#2}}}}
63\newcommand{\CRT}{\global\columnposn=\gcolumnposn}
64
65% Denote newterms in particular font and index them without particular font and in lowercase, e.g., \newterm{abc}.
66% The option parameter provides an index term different from the new term, e.g., \newterm[\texttt{abc}]{abc}
67% The star version does not lowercase the index information, e.g., \newterm*{IBM}.
68\newcommand{\newtermFontInline}{\emph}
69\newcommand{\newterm}{\@ifstar\@snewterm\@newterm}
70\newcommand{\@newterm}[2][\@empty]{\lowercase{\def\temp{#2}}{\newtermFontInline{#2}}\ifx#1\@empty\index{\temp}\else\index{#1@{\protect#2}}\fi}
71\newcommand{\@snewterm}[2][\@empty]{{\newtermFontInline{#2}}\ifx#1\@empty\index{#2}\else\index{#1@{\protect#2}}\fi}
72
73% Latin abbreviation
74\newcommand{\abbrevFont}{\textit}       % set empty for no italics
75\newcommand{\EG}{\abbrevFont{e}.\abbrevFont{g}.}
76\newcommand*{\eg}{%
77        \@ifnextchar{,}{\EG}%
78                {\@ifnextchar{:}{\EG}%
79                        {\EG,\xspace}}%
80}%
81\newcommand{\IE}{\abbrevFont{i}.\abbrevFont{e}.}
82\newcommand*{\ie}{%
83        \@ifnextchar{,}{\IE}%
84                {\@ifnextchar{:}{\IE}%
85                        {\IE,\xspace}}%
86}%
87\newcommand{\ETC}{\abbrevFont{etc}}
88\newcommand*{\etc}{%
89        \@ifnextchar{.}{\ETC}%
90        {\ETC\xspace}%
91}%
92\newcommand{\ETAL}{\abbrevFont{et}\hspace{2pt}\abbrevFont{al}}
93\newcommand*{\etal}{%
94        \@ifnextchar{.}{\protect\ETAL}%
95                {\abbrevFont{\protect\ETAL}.\xspace}%
96}%
97\newcommand{\VIZ}{\abbrevFont{viz}}
98\newcommand*{\viz}{%
99        \@ifnextchar{.}{\VIZ}%
100                {\abbrevFont{\VIZ}.\xspace}%
101}%
102\makeatother
103
104% CFA programming language, based on ANSI C (with some gcc additions)
105\lstdefinelanguage{CFA}[ANSI]{C}{
106        morekeywords={
107                _Alignas, _Alignof, __alignof, __alignof__, asm, __asm, __asm__, _At, __attribute,
108                __attribute__, auto, _Bool, catch, catchResume, choose, _Complex, __complex, __complex__,
109                __const, __const__, disable, dtype, enable, __extension__, fallthrough, fallthru,
110                finally, forall, ftype, _Generic, _Imaginary, inline, __label__, lvalue, _Noreturn, one_t,
111                otype, restrict, _Static_assert, throw, throwResume, trait, try, ttype, typeof, __typeof,
112                __typeof__, virtual, with, zero_t},
113        moredirectives={defined,include_next}%
114}%
115
116\lstset{
117language=CFA,
118columns=fullflexible,
119basicstyle=\linespread{0.9}\sf,                                                 % reduce line spacing and use sanserif font
120stringstyle=\tt,                                                                                % use typewriter font
121tabsize=5,                                                                                              % N space tabbing
122xleftmargin=\parindentlnth,                                                             % indent code to paragraph indentation
123%mathescape=true,                                                                               % LaTeX math escape in CFA code $...$
124escapechar=\$,                                                                                  % LaTeX escape in CFA code
125keepspaces=true,                                                                                %
126showstringspaces=false,                                                                 % do not show spaces with cup
127showlines=true,                                                                                 % show blank lines at end of code
128aboveskip=4pt,                                                                                  % spacing above/below code block
129belowskip=3pt,
130% replace/adjust listing characters that look bad in sanserif
131literate={-}{\makebox[1ex][c]{\raisebox{0.4ex}{\rule{0.8ex}{0.1ex}}}}1 {^}{\raisebox{0.6ex}{$\scriptscriptstyle\land\,$}}1
132        {~}{\raisebox{0.3ex}{$\scriptstyle\sim\,$}}1 % {`}{\ttfamily\upshape\hspace*{-0.1ex}`}1
133        {<-}{$\leftarrow$}2 {=>}{$\Rightarrow$}2 {->}{\makebox[1ex][c]{\raisebox{0.4ex}{\rule{0.8ex}{0.075ex}}}\kern-0.2ex\textgreater}2,
134moredelim=**[is][\color{red}]{`}{`},
135}% lstset
136
137% inline code @...@
138\lstMakeShortInline@%
139
140\lstnewenvironment{cfa}[1][]
141{\lstset{#1}}
142{}
143\lstnewenvironment{C++}[1][]                            % use C++ style
144{\lstset{language=C++,moredelim=**[is][\protect\color{red}]{`}{`},#1}\lstset{#1}}
145{}
146
147
148\title{Generic and Tuple Types with Efficient Dynamic Layout in \protect\CFA}
149
150\author{Aaron Moss, Robert Schluntz, Peter Buhr}
151% \email{a3moss@uwaterloo.ca}
152% \email{rschlunt@uwaterloo.ca}
153% \email{pabuhr@uwaterloo.ca}
154% \affiliation{%
155%       \institution{University of Waterloo}
156%       \department{David R. Cheriton School of Computer Science}
157%       \streetaddress{Davis Centre, University of Waterloo}
158%       \city{Waterloo}
159%       \state{ON}
160%       \postcode{N2L 3G1}
161%       \country{Canada}
162% }
163
164%\terms{generic, tuple, variadic, types}
165%\keywords{generic types, tuple types, variadic types, polymorphic functions, C, Cforall}
166
167\begin{document}
168\maketitle
169
170
171\begin{abstract}
172The C programming language is a foundational technology for modern computing with millions of lines of code implementing everything from commercial operating-systems to hobby projects.
173This installation base and the programmers producing it represent a massive software-engineering investment spanning decades and likely to continue for decades more.
174Nonetheless, C, first standardized over thirty years ago, lacks many features that make programming in more modern languages safer and more productive.
175The goal of the \CFA project is to create an extension of C that provides modern safety and productivity features while still ensuring strong backwards compatibility with C and its programmers.
176Prior projects have attempted similar goals but failed to honour C programming-style; for instance, adding object-oriented or functional programming with garbage collection is a non-starter for many C developers.
177Specifically, \CFA is designed to have an orthogonal feature-set based closely on the C programming paradigm, so that \CFA features can be added \emph{incrementally} to existing C code-bases, and C programmers can learn \CFA extensions on an as-needed basis, preserving investment in existing code and engineers.
178This paper describes two \CFA extensions, generic and tuple types, details how their design avoids shortcomings of similar features in C and other C-like languages, and presents experimental results validating the design.
179\end{abstract}
180
181
182\section{Introduction and Background}
183
184The C programming language is a foundational technology for modern computing with millions of lines of code implementing everything from commercial operating-systems to hobby projects.
185This installation base and the programmers producing it represent a massive software-engineering investment spanning decades and likely to continue for decades more.
186The TIOBE~\cite{TIOBE} ranks the top 5 most popular programming languages as: Java 16\%, \Textbf{C 7\%}, \Textbf{\CC 5\%}, \Csharp 4\%, Python 4\% = 36\%, where the next 50 languages are less than 3\% each with a long tail.
187The top 3 rankings over the past 30 years are:
188\lstDeleteShortInline@%
189\begin{center}
190\setlength{\tabcolsep}{10pt}
191\begin{tabular}{@{}rccccccc@{}}
192                & 2017  & 2012  & 2007  & 2002  & 1997  & 1992  & 1987          \\ \hline
193Java    & 1             & 1             & 1             & 1             & 12    & -             & -                     \\
194\Textbf{C}      & \Textbf{2}& \Textbf{2}& \Textbf{2}& \Textbf{2}& \Textbf{1}& \Textbf{1}& \Textbf{1}    \\
195\CC             & 3             & 3             & 3             & 3             & 2             & 2             & 4                     \\
196\end{tabular}
197\end{center}
198\lstMakeShortInline@%
199Love it or hate it, C is extremely popular, highly used, and one of the few systems languages.
200In many cases, \CC is often used solely as a better C.
201Nonetheless, C, first standardized over thirty years ago, lacks many features that make programming in more modern languages safer and more productive.
202
203\CFA (pronounced ``C-for-all'', and written \CFA or Cforall) is an evolutionary extension of the C programming language that aims to add modern language features to C while maintaining both source compatibility with C and a familiar programming model for programmers.
204The four key design goals for \CFA~\cite{Bilson03} are:
205(1) The behaviour of standard C code must remain the same when translated by a \CFA compiler as when translated by a C compiler;
206(2) Standard C code must be as fast and as small when translated by a \CFA compiler as when translated by a C compiler;
207(3) \CFA code must be at least as portable as standard C code;
208(4) Extensions introduced by \CFA must be translated in the most efficient way possible.
209These goals ensure existing C code-bases can be converted to \CFA incrementally with minimal effort, and C programmers can productively generate \CFA code without training beyond the features being used.
210\CC is used similarly, but has the disadvantages of multiple legacy design-choices that cannot be updated and active divergence of the language model from C, requiring significant effort and training to incrementally add \CC to a C-based project.
211
212\CFA is currently implemented as a source-to-source translator from \CFA to the GCC-dialect of C~\cite{GCCExtensions}, allowing it to leverage the portability and code optimizations provided by GCC, meeting goals (1)--(3).
213Ultimately, a compiler is necessary for advanced features and optimal performance.
214
215This paper identifies shortcomings in existing approaches to generic and variadic data types in C-like languages and presents a design for generic and variadic types avoiding those shortcomings.
216Specifically, the solution is both reusable and type-checked, as well as conforming to the design goals of \CFA with ergonomic use of existing C abstractions.
217The new constructs are empirically compared with both standard C and \CC; the results show the new design is comparable in performance.
218
219
220\subsection{Polymorphic Functions}
221\label{sec:poly-fns}
222
223\CFA{}\hspace{1pt}'s polymorphism was originally formalized by Ditchfield~\cite{Ditchfield92}, and first implemented by Bilson~\cite{Bilson03}.
224The signature feature of \CFA is parametric-polymorphic functions~\cite{forceone:impl,Cormack90,Duggan96} with functions generalized using a @forall@ clause (giving the language its name):
225\begin{lstlisting}
226`forall( otype T )` T identity( T val ) { return val; }
227int forty_two = identity( 42 );                         $\C{// T is bound to int, forty\_two == 42}$
228\end{lstlisting}
229The @identity@ function above can be applied to any complete \emph{object type} (or @otype@).
230The type variable @T@ is transformed into a set of additional implicit parameters encoding sufficient information about @T@ to create and return a variable of that type.
231The \CFA implementation passes the size and alignment of the type represented by an @otype@ parameter, as well as an assignment operator, constructor, copy constructor and destructor.
232If this extra information is not needed, \eg for a pointer, the type parameter can be declared as a \emph{data type} (or @dtype@).
233
234In \CFA, the polymorphism runtime-cost is spread over each polymorphic call, due to passing more arguments to polymorphic functions;
235the experiments in Section~\ref{sec:eval} show this overhead is similar to \CC virtual-function calls.
236A design advantage is that, unlike \CC template-functions, \CFA polymorphic-functions are compatible with C \emph{separate compilation}, preventing compilation and code bloat.
237
238Since bare polymorphic-types provide a restricted set of available operations, \CFA provides a \emph{type assertion}~\cite[pp.~37-44]{Alphard} mechanism to provide further type information, where type assertions may be variable or function declarations that depend on a polymorphic type-variable.
239For example, the function @twice@ can be defined using the \CFA syntax for operator overloading:
240\begin{lstlisting}
241forall( otype T `| { T ?+?(T, T); }` ) T twice( T x ) { return x + x; } $\C{// ? denotes operands}$
242int val = twice( twice( 3.7 ) );
243\end{lstlisting}
244which works for any type @T@ with a matching addition operator.
245The polymorphism is achieved by creating a wrapper function for calling @+@ with @T@ bound to @double@, then passing this function to the first call of @twice@.
246There is now the option of using the same @twice@ and converting the result to @int@ on assignment, or creating another @twice@ with type parameter @T@ bound to @int@ because \CFA uses the return type~\cite{Cormack81,Baker82,Ada}, in its type analysis.
247The first approach has a late conversion from @double@ to @int@ on the final assignment, while the second has an eager conversion to @int@.
248\CFA minimizes the number of conversions and their potential to lose information, so it selects the first approach, which corresponds with C-programmer intuition.
249
250Crucial to the design of a new programming language are the libraries to access thousands of external software features.
251Like \CC, \CFA inherits a massive compatible library-base, where other programming languages must rewrite or provide fragile inter-language communication with C.
252A simple example is leveraging the existing type-unsafe (@void *@) C @bsearch@ to binary search a sorted floating-point array:
253\begin{lstlisting}
254void * bsearch( const void * key, const void * base, size_t nmemb, size_t size,
255                                int (* compar)( const void *, const void * ));
256int comp( const void * t1, const void * t2 ) { return *(double *)t1 < *(double *)t2 ? -1 :
257                                *(double *)t2 < *(double *)t1 ? 1 : 0; }
258double key = 5.0, vals[10] = { /* 10 sorted floating-point values */ };
259double * val = (double *)bsearch( &key, vals, 10, sizeof(vals[0]), comp );      $\C{// search sorted array}$
260\end{lstlisting}
261which can be augmented simply with a generalized, type-safe, \CFA-overloaded wrappers:
262\begin{lstlisting}
263forall( otype T | { int ?<?( T, T ); } ) T * bsearch( T key, const T * arr, size_t size ) {
264        int comp( const void * t1, const void * t2 ) { /* as above with double changed to T */ }
265        return (T *)bsearch( &key, arr, size, sizeof(T), comp ); }
266forall( otype T | { int ?<?( T, T ); } ) unsigned int bsearch( T key, const T * arr, size_t size ) {
267        T * result = bsearch( key, arr, size ); $\C{// call first version}$
268        return result ? result - arr : size; }  $\C{// pointer subtraction includes sizeof(T)}$
269double * val = bsearch( 5.0, vals, 10 );        $\C{// selection based on return type}$
270int posn = bsearch( 5.0, vals, 10 );
271\end{lstlisting}
272The nested function @comp@ provides the hidden interface from typed \CFA to untyped (@void *@) C, plus the cast of the result.
273Providing a hidden @comp@ function in \CC is awkward as lambdas do not use C calling-conventions and template declarations cannot appear at block scope.
274As well, an alternate kind of return is made available: position versus pointer to found element.
275\CC's type-system cannot disambiguate between the two versions of @bsearch@ because it does not use the return type in overload resolution, nor can \CC separately compile a templated @bsearch@.
276
277\CFA has replacement libraries condensing hundreds of existing C functions into tens of \CFA overloaded functions, all without rewriting the actual computations.
278For example, it is possible to write a type-safe \CFA wrapper @malloc@ based on the C @malloc@:
279\begin{lstlisting}
280forall( dtype T | sized(T) ) T * malloc( void ) { return (T *)malloc( sizeof(T) ); }
281int * ip = malloc();                                            $\C{// select type and size from left-hand side}$
282double * dp = malloc();
283struct S {...} * sp = malloc();
284\end{lstlisting}
285where the return type supplies the type/size of the allocation, which is impossible in most type systems.
286
287Call-site inferencing and nested functions provide a localized form of inheritance.
288For example, the \CFA @qsort@ only sorts in ascending order using @<@.
289However, it is trivial to locally change this behaviour:
290\begin{lstlisting}
291forall( otype T | { int ?<?( T, T ); } ) void qsort( const T * arr, size_t size ) { /* use C qsort */ }
292{       int ?<?( double x, double y ) { return x `>` y; }       $\C{// locally override behaviour}$
293        qsort( vals, size );                                    $\C{// descending sort}$
294}
295\end{lstlisting}
296Within the block, the nested version of @?<?@ performs @?>?@ and this local version overrides the built-in @?<?@ so it is passed to @qsort@.
297Hence, programmers can easily form local environments, adding and modifying appropriate functions, to maximize reuse of other existing functions and types.
298
299Finally, \CFA allows variable overloading:
300\begin{lstlisting}
301short int MAX = ...;   int MAX = ...;  double MAX = ...;
302short int s = MAX;    int i = MAX;    double d = MAX;   $\C{// select correct MAX}$
303\end{lstlisting}
304Here, the single name @MAX@ replaces all the C type-specific names: @SHRT_MAX@, @INT_MAX@, @DBL_MAX@.
305As well, restricted constant overloading is allowed for the values @0@ and @1@, which have special status in C, \eg the value @0@ is both an integer and a pointer literal, so its meaning depends on context.
306In addition, several operations are defined in terms values @0@ and @1@, \eg:
307\begin{lstlisting}
308int x;
309if (x) x++                                                                      $\C{// if (x != 0) x += 1;}$
310\end{lstlisting}
311Every @if@ and iteration statement in C compares the condition with @0@, and every increment and decrement operator is semantically equivalent to adding or subtracting the value @1@ and storing the result.
312Due to these rewrite rules, the values @0@ and @1@ have the types @zero_t@ and @one_t@ in \CFA, which allows overloading various operations for new types that seamlessly connect to all special @0@ and @1@ contexts.
313The types @zero_t@ and @one_t@ have special built in implicit conversions to the various integral types, and a conversion to pointer types for @0@, which allows standard C code involving @0@ and @1@ to work as normal.
314
315
316\subsection{Traits}
317
318\CFA provides \emph{traits} to name a group of type assertions, where the trait name allows specifying the same set of assertions in multiple locations, preventing repetition mistakes at each function declaration:
319\begin{lstlisting}
320trait summable( otype T ) {
321        void ?{}( T *, zero_t );                                $\C{// constructor from 0 literal}$
322        T ?+?( T, T );                                                  $\C{// assortment of additions}$
323        T ?+=?( T *, T );
324        T ++?( T * );
325        T ?++( T * ); };
326forall( otype T `| summable( T )` ) T sum( T a[$\,$], size_t size ) {  // use trait
327        `T` total = { `0` };                                    $\C{// instantiate T from 0 by calling its constructor}$
328        for ( unsigned int i = 0; i < size; i += 1 ) total `+=` a[i]; $\C{// select appropriate +}$
329        return total; }
330\end{lstlisting}
331
332In fact, the set of @summable@ trait operators is incomplete, as it is missing assignment for type @T@, but @otype@ is syntactic sugar for the following implicit trait:
333\begin{lstlisting}
334trait otype( dtype T | sized(T) ) {  // sized is a pseudo-trait for types with known size and alignment
335        void ?{}( T * );                                                $\C{// default constructor}$
336        void ?{}( T *, T );                                             $\C{// copy constructor}$
337        void ?=?( T *, T );                                             $\C{// assignment operator}$
338        void ^?{}( T * ); };                                    $\C{// destructor}$
339\end{lstlisting}
340Given the information provided for an @otype@, variables of polymorphic type can be treated as if they were a complete type: stack-allocatable, default or copy-initialized, assigned, and deleted.
341
342In summation, the \CFA type-system uses \emph{nominal typing} for concrete types, matching with the C type-system, and \emph{structural typing} for polymorphic types.
343Hence, trait names play no part in type equivalence;
344the names are simply macros for a list of polymorphic assertions, which are expanded at usage sites.
345Nevertheless, trait names form a logical subtype-hierarchy with @dtype@ at the top, where traits often contain overlapping assertions, \eg operator @+@.
346Traits are used like interfaces in Java or abstract base-classes in \CC, but without the nominal inheritance-relationships.
347Instead, each polymorphic function (or generic type) defines the structural type needed for its execution (polymorphic type-key), and this key is fulfilled at each call site from the lexical environment, which is similar to Go~\cite{Go} interfaces.
348Hence, new lexical scopes and nested functions are used extensively to create local subtypes, as in the @qsort@ example, without having to manage a nominal-inheritance hierarchy.
349(Nominal inheritance can be approximated with traits using marker variables or functions, as is done in Go.)
350
351% Nominal inheritance can be simulated with traits using marker variables or functions:
352% \begin{lstlisting}
353% trait nominal(otype T) {
354%     T is_nominal;
355% };
356% int is_nominal;                                                               $\C{// int now satisfies the nominal trait}$
357% \end{lstlisting}
358%
359% Traits, however, are significantly more powerful than nominal-inheritance interfaces; most notably, traits may be used to declare a relationship \emph{among} multiple types, a property that may be difficult or impossible to represent in nominal-inheritance type systems:
360% \begin{lstlisting}
361% trait pointer_like(otype Ptr, otype El) {
362%     lvalue El *?(Ptr);                                                $\C{// Ptr can be dereferenced into a modifiable value of type El}$
363% }
364% struct list {
365%     int value;
366%     list * next;                                                              $\C{// may omit "struct" on type names as in \CC}$
367% };
368% typedef list * list_iterator;
369%
370% lvalue int *?( list_iterator it ) { return it->value; }
371% \end{lstlisting}
372% In the example above, @(list_iterator, int)@ satisfies @pointer_like@ by the user-defined dereference function, and @(list_iterator, list)@ also satisfies @pointer_like@ by the built-in dereference operator for pointers. Given a declaration @list_iterator it@, @*it@ can be either an @int@ or a @list@, with the meaning disambiguated by context (\eg @int x = *it;@ interprets @*it@ as an @int@, while @(*it).value = 42;@ interprets @*it@ as a @list@).
373% While a nominal-inheritance system with associated types could model one of those two relationships by making @El@ an associated type of @Ptr@ in the @pointer_like@ implementation, few such systems could model both relationships simultaneously.
374
375
376\section{Generic Types}
377
378One of the known shortcomings of standard C is that it does not provide reusable type-safe abstractions for generic data structures and algorithms.
379Broadly speaking, there are three approaches to implement abstract data-structures in C.
380One approach is to write bespoke data-structures for each context in which they are needed.
381While this approach is flexible and supports integration with the C type-checker and tooling, it is also tedious and error-prone, especially for more complex data structures.
382A second approach is to use @void *@--based polymorphism, \eg the C standard-library functions @bsearch@ and @qsort@; an approach which does allow reuse of code for common functionality.
383However, basing all polymorphism on @void *@ eliminates the type-checker's ability to ensure that argument types are properly matched, often requiring a number of extra function parameters, pointer indirection, and dynamic allocation that would not otherwise be needed.
384A third approach to generic code is to use preprocessor macros, which does allow the generated code to be both generic and type-checked, but errors may be difficult to interpret.
385Furthermore, writing and using preprocessor macros can be unnatural and inflexible.
386
387\CC, Java, and other languages use \emph{generic types} to produce type-safe abstract data-types.
388\CFA also implements generic types that integrate efficiently and naturally with the existing polymorphic functions, while retaining backwards compatibility with C and providing separate compilation.
389However, for known concrete parameters, the generic-type definition can be inlined, like \CC templates.
390
391A generic type can be declared by placing a @forall@ specifier on a @struct@ or @union@ declaration, and instantiated using a parenthesized list of types after the type name:
392\begin{lstlisting}
393forall( otype R, otype S ) struct pair {
394        R first;
395        S second;
396};
397forall( otype T ) T value( pair( const char *, T ) p ) { return p.second; }
398forall( dtype F, otype T ) T value_p( pair( F *, T * ) p ) { return * p.second; }
399pair( const char *, int ) p = { "magic", 42 };
400int magic = value( p );
401pair( void *, int * ) q = { 0, &p.second };
402magic = value_p( q );
403double d = 1.0;
404pair( double *, double * ) r = { &d, &d };
405d = value_p( r );
406\end{lstlisting}
407
408\CFA classifies generic types as either \emph{concrete} or \emph{dynamic}.
409Concrete types have a fixed memory layout regardless of type parameters, while dynamic types vary in memory layout depending on their type parameters.
410A type may have polymorphic parameters but still be concrete, called \emph{dtype-static}.
411Polymorphic pointers are an example of dtype-static types, \eg @forall(dtype T) T *@ is a polymorphic type, but for any @T@, @T *@  is a fixed-sized pointer, and therefore, can be represented by a @void *@ in code generation.
412
413\CFA generic types also allow checked argument-constraints.
414For example, the following declaration of a sorted set-type ensures the set key supports equality and relational comparison:
415\begin{lstlisting}
416forall( otype Key | { _Bool ?==?(Key, Key); _Bool ?<?(Key, Key); } ) struct sorted_set;
417\end{lstlisting}
418
419
420\subsection{Concrete Generic-Types}
421
422The \CFA translator template-expands concrete generic-types into new structure types, affording maximal inlining.
423To enable inter-operation among equivalent instantiations of a generic type, the translator saves the set of instantiations currently in scope and reuses the generated structure declarations where appropriate.
424A function declaration that accepts or returns a concrete generic-type produces a declaration for the instantiated structure in the same scope, which all callers may reuse.
425For example, the concrete instantiation for @pair( const char *, int )@ is:
426\begin{lstlisting}
427struct _pair_conc1 {
428        const char * first;
429        int second;
430};
431\end{lstlisting}
432
433A concrete generic-type with dtype-static parameters is also expanded to a structure type, but this type is used for all matching instantiations.
434In the above example, the @pair( F *, T * )@ parameter to @value_p@ is such a type; its expansion is below and it is used as the type of the variables @q@ and @r@ as well, with casts for member access where appropriate:
435\begin{lstlisting}
436struct _pair_conc0 {
437        void * first;
438        void * second;
439};
440\end{lstlisting}
441
442
443\subsection{Dynamic Generic-Types}
444
445Though \CFA implements concrete generic-types efficiently, it also has a fully general system for dynamic generic types.
446As mentioned in Section~\ref{sec:poly-fns}, @otype@ function parameters (in fact all @sized@ polymorphic parameters) come with implicit size and alignment parameters provided by the caller.
447Dynamic generic-types also have an \emph{offset array} containing structure-member offsets.
448A dynamic generic-union needs no such offset array, as all members are at offset 0, but size and alignment are still necessary.
449Access to members of a dynamic structure is provided at runtime via base-displacement addressing with the structure pointer and the member offset (similar to the @offsetof@ macro), moving a compile-time offset calculation to runtime.
450
451The offset arrays are statically generated where possible.
452If a dynamic generic-type is declared to be passed or returned by value from a polymorphic function, the translator can safely assume the generic type is complete (\ie has a known layout) at any call-site, and the offset array is passed from the caller;
453if the generic type is concrete at the call site, the elements of this offset array can even be statically generated using the C @offsetof@ macro.
454As an example, @p.second@ in the @value@ function above is implemented as @*(p + _offsetof_pair[1])@, where @p@ is a @void *@, and @_offsetof_pair@ is the offset array passed into @value@ for @pair( const char *, T )@.
455The offset array @_offsetof_pair@ is generated at the call site as @size_t _offsetof_pair[] = { offsetof(_pair_conc1, first), offsetof(_pair_conc1, second) }@.
456
457In some cases the offset arrays cannot be statically generated.
458For instance, modularity is generally provided in C by including an opaque forward-declaration of a structure and associated accessor and mutator functions in a header file, with the actual implementations in a separately-compiled @.c@ file.
459\CFA supports this pattern for generic types, but the caller does not know the actual layout or size of the dynamic generic-type, and only holds it by a pointer.
460The \CFA translator automatically generates \emph{layout functions} for cases where the size, alignment, and offset array of a generic struct cannot be passed into a function from that function's caller.
461These layout functions take as arguments pointers to size and alignment variables and a caller-allocated array of member offsets, as well as the size and alignment of all @sized@ parameters to the generic structure (un@sized@ parameters are forbidden from being used in a context that affects layout).
462Results of these layout functions are cached so that they are only computed once per type per function. %, as in the example below for @pair@.
463Layout functions also allow generic types to be used in a function definition without reflecting them in the function signature.
464For instance, a function that strips duplicate values from an unsorted @vector(T)@ would likely have a pointer to the vector as its only explicit parameter, but use some sort of @set(T)@ internally to test for duplicate values.
465This function could acquire the layout for @set(T)@ by calling its layout function with the layout of @T@ implicitly passed into the function.
466
467Whether a type is concrete, dtype-static, or dynamic is decided solely on the @forall@'s type parameters.
468This design allows opaque forward declarations of generic types, \eg @forall(otype T)@ @struct Box@ -- like in C, all uses of @Box(T)@ can be separately compiled, and callers from other translation units know the proper calling conventions to use.
469If the definition of a structure type is included in deciding whether a generic type is dynamic or concrete, some further types may be recognized as dtype-static (\eg @forall(otype T)@ @struct unique_ptr { T * p }@ does not depend on @T@ for its layout, but the existence of an @otype@ parameter means that it \emph{could}.), but preserving separate compilation (and the associated C compatibility) in the existing design is judged to be an appropriate trade-off.
470
471
472\subsection{Applications}
473\label{sec:generic-apps}
474
475The reuse of dtype-static structure instantiations enables useful programming patterns at zero runtime cost.
476The most important such pattern is using @forall(dtype T) T *@ as a type-checked replacement for @void *@, \eg creating a lexicographic comparison for pairs of pointers used by @bsearch@ or @qsort@:
477\begin{lstlisting}
478forall(dtype T) int lexcmp( pair( T *, T * ) * a, pair( T *, T * ) * b, int (* cmp)( T *, T * ) ) {
479        return cmp( a->first, b->first ) ? : cmp( a->second, b->second );
480}
481\end{lstlisting}
482Since @pair(T *, T * )@ is a concrete type, there are no implicit parameters passed to @lexcmp@, so the generated code is identical to a function written in standard C using @void *@, yet the \CFA version is type-checked to ensure the fields of both pairs and the arguments to the comparison function match in type.
483
484Another useful pattern enabled by reused dtype-static type instantiations is zero-cost \emph{tag-structures}.
485Sometimes information is only used for type-checking and can be omitted at runtime, \eg:
486\begin{lstlisting}
487forall(dtype Unit) struct scalar { unsigned long value; };
488struct metres {};
489struct litres {};
490
491forall(dtype U) scalar(U) ?+?( scalar(U) a, scalar(U) b ) {
492        return (scalar(U)){ a.value + b.value };
493}
494scalar(metres) half_marathon = { 21093 };
495scalar(litres) swimming_pool = { 2500000 };
496scalar(metres) marathon = half_marathon + half_marathon;
497scalar(litres) two_pools = swimming_pool + swimming_pool;
498marathon + swimming_pool;                                       $\C{// compilation ERROR}$
499\end{lstlisting}
500@scalar@ is a dtype-static type, so all uses have a single structure definition, containing @unsigned long@, and can share the same implementations of common functions like @?+?@.
501These implementations may even be separately compiled, unlike \CC template functions.
502However, the \CFA type-checker ensures matching types are used by all calls to @?+?@, preventing nonsensical computations like adding a length to a volume.
503
504
505\section{Tuples}
506\label{sec:tuples}
507
508In many languages, functions can return at most one value;
509however, many operations have multiple outcomes, some exceptional.
510Consider C's @div@ and @remquo@ functions, which return the quotient and remainder for a division of integer and floating-point values, respectively.
511\begin{lstlisting}
512typedef struct { int quo, rem; } div_t;         $\C{// from include stdlib.h}$
513div_t div( int num, int den );
514double remquo( double num, double den, int * quo );
515div_t qr = div( 13, 5 );                                        $\C{// return quotient/remainder aggregate}$
516int q;
517double r = remquo( 13.5, 5.2, &q );                     $\C{// return remainder, alias quotient}$
518\end{lstlisting}
519@div@ aggregates the quotient/remainder in a structure, while @remquo@ aliases a parameter to an argument.
520Both approaches are awkward.
521Alternatively, a programming language can directly support returning multiple values, \eg in \CFA:
522\begin{lstlisting}
523[ int, int ] div( int num, int den );           $\C{// return two integers}$
524[ double, double ] div( double num, double den ); $\C{// return two doubles}$
525int q, r;                                                                       $\C{// overloaded variable names}$
526double q, r;
527[ q, r ] = div( 13, 5 );                                        $\C{// select appropriate div and q, r}$
528[ q, r ] = div( 13.5, 5.2 );                            $\C{// assign into tuple}$
529\end{lstlisting}
530Clearly, this approach is straightforward to understand and use;
531therefore, why do few programming languages support this obvious feature or provide it awkwardly?
532The answer is that there are complex consequences that cascade through multiple aspects of the language, especially the type-system.
533This section show these consequences and how \CFA handles them.
534
535
536\subsection{Tuple Expressions}
537
538The addition of multiple-return-value functions (MRVF) are useless without a syntax for accepting multiple values at the call-site.
539The simplest mechanism for capturing the return values is variable assignment, allowing the values to be retrieved directly.
540As such, \CFA allows assigning multiple values from a function into multiple variables, using a square-bracketed list of lvalue expressions (as above), called a \emph{tuple}.
541
542However, functions also use \emph{composition} (nested calls), with the direct consequence that MRVFs must also support composition to be orthogonal with single-returning-value functions (SRVF), \eg:
543\begin{lstlisting}
544printf( "%d %d\n", div( 13, 5 ) );                      $\C{// return values seperated into arguments}$
545\end{lstlisting}
546Here, the values returned by @div@ are composed with the call to @printf@ by flattening the tuple into separate arguments.
547However, the \CFA type-system must support significantly more complex composition:
548\begin{lstlisting}
549[ int, int ] foo$\(_1\)$( int );                        $\C{// overloaded foo functions}$
550[ double ] foo$\(_2\)$( int );
551void bar( int, double, double );
552bar( foo( 3 ), foo( 3 ) );
553\end{lstlisting}
554The type-resolver only has the tuple return-types to resolve the call to @bar@ as the @foo@ parameters are identical, which involves unifying the possible @foo@ functions with @bar@'s parameter list.
555No combination of @foo@s are an exact match with @bar@'s parameters, so the resolver applies C conversions.
556The minimal cost is @bar( foo@$_1$@( 3 ), foo@$_2$@( 3 ) )@, giving (@int@, {\color{ForestGreen}@int@}, @double@) to (@int@, {\color{ForestGreen}@double@}, @double@) with one {\color{ForestGreen}safe} (widening) conversion from @int@ to @double@ versus ({\color{red}@double@}, {\color{ForestGreen}@int@}, {\color{ForestGreen}@int@}) to ({\color{red}@int@}, {\color{ForestGreen}@double@}, {\color{ForestGreen}@double@}) with one {\color{red}unsafe} (narrowing) conversion from @double@ to @int@ and two safe conversions.
557
558
559\subsection{Tuple Variables}
560
561An important observation from function composition is that new variable names are not required to initialize parameters from an MRVF.
562\CFA also allows declaration of tuple variables that can be initialized from an MRVF, since it can be awkward to declare multiple variables of different types, \eg:
563\begin{lstlisting}
564[ int, int ] qr = div( 13, 5 );                         $\C{// tuple-variable declaration and initialization}$
565[ double, double ] qr = div( 13.5, 5.2 );
566\end{lstlisting}
567where the tuple variable-name serves the same purpose as the parameter name(s).
568Tuple variables can be composed of any types, except for array types, since array sizes are generally unknown in C.
569
570One way to access the tuple-variable components is with assignment or composition:
571\begin{lstlisting}
572[ q, r ] = qr;                                                          $\C{// access tuple-variable components}$
573printf( "%d %d\n", qr );
574\end{lstlisting}
575\CFA also supports \emph{tuple indexing} to access single components of a tuple expression:
576\begin{lstlisting}
577[int, int] * p = &qr;                                           $\C{// tuple pointer}$
578int rem = qr`.1`;                                                       $\C{// access remainder}$
579int quo = div( 13, 5 )`.0`;                                     $\C{// access quotient}$
580p`->0` = 5;                                                                     $\C{// change quotient}$
581bar( qr`.1`, qr );                                                      $\C{// pass remainder and quotient/remainder}$
582rem = [div( 13, 5 ), 42]`.0.1`;                         $\C{// access 2nd component of 1st component of tuple expression}$
583\end{lstlisting}
584
585
586\subsection{Flattening and Restructuring}
587
588In function call contexts, tuples support implicit flattening and restructuring conversions.
589Tuple flattening recursively expands a tuple into the list of its basic components.
590Tuple structuring packages a list of expressions into a value of tuple type, \eg:
591%\lstDeleteShortInline@%
592%\par\smallskip
593%\begin{tabular}{@{}l@{\hspace{1.5\parindent}}||@{\hspace{1.5\parindent}}l@{}}
594\begin{lstlisting}
595int f( int, int );
596int g( [int, int] );
597int h( int, [int, int] );
598[int, int] x;
599int y;
600f( x );                 $\C{// flatten}$
601g( y, 10 );             $\C{// structure}$
602h( x, y );              $\C{// flatten and structure}$
603\end{lstlisting}
604%\end{lstlisting}
605%&
606%\begin{lstlisting}
607%\end{tabular}
608%\smallskip\par\noindent
609%\lstMakeShortInline@%
610In the call to @f@, @x@ is implicitly flattened so the components of @x@ are passed as the two arguments.
611In the call to @g@, the values @y@ and @10@ are structured into a single argument of type @[int, int]@ to match the parameter type of @g@.
612Finally, in the call to @h@, @x@ is flattened to yield an argument list of length 3, of which the first component of @x@ is passed as the first parameter of @h@, and the second component of @x@ and @y@ are structured into the second argument of type @[int, int]@.
613The flexible structure of tuples permits a simple and expressive function call syntax to work seamlessly with both SRVF and MRVF, and with any number of arguments of arbitrarily complex structure.
614
615
616\subsection{Tuple Assignment}
617
618An assignment where the left side is a tuple type is called \emph{tuple assignment}.
619There are two kinds of tuple assignment depending on whether the right side of the assignment operator has a tuple type or a non-tuple type, called \emph{multiple} and \emph{mass assignment}, respectively.
620%\lstDeleteShortInline@%
621%\par\smallskip
622%\begin{tabular}{@{}l@{\hspace{1.5\parindent}}||@{\hspace{1.5\parindent}}l@{}}
623\begin{lstlisting}
624int x = 10;
625double y = 3.5;
626[int, double] z;
627z = [x, y];                                                                     $\C{// multiple assignment}$
628[x, y] = z;                                                                     $\C{// multiple assignment}$
629z = 10;                                                                         $\C{// mass assignment}$
630[y, x] = 3.14;                                                          $\C{// mass assignment}$
631\end{lstlisting}
632%\end{lstlisting}
633%&
634%\begin{lstlisting}
635%\end{tabular}
636%\smallskip\par\noindent
637%\lstMakeShortInline@%
638Both kinds of tuple assignment have parallel semantics, so that each value on the left and right side is evaluated before any assignments occur.
639As a result, it is possible to swap the values in two variables without explicitly creating any temporary variables or calling a function, \eg, @[x, y] = [y, x]@.
640This semantics means mass assignment differs from C cascading assignment (\eg @a = b = c@) in that conversions are applied in each individual assignment, which prevents data loss from the chain of conversions that can happen during a cascading assignment.
641For example, @[y, x] = 3.14@ performs the assignments @y = 3.14@ and @x = 3.14@, yielding @y == 3.14@ and @x == 3@;
642whereas, C cascading assignment @y = x = 3.14@ performs the assignments @x = 3.14@ and @y = x@, yielding @3@ in @y@ and @x@.
643Finally, tuple assignment is an expression where the result type is the type of the left-hand side of the assignment, just like all other assignment expressions in C.
644This example shows mass, multiple, and cascading assignment used in one expression:
645\begin{lstlisting}
646void f( [int, int] );
647f( [x, y] = z = 1.5 );                                          $\C{// assignments in parameter list}$
648\end{lstlisting}
649
650
651\subsection{Member Access}
652
653It is also possible to access multiple fields from a single expression using a \emph{member-access}.
654The result is a single tuple-valued expression whose type is the tuple of the types of the members, \eg:
655\begin{lstlisting}
656struct S { int x; double y; char * z; } s;
657s.[x, y, z] = 0;
658\end{lstlisting}
659Here, the mass assignment sets all members of @s@ to zero.
660Since tuple-index expressions are a form of member-access expression, it is possible to use tuple-index expressions in conjunction with member tuple expressions to manually restructure a tuple (\eg rearrange, drop, and duplicate components).
661%\lstDeleteShortInline@%
662%\par\smallskip
663%\begin{tabular}{@{}l@{\hspace{1.5\parindent}}||@{\hspace{1.5\parindent}}l@{}}
664\begin{lstlisting}
665[int, int, long, double] x;
666void f( double, long );
667x.[0, 1] = x.[1, 0];                                            $\C{// rearrange: [x.0, x.1] = [x.1, x.0]}$
668f( x.[0, 3] );                                                          $\C{// drop: f(x.0, x.3)}$
669[int, int, int] y = x.[2, 0, 2];                        $\C{// duplicate: [y.0, y.1, y.2] = [x.2, x.0.x.2]}$
670\end{lstlisting}
671%\end{lstlisting}
672%&
673%\begin{lstlisting}
674%\end{tabular}
675%\smallskip\par\noindent
676%\lstMakeShortInline@%
677It is also possible for a member access to contain other member accesses, \eg:
678\begin{lstlisting}
679struct A { double i; int j; };
680struct B { int * k; short l; };
681struct C { int x; A y; B z; } v;
682v.[x, y.[i, j], z.k];                                           $\C{// [v.x, [v.y.i, v.y.j], v.z.k]}$
683\end{lstlisting}
684
685
686\begin{comment}
687\subsection{Casting}
688
689In C, the cast operator is used to explicitly convert between types.
690In \CFA, the cast operator has a secondary use as type ascription.
691That is, a cast can be used to select the type of an expression when it is ambiguous, as in the call to an overloaded function:
692\begin{lstlisting}
693int f();     // (1)
694double f()// (2)
695
696f();       // ambiguous - (1),(2) both equally viable
697(int)f()// choose (2)
698\end{lstlisting}
699
700Since casting is a fundamental operation in \CFA, casts should be given a meaningful interpretation in the context of tuples.
701Taking a look at standard C provides some guidance with respect to the way casts should work with tuples:
702\begin{lstlisting}
703int f();
704void g();
705
706(void)f()// (1)
707(int)g()// (2)
708\end{lstlisting}
709In C, (1) is a valid cast, which calls @f@ and discards its result.
710On the other hand, (2) is invalid, because @g@ does not produce a result, so requesting an @int@ to materialize from nothing is nonsensical.
711Generalizing these principles, any cast wherein the number of components increases as a result of the cast is invalid, while casts that have the same or fewer number of components may be valid.
712
713Formally, a cast to tuple type is valid when $T_n \leq S_m$, where $T_n$ is the number of components in the target type and $S_m$ is the number of components in the source type, and for each $i$ in $[0, n)$, $S_i$ can be cast to $T_i$.
714Excess elements ($S_j$ for all $j$ in $[n, m)$) are evaluated, but their values are discarded so that they are not included in the result expression.
715This approach follows naturally from the way that a cast to @void@ works in C.
716
717For example, in
718\begin{lstlisting}
719[int, int, int] f();
720[int, [int, int], int] g();
721
722([int, double])f();           $\C{// (1)}$
723([int, int, int])g();         $\C{// (2)}$
724([void, [int, int]])g();      $\C{// (3)}$
725([int, int, int, int])g();    $\C{// (4)}$
726([int, [int, int, int]])g()$\C{// (5)}$
727\end{lstlisting}
728
729(1) discards the last element of the return value and converts the second element to @double@.
730Since @int@ is effectively a 1-element tuple, (2) discards the second component of the second element of the return value of @g@.
731If @g@ is free of side effects, this expression is equivalent to @[(int)(g().0), (int)(g().1.0), (int)(g().2)]@.
732Since @void@ is effectively a 0-element tuple, (3) discards the first and third return values, which is effectively equivalent to @[(int)(g().1.0), (int)(g().1.1)]@).
733
734Note that a cast is not a function call in \CFA, so flattening and structuring conversions do not occur for cast expressions\footnote{User-defined conversions have been considered, but for compatibility with C and the existing use of casts as type ascription, any future design for such conversions would require more precise matching of types than allowed for function arguments and parameters.}.
735As such, (4) is invalid because the cast target type contains 4 components, while the source type contains only 3.
736Similarly, (5) is invalid because the cast @([int, int, int])(g().1)@ is invalid.
737That is, it is invalid to cast @[int, int]@ to @[int, int, int]@.
738\end{comment}
739
740
741\subsection{Polymorphism}
742
743Tuples also integrate with \CFA polymorphism as a kind of generic type.
744Due to the implicit flattening and structuring conversions involved in argument passing, @otype@ and @dtype@ parameters are restricted to matching only with non-tuple types, \eg:
745\begin{lstlisting}
746forall(otype T, dtype U) void f( T x, U * y );
747f( [5, "hello"] );
748\end{lstlisting}
749where @[5, "hello"]@ is flattened, giving argument list @5, "hello"@, and @T@ binds to @int@ and @U@ binds to @const char@.
750Tuples, however, may contain polymorphic components.
751For example, a plus operator can be written to add two triples together.
752\begin{lstlisting}
753forall(otype T | { T ?+?( T, T ); }) [T, T, T] ?+?( [T, T, T] x, [T, T, T] y ) {
754        return [x.0 + y.0, x.1 + y.1, x.2 + y.2];
755}
756[int, int, int] x;
757int i1, i2, i3;
758[i1, i2, i3] = x + ([10, 20, 30]);
759\end{lstlisting}
760
761Flattening and restructuring conversions are also applied to tuple types in polymorphic type assertions.
762\begin{lstlisting}
763int f( [int, double], double );
764forall(otype T, otype U | { T f( T, U, U ); }) void g( T, U );
765g( 5, 10.21 );
766\end{lstlisting}
767Hence, function parameter and return lists are flattened for the purposes of type unification allowing the example to pass expression resolution.
768This relaxation is possible by extending the thunk scheme described by Bilson~\cite{Bilson03}.
769Whenever a candidate's parameter structure does not exactly match the formal parameter's structure, a thunk is generated to specialize calls to the actual function:
770\begin{lstlisting}
771int _thunk( int _p0, double _p1, double _p2 ) { return f( [_p0, _p1], _p2 ); }
772\end{lstlisting}
773so the thunk provides flattening and structuring conversions to inferred functions, improving the compatibility of tuples and polymorphism.
774These thunks take advantage of GCC C nested-functions to produce closures that have the usual function-pointer signature.
775
776
777\subsection{Variadic Tuples}
778\label{sec:variadic-tuples}
779
780To define variadic functions, \CFA adds a new kind of type parameter, @ttype@ (tuple type).
781Matching against a @ttype@ parameter consumes all remaining argument components and packages them into a tuple, binding to the resulting tuple of types.
782In a given parameter list, there must be at most one @ttype@ parameter that occurs last, which matches normal variadic semantics, with a strong feeling of similarity to \CCeleven variadic templates.
783As such, @ttype@ variables are also called \emph{argument packs}.
784
785Like variadic templates, the main way to manipulate @ttype@ polymorphic functions is via recursion.
786Since nothing is known about a parameter pack by default, assertion parameters are key to doing anything meaningful.
787Unlike variadic templates, @ttype@ polymorphic functions can be separately compiled.
788For example, a generalized @sum@ function written using @ttype@:
789\begin{lstlisting}
790int sum$\(_0\)$() { return 0; }
791forall(ttype Params | { int sum( Params ); } ) int sum$\(_1\)$( int x, Params rest ) {
792        return x + sum( rest );
793}
794sum( 10, 20, 30 );
795\end{lstlisting}
796Since @sum@\(_0\) does not accept any arguments, it is not a valid candidate function for the call @sum(10, 20, 30)@.
797In order to call @sum@\(_1\), @10@ is matched with @x@, and the argument resolution moves on to the argument pack @rest@, which consumes the remainder of the argument list and @Params@ is bound to @[20, 30]@.
798The process continues unitl @Params@ is bound to @[]@, requiring an assertion @int sum()@, which matches @sum@\(_0\) and terminates the recursion.
799Effectively, this algorithm traces as @sum(10, 20, 30)@ $\rightarrow$ @10 + sum(20, 30)@ $\rightarrow$ @10 + (20 + sum(30))@ $\rightarrow$ @10 + (20 + (30 + sum()))@ $\rightarrow$ @10 + (20 + (30 + 0))@.
800
801It is reasonable to take the @sum@ function a step further to enforce a minimum number of arguments:
802\begin{lstlisting}
803int sum( int x, int y ) { return x + y; }
804forall(ttype Params | { int sum( int, Params ); } ) int sum( int x, int y, Params rest ) {
805        return sum( x + y, rest );
806}
807\end{lstlisting}
808One more step permits the summation of any summable type with all arguments of the same type:
809\begin{lstlisting}
810trait summable(otype T) {
811        T ?+?( T, T );
812};
813forall(otype R | summable( R ) ) R sum( R x, R y ) {
814        return x + y;
815}
816forall(otype R, ttype Params | summable(R) | { R sum(R, Params); } ) R sum(R x, R y, Params rest) {
817        return sum( x + y, rest );
818}
819\end{lstlisting}
820Unlike C variadic functions, it is unnecessary to hard code the number and expected types.
821Furthermore, this code is extendable for any user-defined type with a @?+?@ operator.
822Summing arbitrary heterogeneous lists is possible with similar code by adding the appropriate type variables and addition operators.
823
824It is also possible to write a type-safe variadic print function to replace @printf@:
825\begin{lstlisting}
826struct S { int x, y; };
827forall(otype T, ttype Params | { void print(T); void print(Params); }) void print(T arg, Params rest) {
828        print(arg);  print(rest);
829}
830void print( char * x ) { printf( "%s", x ); }
831void print( int x ) { printf( "%d", x ); }
832void print( S s ) { print( "{ ", s.x, ",", s.y, " }" ); }
833print( "s = ", (S){ 1, 2 }, "\n" );
834\end{lstlisting}
835This example showcases a variadic-template-like decomposition of the provided argument list.
836The individual @print@ functions allow printing a single element of a type.
837The polymorphic @print@ allows printing any list of types, where as each individual type has a @print@ function.
838The individual print functions can be used to build up more complicated @print@ functions, such as @S@, which cannot be done with @printf@ in C.
839
840Finally, it is possible to use @ttype@ polymorphism to provide arbitrary argument forwarding functions.
841For example, it is possible to write @new@ as a library function:
842\begin{lstlisting}
843forall( otype R, otype S ) void ?{}( pair(R, S) *, R, S );
844forall( dtype T, ttype Params | sized(T) | { void ?{}( T *, Params ); } ) T * new( Params p ) {
845        return ((T *)malloc()){ p };                    $\C{// construct into result of malloc}$
846}
847pair( int, char ) * x = new( 42, '!' );
848\end{lstlisting}
849The @new@ function provides the combination of type-safe @malloc@ with a \CFA constructor call, making it impossible to forget constructing dynamically allocated objects.
850This function provides the type-safety of @new@ in \CC, without the need to specify the allocated type again, thanks to return-type inference.
851
852
853\subsection{Implementation}
854
855Tuples are implemented in the \CFA translator via a transformation into \emph{generic types}.
856For each $N$, the first time an $N$-tuple is seen in a scope a generic type with $N$ type parameters is generated, \eg:
857\begin{lstlisting}
858[int, int] f() {
859        [double, double] x;
860        [int, double, int] y;
861}
862\end{lstlisting}
863is transformed into:
864\begin{lstlisting}
865forall(dtype T0, dtype T1 | sized(T0) | sized(T1)) struct _tuple2 {
866        T0 field_0;                                                             $\C{// generated before the first 2-tuple}$
867        T1 field_1;
868};
869_tuple2(int, int) f() {
870        _tuple2(double, double) x;
871        forall(dtype T0, dtype T1, dtype T2 | sized(T0) | sized(T1) | sized(T2)) struct _tuple3 {
872                T0 field_0;                                                     $\C{// generated before the first 3-tuple}$
873                T1 field_1;
874                T2 field_2;
875        };
876        _tuple3(int, double, int) y;
877}
878\end{lstlisting}
879\begin{sloppypar}
880Tuple expressions are then simply converted directly into compound literals, \eg @[5, 'x', 1.24]@ becomes @(_tuple3(int, char, double)){ 5, 'x', 1.24 }@.
881\end{sloppypar}
882
883\begin{comment}
884Since tuples are essentially structures, tuple indexing expressions are just field accesses:
885\begin{lstlisting}
886void f(int, [double, char]);
887[int, double] x;
888
889x.0+x.1;
890printf("%d %g\n", x);
891f(x, 'z');
892\end{lstlisting}
893Is transformed into:
894\begin{lstlisting}
895void f(int, _tuple2(double, char));
896_tuple2(int, double) x;
897
898x.field_0+x.field_1;
899printf("%d %g\n", x.field_0, x.field_1);
900f(x.field_0, (_tuple2){ x.field_1, 'z' });
901\end{lstlisting}
902Note that due to flattening, @x@ used in the argument position is converted into the list of its fields.
903In the call to @f@, the second and third argument components are structured into a tuple argument.
904Similarly, tuple member expressions are recursively expanded into a list of member access expressions.
905
906Expressions that may contain side effects are made into \emph{unique expressions} before being expanded by the flattening conversion.
907Each unique expression is assigned an identifier and is guaranteed to be executed exactly once:
908\begin{lstlisting}
909void g(int, double);
910[int, double] h();
911g(h());
912\end{lstlisting}
913Internally, this expression is converted to two variables and an expression:
914\begin{lstlisting}
915void g(int, double);
916[int, double] h();
917
918_Bool _unq0_finished_ = 0;
919[int, double] _unq0;
920g(
921        (_unq0_finished_ ? _unq0 : (_unq0 = f(), _unq0_finished_ = 1, _unq0)).0,
922        (_unq0_finished_ ? _unq0 : (_unq0 = f(), _unq0_finished_ = 1, _unq0)).1,
923);
924\end{lstlisting}
925Since argument evaluation order is not specified by the C programming language, this scheme is built to work regardless of evaluation order.
926The first time a unique expression is executed, the actual expression is evaluated and the accompanying boolean is set to true.
927Every subsequent evaluation of the unique expression then results in an access to the stored result of the actual expression.
928Tuple member expressions also take advantage of unique expressions in the case of possible impurity.
929
930Currently, the \CFA translator has a very broad, imprecise definition of impurity, where any function call is assumed to be impure.
931This notion could be made more precise for certain intrinsic, auto-generated, and builtin functions, and could analyze function bodies when they are available to recursively detect impurity, to eliminate some unique expressions.
932
933The various kinds of tuple assignment, constructors, and destructors generate GNU C statement expressions.
934A variable is generated to store the value produced by a statement expression, since its fields may need to be constructed with a non-trivial constructor and it may need to be referred to multiple time, \eg in a unique expression.
935The use of statement expressions allows the translator to arbitrarily generate additional temporary variables as needed, but binds the implementation to a non-standard extension of the C language.
936However, there are other places where the \CFA translator makes use of GNU C extensions, such as its use of nested functions, so this restriction is not new.
937\end{comment}
938
939
940\section{Control Structures}
941
942
943\subsection{\texorpdfstring{Labelled \LstKeywordStyle{continue} / \LstKeywordStyle{break}}{Labelled continue / break}}
944
945While C provides @continue@ and @break@ statements for altering control flow, both are restricted to one level of nesting for a particular control structure.
946Unfortunately, this restriction forces programmers to use @goto@ to achieve the equivalent control-flow for more than one level of nesting.
947To prevent having to switch to the @goto@, \CFA extends the @continue@ and @break@ with a target label to support static multi-level exit~\cite{Buhr85}, as in Java.
948For both @continue@ and @break@, the target label must be directly associated with a @for@, @while@ or @do@ statement;
949for @break@, the target label can also be associated with a @switch@, @if@ or compound (@{}@) statement.
950Figure~\ref{f:MultiLevelExit} shows @continue@ and @break@ indicating the specific control structure, and the corresponding C program using only @goto@ and labels.
951The innermost loop has 7 exit points, which cause continuation or termination of one or more of the 7 nested control-structures.
952
953\begin{figure}
954\lstDeleteShortInline@%
955\begin{tabular}{@{\hspace{\parindentlnth}}l@{\hspace{\parindentlnth}}l@{\hspace{\parindentlnth}}l@{}}
956\multicolumn{1}{@{\hspace{\parindentlnth}}c@{\hspace{\parindentlnth}}}{\textbf{\CFA}}   & \multicolumn{1}{@{\hspace{\parindentlnth}}c}{\textbf{C}}      \\
957\begin{cfa}
958`LC:` {
959        ... $declarations$ ...
960        `LS:` switch ( ... ) {
961          case 3:
962                `LIF:` if ( ... ) {
963                        `LF:` for ( ... ) {
964                                `LW:` while ( ... ) {
965                                        ... break `LC`; ...
966                                        ... break `LS`; ...
967                                        ... break `LIF`; ...
968                                        ... continue `LF;` ...
969                                        ... break `LF`; ...
970                                        ... continue `LW`; ...
971                                        ... break `LW`; ...
972                                } // while
973                        } // for
974                } else {
975                        ... break `LIF`; ...
976                } // if
977        } // switch
978} // compound
979\end{cfa}
980&
981\begin{cfa}
982{
983        ... $declarations$ ...
984        switch ( ... ) {
985          case 3:
986                if ( ... ) {
987                        for ( ... ) {
988                                while ( ... ) {
989                                        ... goto `LC`; ...
990                                        ... goto `LS`; ...
991                                        ... goto `LIF`; ...
992                                        ... goto `LFC`; ...
993                                        ... goto `LFB`; ...
994                                        ... goto `LWC`; ...
995                                        ... goto `LWB`; ...
996                                  `LWC`: ; } `LWB:` ;
997                          `LFC:` ; } `LFB:` ;
998                } else {
999                        ... goto `LIF`; ...
1000                } `L3:` ;
1001        } `LS:` ;
1002} `LC:` ;
1003\end{cfa}
1004&
1005\begin{cfa}
1006
1007
1008
1009
1010
1011
1012
1013// terminate compound
1014// terminate switch
1015// terminate if
1016// continue loop
1017// terminate loop
1018// continue loop
1019// terminate loop
1020
1021
1022
1023// terminate if
1024
1025
1026
1027\end{cfa}
1028\end{tabular}
1029\lstMakeShortInline@%
1030\caption{Multi-level Exit}
1031\label{f:MultiLevelExit}
1032\end{figure}
1033
1034Both labelled @continue@ and @break@ are a @goto@ restricted in the following ways:
1035\begin{itemize}
1036\item
1037They cannot create a loop, which means only the looping constructs cause looping.
1038This restriction means all situations resulting in repeated execution are clearly delineated.
1039\item
1040They cannot branch into a control structure.
1041This restriction prevents missing declarations and/or initializations at the start of a control structure resulting in undefined behaviour.
1042\end{itemize}
1043The advantage of the labelled @continue@/@break@ is allowing static multi-level exits without having to use the @goto@ statement, and tying control flow to the target control structure rather than an arbitrary point in a program.
1044Furthermore, the location of the label at the \emph{beginning} of the target control structure informs the reader (eye candy) that complex control-flow is occurring in the body of the control structure.
1045With @goto@, the label is at the end of the control structure, which fails to convey this important clue early enough to the reader.
1046Finally, using an explicit target for the transfer instead of an implicit target allows new constructs to be added or removed without affecting existing constructs.
1047The implicit targets of the current @continue@ and @break@, \ie the closest enclosing loop or @switch@, change as certain constructs are added or removed.
1048
1049\TODO{choose and fallthrough here as well?}
1050
1051
1052\subsection{\texorpdfstring{\LstKeywordStyle{with} Clause / Statement}{with Clause / Statement}}
1053\label{s:WithClauseStatement}
1054
1055Grouping heterogenous data into \newterm{aggregate}s (structure/union) is a common programming practice, and an aggregate can be further organized into more complex structures, such as arrays and containers:
1056\begin{cfa}
1057struct S {                                                                      $\C{// aggregate}$
1058        char c;                                                                 $\C{// fields}$
1059        int i;
1060        double d;
1061};
1062S s, as[10];
1063\end{cfa}
1064However, routines manipulating aggregates must repeat the aggregate name to access its containing fields:
1065\begin{cfa}
1066void f( S s ) {
1067        `s.`c; `s.`i; `s.`d;                                    $\C{// access containing fields}$
1068}
1069\end{cfa}
1070A similar situation occurs in object-oriented programming, \eg \CC:
1071\begin{C++}
1072class C {
1073        char c;                                                                 $\C{// fields}$
1074        int i;
1075        double d;
1076        int mem() {                                                             $\C{// implicit "this" parameter}$
1077                `this->`c; `this->`i; `this->`d;        $\C{// access containing fields}$
1078        }
1079}
1080\end{C++}
1081Nesting of member routines in a \lstinline[language=C++]@class@ allows eliding \lstinline[language=C++]@this->@ because of lexical scoping.
1082However, for other aggregate parameters, qualification is necessary:
1083\begin{cfa}
1084struct T { double m, n; };
1085int C::mem( T & t ) {                                           $\C{// multiple aggregate parameters}$
1086        c; i; d;                                                                $\C{\color{red}// this-\textgreater.c, this-\textgreater.i, this-\textgreater.d}$
1087        `t.`m; `t.`n;                                                   $\C{// must qualify}$
1088}
1089\end{cfa}
1090
1091% In object-oriented programming, there is an implicit first parameter, often names @self@ or @this@, which is elided.
1092% In any programming language, some functions have a naturally close relationship with a particular data type.
1093% Object-oriented programming allows this close relationship to be codified in the language by making such functions \emph{class methods} of their related data type.
1094% Class methods have certain privileges with respect to their associated data type, notably un-prefixed access to the fields of that data type.
1095% When writing C functions in an object-oriented style, this un-prefixed access is swiftly missed, as access to fields of a @Foo* f@ requires an extra three characters @f->@ every time, which disrupts coding flow and clutters the produced code.
1096%
1097% \TODO{Fill out section. Be sure to mention arbitrary expressions in with-blocks, recent change driven by Thierry to prioritize field name over parameters.}
1098
1099To simplify the programmer experience, \CFA provides a @with@ clause/statement (see Pascal~\cite[\S~4.F]{Pascal}) to elide aggregate qualification to fields by opening a scope containing the field identifiers.
1100Hence, the qualified fields become variables with the side-effect that it is easier to optimizing field references in a block.
1101\begin{cfa}
1102void f( S s ) `with( s )` {                                     $\C{// with clause}$
1103        c; i; d;                                                                $\C{\color{red}// s.c, s.i, s.d}$
1104}
1105\end{cfa}
1106and the equivalence for object-style programming is:
1107\begin{cfa}
1108int mem( S & this ) `with( this )` {            $\C{// with clause}$
1109        c; i; d;                                                                $\C{\color{red}// this.c, this.i, this.d}$
1110}
1111\end{cfa}
1112with the generality of opening multiple aggregate-parameters:
1113\begin{cfa}
1114int mem( S & s, T & t ) `with( s, t )` {        $\C{// multiple aggregate parameters}$
1115        c; i; d;                                                                $\C{\color{red}// s.c, s.i, s.d}$
1116        m; n;                                                                   $\C{\color{red}// t.m, t.n}$
1117}
1118\end{cfa}
1119
1120In detail, the @with@ clause/statement has the form:
1121\begin{cfa}
1122$\emph{with-statement}$:
1123        'with' '(' $\emph{expression-list}$ ')' $\emph{compound-statement}$
1124\end{cfa}
1125and may appear as the body of a routine or nested within a routine body.
1126Each expression in the expression-list provides a type and object.
1127The type must be an aggregate type.
1128(Enumerations are already opened.)
1129The object is the implicit qualifier for the open structure-fields.
1130
1131All expressions in the expression list are open in ``parallel'' within the compound statement.
1132This semantic is different from Pascal, which nests the openings.
1133The difference between parallel and nesting occurs for fields with the same name but different type:
1134\begin{cfa}
1135struct S { int i; int j; double m; } s, w;
1136struct T { int i; int k; int m } t, w;
1137with( s, t ) {
1138        j + k;                                                                  $\C{// unambiguous, s.j + t.m}$
1139        m = 5.0;                                                                $\C{// unambiguous, t.m = 5.0}$
1140        m = 1;                                                                  $\C{// unambiguous, s.m = 1}$
1141        int a = s.i + m;                                                $\C{// unambiguous, a = s.i + t.i}$
1142        int b = s.i + t.i;                                              $\C{// unambiguous, qualification}$
1143        sout | (double)m | endl;                                $\C{// unambiguous, cast}$
1144        i;                                                                              $\C{// ambiguous}$
1145}
1146\end{cfa}
1147\CFA's ability to overload variables means usages of field with the same names can be automatically disambiguated, eliminating most qualification.
1148Qualification or a cast is used to disambiguate.
1149A cast may be necessary to disambiguate between the overload variables in a @with@ expression:
1150\begin{cfa}
1151with( w ) { ... }                                                       $\C{// ambiguous, same name and no context}$
1152with( (S)w ) { ... }                                            $\C{// unambiguous}$
1153\end{cfa}
1154
1155\begin{cfa}
1156struct S { int i, j; } sv;
1157with( sv ) {
1158        S & sr = sv;
1159        with( sr ) {
1160                S * sp = &sv;
1161                with( *sp ) {
1162                        i = 3; j = 4;                                   $\C{\color{red}// sp-{\textgreater}i, sp-{\textgreater}j}$
1163                }
1164                i = 3; j = 4;                                           $\C{\color{red}// sr.i, sr.j}$
1165        }
1166        i = 3; j = 4;                                                   $\C{\color{red}// sv.i, sv.j}$
1167}
1168\end{cfa}
1169
1170The statement form is used within a block:
1171\begin{cfa}
1172int foo() {
1173        struct S1 { ... } s1;
1174        struct S2 { ... } s2;
1175        `with( s1 )` {                                                  $\C{// with statement}$
1176                // access fields of s1 without qualification
1177                `with( s2 )` {                                          $\C{// nesting}$
1178                        // access fields of s1 and s2 without qualification
1179                }
1180        }
1181        `with( s1, s2 )` {
1182                // access unambiguous fields of s1 and s2 without qualification
1183        }
1184}
1185\end{cfa}
1186
1187
1188\subsection{Exception Handling ???}
1189
1190
1191\section{Declarations}
1192
1193It is important to the design team that \CFA subjectively ``feel like'' C to user programmers.
1194An important part of this subjective feel is maintaining C's procedural programming paradigm, as opposed to the object-oriented paradigm of other systems languages such as \CC and Rust.
1195Maintaining this procedural paradigm means that coding patterns that work in C will remain not only functional but idiomatic in \CFA, reducing the mental burden of retraining C programmers and switching between C and \CFA development.
1196Nonetheless, some features of object-oriented languages are undeniably convienient, and the \CFA design team has attempted to adapt them to a procedural paradigm so as to incorporate their benefits into \CFA; two of these features are resource management and name scoping.
1197
1198
1199\subsection{Alternative Declaration Syntax}
1200
1201\newcommand{\R}[1]{\Textbf{#1}}
1202\newcommand{\B}[1]{{\Textbf[blue]{#1}}}
1203\newcommand{\G}[1]{{\Textbf[OliveGreen]{#1}}}
1204
1205C declaration syntax is notoriously confusing and error prone.
1206For example, many C programmers are confused by a declaration as simple as:
1207\begin{flushleft}
1208\lstDeleteShortInline@%
1209\begin{tabular}{@{}ll@{}}
1210\begin{cfa}
1211int * x[5]
1212\end{cfa}
1213&
1214\raisebox{-0.75\totalheight}{\input{Cdecl}}
1215\end{tabular}
1216\lstMakeShortInline@%
1217\end{flushleft}
1218Is this an array of 5 pointers to integers or a pointer to an array of 5 integers?
1219The fact this declaration is unclear to many C programmers means there are productivity and safety issues even for basic programs.
1220Another example of confusion results from the fact that a routine name and its parameters are embedded within the return type, mimicking the way the return value is used at the routine's call site.
1221For example, a routine returning a pointer to an array of integers is defined and used in the following way:
1222\begin{cfa}
1223int `(*`f`())[`5`]` {...};                              $\C{// definition}$
1224 ... `(*`f`())[`3`]` += 1;                              $\C{// usage}$
1225\end{cfa}
1226Essentially, the return type is wrapped around the routine name in successive layers (like an onion).
1227While attempting to make the two contexts consistent is a laudable goal, it has not worked out in practice.
1228
1229\CFA provides its own type, variable and routine declarations, using a different syntax.
1230The new declarations place qualifiers to the left of the base type, while C declarations place qualifiers to the right of the base type.
1231In the following example, \R{red} is the base type and \B{blue} is qualifiers.
1232The \CFA declarations move the qualifiers to the left of the base type, \ie move the blue to the left of the red, while the qualifiers have the same meaning but are ordered left to right to specify a variable's type.
1233\begin{quote}
1234\lstDeleteShortInline@%
1235\lstset{moredelim=**[is][\color{blue}]{+}{+}}
1236\begin{tabular}{@{}l@{\hspace{3em}}l@{}}
1237\multicolumn{1}{c@{\hspace{3em}}}{\textbf{\CFA}}        & \multicolumn{1}{c}{\textbf{C}}        \\
1238\begin{cfa}
1239+[5] *+ `int` x1;
1240+* [5]+ `int` x2;
1241+[* [5] int]+ f`( int p )`;
1242\end{cfa}
1243&
1244\begin{cfa}
1245`int` +*+ x1 +[5]+;
1246`int` +(*+x2+)[5]+;
1247+int (*+f`( int p )`+)[5]+;
1248\end{cfa}
1249\end{tabular}
1250\lstMakeShortInline@%
1251\end{quote}
1252The only exception is bit field specification, which always appear to the right of the base type.
1253% Specifically, the character ©*© is used to indicate a pointer, square brackets ©[©\,©]© are used to represent an array or function return value, and parentheses ©()© are used to indicate a routine parameter.
1254However, unlike C, \CFA type declaration tokens are distributed across all variables in the declaration list.
1255For instance, variables ©x© and ©y© of type pointer to integer are defined in \CFA as follows:
1256\begin{quote}
1257\lstDeleteShortInline@%
1258\begin{tabular}{@{}l@{\hspace{3em}}l@{}}
1259\multicolumn{1}{c@{\hspace{3em}}}{\textbf{\CFA}}        & \multicolumn{1}{c}{\textbf{C}}        \\
1260\begin{cfa}
1261`*` int x, y;
1262\end{cfa}
1263&
1264\begin{cfa}
1265int `*`x, `*`y;
1266\end{cfa}
1267\end{tabular}
1268\lstMakeShortInline@%
1269\end{quote}
1270The downside of this semantics is the need to separate regular and pointer declarations:
1271\begin{quote}
1272\lstDeleteShortInline@%
1273\begin{tabular}{@{}l@{\hspace{3em}}l@{}}
1274\multicolumn{1}{c@{\hspace{3em}}}{\textbf{\CFA}}        & \multicolumn{1}{c}{\textbf{C}}        \\
1275\begin{cfa}
1276`*` int x;
1277int y;
1278\end{cfa}
1279&
1280\begin{cfa}
1281int `*`x, y;
1282
1283\end{cfa}
1284\end{tabular}
1285\lstMakeShortInline@%
1286\end{quote}
1287which is prescribing a safety benefit.
1288Other examples are:
1289\begin{quote}
1290\lstDeleteShortInline@%
1291\begin{tabular}{@{}l@{\hspace{3em}}l@{\hspace{2em}}l@{}}
1292\multicolumn{1}{c@{\hspace{3em}}}{\textbf{\CFA}}        & \multicolumn{1}{c@{\hspace{2em}}}{\textbf{C}} \\
1293\begin{cfa}
1294[ 5 ] int z;
1295[ 5 ] * char w;
1296* [ 5 ] double v;
1297struct s {
1298        int f0:3;
1299        * int f1;
1300        [ 5 ] * int f2;
1301};
1302\end{cfa}
1303&
1304\begin{cfa}
1305int z[ 5 ];
1306char * w[ 5 ];
1307double (* v)[ 5 ];
1308struct s {
1309        int f0:3;
1310        int * f1;
1311        int * f2[ 5 ]
1312};
1313\end{cfa}
1314&
1315\begin{cfa}
1316// array of 5 integers
1317// array of 5 pointers to char
1318// pointer to array of 5 doubles
1319
1320// common bit field syntax
1321
1322
1323
1324\end{cfa}
1325\end{tabular}
1326\lstMakeShortInline@%
1327\end{quote}
1328
1329All type qualifiers, \eg ©const©, ©volatile©, etc., are used in the normal way with the new declarations and also appear left to right, \eg:
1330\begin{quote}
1331\lstDeleteShortInline@%
1332\begin{tabular}{@{}l@{\hspace{1em}}l@{\hspace{1em}}l@{}}
1333\multicolumn{1}{c@{\hspace{1em}}}{\textbf{\CFA}}        & \multicolumn{1}{c@{\hspace{1em}}}{\textbf{C}} \\
1334\begin{cfa}
1335const * const int x;
1336const * [ 5 ] const int y;
1337\end{cfa}
1338&
1339\begin{cfa}
1340int const * const x;
1341const int (* const y)[ 5 ]
1342\end{cfa}
1343&
1344\begin{cfa}
1345// const pointer to const integer
1346// const pointer to array of 5 const integers
1347\end{cfa}
1348\end{tabular}
1349\lstMakeShortInline@%
1350\end{quote}
1351All declaration qualifiers, \eg ©extern©, ©static©, etc., are used in the normal way with the new declarations but can only appear at the start of a \CFA routine declaration,\footnote{\label{StorageClassSpecifier}
1352The placement of a storage-class specifier other than at the beginning of the declaration specifiers in a declaration is an obsolescent feature.~\cite[\S~6.11.5(1)]{C11}} \eg:
1353\begin{quote}
1354\lstDeleteShortInline@%
1355\begin{tabular}{@{}l@{\hspace{3em}}l@{\hspace{2em}}l@{}}
1356\multicolumn{1}{c@{\hspace{3em}}}{\textbf{\CFA}}        & \multicolumn{1}{c@{\hspace{2em}}}{\textbf{C}} \\
1357\begin{cfa}
1358extern [ 5 ] int x;
1359static * const int y;
1360\end{cfa}
1361&
1362\begin{cfa}
1363int extern x[ 5 ];
1364const int static * y;
1365\end{cfa}
1366&
1367\begin{cfa}
1368// externally visible array of 5 integers
1369// internally visible pointer to constant int
1370\end{cfa}
1371\end{tabular}
1372\lstMakeShortInline@%
1373\end{quote}
1374
1375The new declaration syntax can be used in other contexts where types are required, \eg casts and the pseudo-routine ©sizeof©:
1376\begin{quote}
1377\lstDeleteShortInline@%
1378\begin{tabular}{@{}l@{\hspace{3em}}l@{}}
1379\multicolumn{1}{c@{\hspace{3em}}}{\textbf{\CFA}}        & \multicolumn{1}{c}{\textbf{C}}        \\
1380\begin{cfa}
1381y = (`* int`)x;
1382i = sizeof(`[ 5 ] * int`);
1383\end{cfa}
1384&
1385\begin{cfa}
1386y = (`int *`)x;
1387i = sizeof(`int * [ 5 ]`);
1388\end{cfa}
1389\end{tabular}
1390\lstMakeShortInline@%
1391\end{quote}
1392
1393Finally, new \CFA declarations may appear together with C declarations in the same program block, but cannot be mixed within a specific declaration.
1394Therefore, a programmer has the option of either continuing to use traditional C declarations or take advantage of the new style.
1395Clearly, both styles need to be supported for some time due to existing C-style header-files, particularly for UNIX systems.
1396
1397
1398\subsection{References}
1399
1400All variables in C have an \emph{address}, a \emph{value}, and a \emph{type}; at the position in the program's memory denoted by the address, there exists a sequence of bits (the value), with the length and semantic meaning of this bit sequence defined by the type.
1401The C type system does not always track the relationship between a value and its address; a value that does not have a corresponding address is called a \emph{rvalue} (for ``right-hand value''), while a value that does have an address is called a \emph{lvalue} (for ``left-hand value''); in @int x; x = 42;@ the variable expression @x@ on the left-hand-side of the assignment is a lvalue, while the constant expression @42@ on the right-hand-side of the assignment is a rvalue.
1402Which address a value is located at is sometimes significant; the imperative programming paradigm of C relies on the mutation of values at specific addresses.
1403Within a lexical scope, lvalue exressions can be used in either their \emph{address interpretation} to determine where a mutated value should be stored or in their \emph{value interpretation} to refer to their stored value; in @x = y;@ in @{ int x, y = 7; x = y; }@, @x@ is used in its address interpretation, while y is used in its value interpretation.
1404Though this duality of interpretation is useful, C lacks a direct mechanism to pass lvalues between contexts, instead relying on \emph{pointer types} to serve a similar purpose.
1405In C, for any type @T@ there is a pointer type @T*@, the value of which is the address of a value of type @T@; a pointer rvalue can be explicitly \emph{dereferenced} to the pointed-to lvalue with the dereference operator @*?@, while the rvalue representing the address of a lvalue can be obtained with the address-of operator @&?@.
1406
1407\begin{cfa}
1408int x = 1, y = 2, * p1, * p2, ** p3;
1409p1 = &x;  $\C{// p1 points to x}$
1410p2 = &y;  $\C{// p2 points to y}$
1411p3 = &p1;  $\C{// p3 points to p1}$
1412*p2 = ((*p1 + *p2) * (**p3 - *p1)) / (**p3 - 15);
1413\end{cfa}
1414
1415Unfortunately, the dereference and address-of operators introduce a great deal of syntactic noise when dealing with pointed-to values rather than pointers, as well as the potential for subtle bugs.
1416For both brevity and clarity, it would be desirable to have the compiler figure out how to elide the dereference operators in a complex expression such as the assignment to @*p2@ above.
1417However, since C defines a number of forms of \emph{pointer arithmetic}, two similar expressions involving pointers to arithmetic types (\eg @*p1 + x@ and @p1 + x@) may each have well-defined but distinct semantics, introducing the possibility that a user programmer may write one when they mean the other, and precluding any simple algorithm for elision of dereference operators.
1418To solve these problems, \CFA introduces reference types @T&@; a @T&@ has exactly the same value as a @T*@, but where the @T*@ takes the address interpretation by default, a @T&@ takes the value interpretation by default, as below:
1419
1420\begin{cfa}
1421inx x = 1, y = 2, & r1, & r2, && r3;
1422&r1 = &x;  $\C{// r1 points to x}$
1423&r2 = &y;  $\C{// r2 points to y}$
1424&&r3 = &&r1;  $\C{// r3 points to r2}$
1425r2 = ((r1 + r2) * (r3 - r1)) / (r3 - 15);  $\C{// implicit dereferencing}$
1426\end{cfa}
1427
1428Except for auto-dereferencing by the compiler, this reference example is exactly the same as the previous pointer example.
1429Hence, a reference behaves like a variable name -- an lvalue expression which is interpreted as a value, but also has the type system track the address of that value.
1430One way to conceptualize a reference is via a rewrite rule, where the compiler inserts a dereference operator before the reference variable for each reference qualifier in the reference variable declaration, so the previous example implicitly acts like:
1431
1432\begin{cfa}
1433`*`r2 = ((`*`r1 + `*`r2) * (`**`r3 - `*`r1)) / (`**`r3 - 15);
1434\end{cfa}
1435
1436References in \CFA are similar to those in \CC, but with a couple important improvements, both of which can be seen in the example above.
1437Firstly, \CFA does not forbid references to references, unlike \CC.
1438This provides a much more orthogonal design for library implementors, obviating the need for workarounds such as @std::reference_wrapper@.
1439
1440Secondly, unlike the references in \CC which always point to a fixed address, \CFA references are rebindable.
1441This allows \CFA references to be default-initialized (\eg to a null pointer), and also to point to different addresses throughout their lifetime.
1442This rebinding is accomplished without adding any new syntax to \CFA, but simply by extending the existing semantics of the address-of operator in C.
1443In C, the address of a lvalue is always a rvalue, as in general that address is not stored anywhere in memory, and does not itself have an address.
1444In \CFA, the address of a @T&@ is a lvalue @T*@, as the address of the underlying @T@ is stored in the reference, and can thus be mutated there.
1445The result of this rule is that any reference can be rebound using the existing pointer assignment semantics by assigning a compatible pointer into the address of the reference, \eg @&r1 = &x;@ above.
1446This rebinding can occur to an arbitrary depth of reference nesting; loosely speaking, nested address-of operators will produce an lvalue nested pointer up to as deep as the reference they're applied to.
1447These explicit address-of operators can be thought of as ``cancelling out'' the implicit dereference operators, \eg @(&`*`)r1 = &x@ or @(&(&`*`)`*`)r3 = &(&`*`)r1@ or even @(&`*`)r2 = (&`*`)`*`r3@ for @&r2 = &r3@.
1448More precisely:
1449\begin{itemize}
1450        \item
1451        if @R@ is an rvalue of type {@T &@$_1 \cdots$@ &@$_r$} where $r \ge 1$ references (@&@ symbols) than @&R@ has type {@T `*`&@$_{\color{red}2} \cdots$@ &@$_{\color{red}r}$}, \\ \ie @T@ pointer with $r-1$ references (@&@ symbols).
1452       
1453        \item
1454        if @L@ is an lvalue of type {@T &@$_1 \cdots$@ &@$_l$} where $l \ge 0$ references (@&@ symbols) then @&L@ has type {@T `*`&@$_{\color{red}1} \cdots$@ &@$_{\color{red}l}$}, \\ \ie @T@ pointer with $l$ references (@&@ symbols).
1455\end{itemize}
1456
1457Since pointers and references share the same internal representation, code using either is equally performant; in fact the \CFA compiler converts references to pointers internally, and the choice between them in user code can be made based solely on convenience.
1458By analogy to pointers, \CFA references also allow cv-qualifiers:
1459
1460\begin{cfa}
1461const int cx = 5;               $\C{// cannot change cx}$
1462const int & cr = cx;    $\C{// cannot change cr's referred value}$
1463&cr = &cx;                              $\C{// rebinding cr allowed}$
1464cr = 7;                                 $\C{// ERROR, cannot change cr}$
1465int & const rc = x;             $\C{// must be initialized, like in \CC}$
1466&rc = &x;                               $\C{// ERROR, cannot rebind rc}$
1467rc = 7;                                 $\C{// x now equal to 7}$
1468\end{cfa}
1469
1470Given that a reference is meant to represent a lvalue, \CFA provides some syntactic shortcuts when initializing references.
1471There are three initialization contexts in \CFA: declaration initialization, argument/parameter binding, and return/temporary binding.
1472In each of these contexts, the address-of operator on the target lvalue may (in fact, must) be elided.
1473The syntactic motivation for this is clearest when considering overloaded operator-assignment, \eg @int ?+=?(int &, int)@; given @int x, y@, the expected call syntax is @x += y@, not @&x += y@.
1474
1475More generally, this initialization of references from lvalues rather than pointers is an instance of a ``lvalue-to-reference'' conversion rather than an elision of the address-of operator; this conversion can actually be used in any context in \CFA an implicit conversion would be allowed.
1476Similarly, use of a the value pointed to by a reference in an rvalue context can be thought of as a ``reference-to-rvalue'' conversion, and \CFA also includes a qualifier-adding ``reference-to-reference'' conversion, analagous to the @T *@ to @const T *@ conversion in standard C.
1477The final reference conversion included in \CFA is ``rvalue-to-reference'' conversion, implemented by means of an implicit temporary.
1478When an rvalue is used to initialize a reference, it is instead used to initialize a hidden temporary value with the same lexical scope as the reference, and the reference is initialized to the address of this temporary.
1479This allows complex values to be succinctly and efficiently passed to functions, without the syntactic overhead of explicit definition of a temporary variable or the runtime cost of pass-by-value.
1480\CC allows a similar binding, but only for @const@ references; the more general semantics of \CFA are an attempt to avoid the \emph{const hell} problem, in which addition of a @const@ qualifier to one reference requires a cascading chain of added qualifiers.
1481
1482\subsection{Constructors and Destructors}
1483
1484One of the strengths of C is the control over memory management it gives programmers, allowing resource release to be more consistent and precisely timed than is possible with garbage-collected memory management.
1485However, this manual approach to memory management is often verbose, and it is useful to manage resources other than memory (\eg file handles) using the same mechanism as memory.
1486\CC is well-known for an approach to manual memory management that addresses both these issues, Resource Aquisition Is Initialization (RAII), implemented by means of special \emph{constructor} and \emph{destructor} functions; we have implemented a similar feature in \CFA.
1487While RAII is a common feature of object-oriented programming languages, its inclusion in \CFA does not violate the design principle that \CFA retain the same procedural paradigm as C.
1488In particular, \CFA does not implement class-based encapsulation: neither the constructor nor any other function has privileged access to the implementation details of a type, except through the translation-unit-scope method of opaque structs provided by C.
1489
1490In \CFA, a constructor is a function named @?{}@, while a destructor is a function named @^?{}@; like other \CFA operators, these names represent the syntax used to call the constructor or destructor, \eg @x{ ... };@ or @^x{};@.
1491Every constructor and destructor must have a return type of @void@, and its first parameter must have a reference type whose base type is the type of the object the function constructs or destructs.
1492This first parameter is informally called the @this@ parameter, as in many object-oriented languages, though a programmer may give it an arbitrary name.
1493Destructors must have exactly one parameter, while constructors allow passing of zero or more additional arguments along with the @this@ parameter.
1494
1495\begin{cfa}
1496struct Array {
1497        int * data;
1498        int len;
1499};
1500
1501void ?{}( Array& arr ) {
1502        arr.len = 10;
1503        arr.data = calloc( arr.len, sizeof(int) );
1504}
1505
1506void ^?{}( Array& arr ) {
1507        free( arr.data );
1508}
1509
1510{
1511        Array x;
1512        `?{}(x);`       $\C{// implicitly compiler-generated}$
1513        // ... use x
1514        `^?{}(x);`      $\C{// implicitly compiler-generated}$
1515}
1516\end{cfa}
1517
1518In the example above, a \emph{default constructor} (\ie one with no parameters besides the @this@ parameter) and destructor are defined for the @Array@ struct, a dynamic array of @int@.
1519@Array@ is an example of a \emph{managed type} in \CFA, a type with a non-trivial constructor or destructor, or with a field of a managed type.
1520As in the example, all instances of managed types are implicitly constructed upon allocation, and destructed upon deallocation; this ensures proper initialization and cleanup of resources contained in managed types, in this case the @data@ array on the heap.
1521The exact details of the placement of these implicit constructor and destructor calls are omitted here for brevity, the interested reader should consult \cite{Schluntz17}.
1522
1523Constructor calls are intended to seamlessly integrate with existing C initialization syntax, providing a simple and familiar syntax to veteran C programmers and allowing constructor calls to be inserted into legacy C code with minimal code changes.
1524As such, \CFA also provides syntax for \emph{copy initialization} and \emph{initialization parameters}:
1525
1526\begin{cfa}
1527void ?{}( Array& arr, Array other );
1528
1529void ?{}( Array& arr, int size, int fill );
1530
1531Array y = { 20, 0xDEADBEEF }, z = y;
1532\end{cfa}
1533
1534Copy constructors have exactly two parameters, the second of which has the same type as the base type of the @this@ parameter; appropriate care is taken in the implementation to avoid recursive calls to the copy constructor when initializing this second parameter.
1535Other constructor calls look just like C initializers, except rather than using field-by-field initialization (as in C), an initialization which matches a defined constructor will call the constructor instead.
1536
1537In addition to initialization syntax, \CFA provides two ways to explicitly call constructors and destructors.
1538Explicit calls to constructors double as a placement syntax, useful for construction of member fields in user-defined constructors and reuse of large storage allocations.
1539While the existing function-call syntax works for explicit calls to constructors and destructors, \CFA also provides a more concise \emph{operator syntax} for both:
1540
1541\begin{cfa}
1542Array a, b;
1543a{};                            $\C{// default construct}$
1544b{ a };                         $\C{// copy construct}$
1545^a{};                           $\C{// destruct}$
1546a{ 5, 0xFFFFFFFF };     $\C{// explicit constructor call}$
1547\end{cfa}
1548
1549To provide a uniform type interface for @otype@ polymorphism, the \CFA compiler automatically generates a default constructor, copy constructor, assignment operator, and destructor for all types.
1550These default functions can be overridden by user-generated versions of them.
1551For compatibility with the standard behaviour of C, the default constructor and destructor for all basic, pointer, and reference types do nothing, while the copy constructor and assignment operator are bitwise copies; if default zero-initialization is desired, the default constructors can be overridden.
1552For user-generated types, the four functions are also automatically generated.
1553@enum@ types are handled the same as their underlying integral type, and unions are also bitwise copied and no-op initialized and destructed.
1554For compatibility with C, a copy constructor from the first union member type is also defined.
1555For @struct@ types, each of the four functions are implicitly defined to call their corresponding functions on each member of the struct.
1556To better simulate the behaviour of C initializers, a set of \emph{field constructors} is also generated for structures.
1557A constructor is generated for each non-empty prefix of a structure's member-list which copy-constructs the members passed as parameters and default-constructs the remaining members.
1558To allow users to limit the set of constructors available for a type, when a user declares any constructor or destructor, the corresponding generated function and all field constructors for that type are hidden from expression resolution; similarly, the generated default constructor is hidden upon declaration of any constructor.
1559These semantics closely mirror the rule for implicit declaration of constructors in \CC\cite[p.~186]{ANSI98:C++}.
1560
1561In rare situations user programmers may not wish to have constructors and destructors called; in these cases, \CFA provides an ``escape hatch'' to not call them.
1562If a variable is initialized using the syntax \lstinline|S x @= {}| it will be an \emph{unmanaged object}, and will not have constructors or destructors called.
1563Any C initializer can be the right-hand side of an \lstinline|@=| initializer, \eg  \lstinline|Array a @= { 0, 0x0 }|, with the usual C initialization semantics.
1564In addition to the expressive power, \lstinline|@=| provides a simple path for migrating legacy C code to \CFA, by providing a mechanism to incrementally convert initializers; the \CFA design team decided to introduce a new syntax for this escape hatch because we believe that our RAII implementation will handle the vast majority of code in a desirable way, and we wished to maintain familiar syntax for this common case.
1565
1566\subsection{Default Parameters}
1567
1568
1569\section{Literals}
1570
1571
1572\subsection{0/1}
1573
1574\TODO{Some text already at the end of Section~\ref{sec:poly-fns}}
1575
1576
1577\subsection{Units}
1578
1579Alternative call syntax (literal argument before routine name) to convert basic literals into user literals.
1580
1581{\lstset{language=CFA,deletedelim=**[is][]{`}{`},moredelim=**[is][\color{red}]{@}{@}}
1582\begin{cfa}
1583struct Weight { double stones; };
1584
1585void ?{}( Weight & w ) { w.stones = 0; } $\C{// operations}$
1586void ?{}( Weight & w, double w ) { w.stones = w; }
1587Weight ?+?( Weight l, Weight r ) { return (Weight){ l.stones + r.stones }; }
1588
1589Weight @?`st@( double w ) { return (Weight){ w }; } $\C{// backquote for units}$
1590Weight @?`lb@( double w ) { return (Weight){ w / 14.0 }; }
1591Weight @?`kg@( double w ) { return (Weight) { w * 0.1575}; }
1592
1593int main() {
1594        Weight w, hw = { 14 };                  $\C{// 14 stone}$
1595        w = 11@`st@ + 1@`lb@;
1596        w = 70.3@`kg@;
1597        w = 155@`lb@;
1598        w = 0x_9b_u@`lb@;                               $\C{// hexadecimal unsigned weight (155)}$
1599        w = 0_233@`lb@;                                 $\C{// octal weight (155)}$
1600        w = 5@`st@ + 8@`kg@ + 25@`lb@ + hw;
1601}
1602\end{cfa}
1603}%
1604
1605
1606\section{Evaluation}
1607\label{sec:eval}
1608
1609Though \CFA provides significant added functionality over C, these features have a low runtime penalty.
1610In fact, \CFA's features for generic programming can enable faster runtime execution than idiomatic @void *@-based C code.
1611This claim is demonstrated through a set of generic-code-based micro-benchmarks in C, \CFA, and \CC (see stack implementations in Appendix~\ref{sec:BenchmarkStackImplementation}).
1612Since all these languages share a subset essentially comprising standard C, maximal-performance benchmarks would show little runtime variance, other than in length and clarity of source code.
1613A more illustrative benchmark measures the costs of idiomatic usage of each language's features.
1614Figure~\ref{fig:BenchmarkTest} shows the \CFA benchmark tests for a generic stack based on a singly linked-list, a generic pair-data-structure, and a variadic @print@ routine similar to that in Section~\ref{sec:variadic-tuples}.
1615The benchmark test is similar for C and \CC.
1616The experiment uses element types @int@ and @pair(_Bool, char)@, and pushes $N=40M$ elements on a generic stack, copies the stack, clears one of the stacks, finds the maximum value in the other stack, and prints $N/2$ (to reduce graph height) constants.
1617
1618\begin{figure}
1619\begin{lstlisting}[xleftmargin=3\parindentlnth,aboveskip=0pt,belowskip=0pt]
1620int main( int argc, char * argv[] ) {
1621        FILE * out = fopen( "cfa-out.txt", "w" );
1622        int maxi = 0, vali = 42;
1623        stack(int) si, ti;
1624
1625        REPEAT_TIMED( "push_int", N, push( &si, vali ); )
1626        TIMED( "copy_int", ti = si; )
1627        TIMED( "clear_int", clear( &si ); )
1628        REPEAT_TIMED( "pop_int", N,
1629                int xi = pop( &ti ); if ( xi > maxi ) { maxi = xi; } )
1630        REPEAT_TIMED( "print_int", N/2, print( out, vali, ":", vali, "\n" ); )
1631
1632        pair(_Bool, char) maxp = { (_Bool)0, '\0' }, valp = { (_Bool)1, 'a' };
1633        stack(pair(_Bool, char)) sp, tp;
1634
1635        REPEAT_TIMED( "push_pair", N, push( &sp, valp ); )
1636        TIMED( "copy_pair", tp = sp; )
1637        TIMED( "clear_pair", clear( &sp ); )
1638        REPEAT_TIMED( "pop_pair", N,
1639                pair(_Bool, char) xp = pop( &tp ); if ( xp > maxp ) { maxp = xp; } )
1640        REPEAT_TIMED( "print_pair", N/2, print( out, valp, ":", valp, "\n" ); )
1641        fclose(out);
1642}
1643\end{lstlisting}
1644\caption{\protect\CFA Benchmark Test}
1645\label{fig:BenchmarkTest}
1646\end{figure}
1647
1648The structure of each benchmark implemented is: C with @void *@-based polymorphism, \CFA with the presented features, \CC with templates, and \CC using only class inheritance for polymorphism, called \CCV.
1649The \CCV variant illustrates an alternative object-oriented idiom where all objects inherit from a base @object@ class, mimicking a Java-like interface;
1650hence runtime checks are necessary to safely down-cast objects.
1651The most notable difference among the implementations is in memory layout of generic types: \CFA and \CC inline the stack and pair elements into corresponding list and pair nodes, while C and \CCV lack such a capability and instead must store generic objects via pointers to separately-allocated objects.
1652For the print benchmark, idiomatic printing is used: the C and \CFA variants used @stdio.h@, while the \CC and \CCV variants used @iostream@; preliminary tests show this distinction has negligible runtime impact.
1653Note, the C benchmark uses unchecked casts as there is no runtime mechanism to perform such checks, while \CFA and \CC provide type-safety statically.
1654
1655Figure~\ref{fig:eval} and Table~\ref{tab:eval} show the results of running the benchmark in Figure~\ref{fig:BenchmarkTest} and its C, \CC, and \CCV equivalents.
1656The graph plots the median of 5 consecutive runs of each program, with an initial warm-up run omitted.
1657All code is compiled at \texttt{-O2} by GCC or G++ 6.2.0, with all \CC code compiled as \CCfourteen.
1658The benchmarks are run on an Ubuntu 16.04 workstation with 16 GB of RAM and a 6-core AMD FX-6300 CPU with 3.5 GHz maximum clock frequency.
1659
1660\begin{figure}
1661\centering
1662\input{timing}
1663\caption{Benchmark Timing Results (smaller is better)}
1664\label{fig:eval}
1665\end{figure}
1666
1667\begin{table}
1668\caption{Properties of benchmark code}
1669\label{tab:eval}
1670\newcommand{\CT}[1]{\multicolumn{1}{c}{#1}}
1671\begin{tabular}{rrrrr}
1672                                                                        & \CT{C}        & \CT{\CFA}     & \CT{\CC}      & \CT{\CCV}             \\ \hline
1673maximum memory usage (MB)                       & 10001         & 2502          & 2503          & 11253                 \\
1674source code size (lines)                        & 247           & 222           & 165           & 339                   \\
1675redundant type annotations (lines)      & 39            & 2                     & 2                     & 15                    \\
1676binary size (KB)                                        & 14            & 229           & 18            & 38                    \\
1677\end{tabular}
1678\end{table}
1679
1680The C and \CCV variants are generally the slowest with the largest memory footprint, because of their less-efficient memory layout and the pointer-indirection necessary to implement generic types;
1681this inefficiency is exacerbated by the second level of generic types in the pair-based benchmarks.
1682By contrast, the \CFA and \CC variants run in roughly equivalent time for both the integer and pair of @_Bool@ and @char@ because the storage layout is equivalent, with the inlined libraries (\ie no separate compilation) and greater maturity of the \CC compiler contributing to its lead.
1683\CCV is slower than C largely due to the cost of runtime type-checking of down-casts (implemented with @dynamic_cast@);
1684There are two outliers in the graph for \CFA: all prints and pop of @pair@.
1685Both of these cases result from the complexity of the C-generated polymorphic code, so that the GCC compiler is unable to optimize some dead code and condense nested calls.
1686A compiler designed for \CFA could easily perform these optimizations.
1687Finally, the binary size for \CFA is larger because of static linking with the \CFA libraries.
1688
1689\CFA is also competitive in terms of source code size, measured as a proxy for programmer effort. The line counts in Table~\ref{tab:eval} include implementations of @pair@ and @stack@ types for all four languages for purposes of direct comparison, though it should be noted that \CFA and \CC have pre-written data structures in their standard libraries that programmers would generally use instead. Use of these standard library types has minimal impact on the performance benchmarks, but shrinks the \CFA and \CC benchmarks to 73 and 54 lines, respectively.
1690On the other hand, C does not have a generic collections-library in its standard distribution, resulting in frequent reimplementation of such collection types by C programmers.
1691\CCV does not use the \CC standard template library by construction, and in fact includes the definition of @object@ and wrapper classes for @bool@, @char@, @int@, and @const char *@ in its line count, which inflates this count somewhat, as an actual object-oriented language would include these in the standard library;
1692with their omission, the \CCV line count is similar to C.
1693We justify the given line count by noting that many object-oriented languages do not allow implementing new interfaces on library types without subclassing or wrapper types, which may be similarly verbose.
1694
1695Raw line-count, however, is a fairly rough measure of code complexity;
1696another important factor is how much type information the programmer must manually specify, especially where that information is not checked by the compiler.
1697Such unchecked type information produces a heavier documentation burden and increased potential for runtime bugs, and is much less common in \CFA than C, with its manually specified function pointers arguments and format codes, or \CCV, with its extensive use of un-type-checked downcasts (\eg @object@ to @integer@ when popping a stack, or @object@ to @printable@ when printing the elements of a @pair@).
1698To quantify this, the ``redundant type annotations'' line in Table~\ref{tab:eval} counts the number of lines on which the type of a known variable is re-specified, either as a format specifier, explicit downcast, type-specific function, or by name in a @sizeof@, struct literal, or @new@ expression.
1699The \CC benchmark uses two redundant type annotations to create a new stack nodes, while the C and \CCV benchmarks have several such annotations spread throughout their code.
1700The two instances in which the \CFA benchmark still uses redundant type specifiers are to cast the result of a polymorphic @malloc@ call (the @sizeof@ argument is inferred by the compiler).
1701These uses are similar to the @new@ expressions in \CC, though the \CFA compiler's type resolver should shortly render even these type casts superfluous.
1702
1703
1704\section{Related Work}
1705
1706
1707\subsection{Polymorphism}
1708
1709\CC is the most similar language to \CFA;
1710both are extensions to C with source and runtime backwards compatibility.
1711The fundamental difference is in their engineering approach to C compatibility and programmer expectation.
1712While \CC provides good backwards compatibility with C, it has a steep learning curve for many of its extensions.
1713For example, polymorphism is provided via three disjoint mechanisms: overloading, inheritance, and templates.
1714The overloading is restricted because resolution does not use the return type, inheritance requires learning object-oriented programming and coping with a restricted nominal-inheritance hierarchy, templates cannot be separately compiled resulting in compilation/code bloat and poor error messages, and determining how these mechanisms interact and which to use is confusing.
1715In contrast, \CFA has a single facility for polymorphic code supporting type-safe separate-compilation of polymorphic functions and generic (opaque) types, which uniformly leverage the C procedural paradigm.
1716The key mechanism to support separate compilation is \CFA's \emph{explicit} use of assumed properties for a type.
1717Until \CC concepts~\cite{C++Concepts} are standardized (anticipated for \CCtwenty), \CC provides no way to specify the requirements of a generic function in code beyond compilation errors during template expansion;
1718furthermore, \CC concepts are restricted to template polymorphism.
1719
1720Cyclone~\cite{Grossman06} also provides capabilities for polymorphic functions and existential types, similar to \CFA's @forall@ functions and generic types.
1721Cyclone existential types can include function pointers in a construct similar to a virtual function-table, but these pointers must be explicitly initialized at some point in the code, a tedious and potentially error-prone process.
1722Furthermore, Cyclone's polymorphic functions and types are restricted to abstraction over types with the same layout and calling convention as @void *@, \ie only pointer types and @int@.
1723In \CFA terms, all Cyclone polymorphism must be dtype-static.
1724While the Cyclone design provides the efficiency benefits discussed in Section~\ref{sec:generic-apps} for dtype-static polymorphism, it is more restrictive than \CFA's general model.
1725Smith and Volpano~\cite{Smith98} present Polymorphic C, an ML dialect with polymorphic functions, C-like syntax, and pointer types; it lacks many of C's features, however, most notably structure types, and so is not a practical C replacement.
1726
1727Objective-C~\cite{obj-c-book} is an industrially successful extension to C.
1728However, Objective-C is a radical departure from C, using an object-oriented model with message-passing.
1729Objective-C did not support type-checked generics until recently \cite{xcode7}, historically using less-efficient runtime checking of object types.
1730The GObject~\cite{GObject} framework also adds object-oriented programming with runtime type-checking and reference-counting garbage-collection to C;
1731these features are more intrusive additions than those provided by \CFA, in addition to the runtime overhead of reference-counting.
1732Vala~\cite{Vala} compiles to GObject-based C, adding the burden of learning a separate language syntax to the aforementioned demerits of GObject as a modernization path for existing C code-bases.
1733Java~\cite{Java8} included generic types in Java~5, which are type-checked at compilation and type-erased at runtime, similar to \CFA's.
1734However, in Java, each object carries its own table of method pointers, while \CFA passes the method pointers separately to maintain a C-compatible layout.
1735Java is also a garbage-collected, object-oriented language, with the associated resource usage and C-interoperability burdens.
1736
1737D~\cite{D}, Go, and Rust~\cite{Rust} are modern, compiled languages with abstraction features similar to \CFA traits, \emph{interfaces} in D and Go and \emph{traits} in Rust.
1738However, each language represents a significant departure from C in terms of language model, and none has the same level of compatibility with C as \CFA.
1739D and Go are garbage-collected languages, imposing the associated runtime overhead.
1740The necessity of accounting for data transfer between managed runtimes and the unmanaged C runtime complicates foreign-function interfaces to C.
1741Furthermore, while generic types and functions are available in Go, they are limited to a small fixed set provided by the compiler, with no language facility to define more.
1742D restricts garbage collection to its own heap by default, while Rust is not garbage-collected, and thus has a lighter-weight runtime more interoperable with C.
1743Rust also possesses much more powerful abstraction capabilities for writing generic code than Go.
1744On the other hand, Rust's borrow-checker provides strong safety guarantees but is complex and difficult to learn and imposes a distinctly idiomatic programming style.
1745\CFA, with its more modest safety features, allows direct ports of C code while maintaining the idiomatic style of the original source.
1746
1747
1748\subsection{Tuples/Variadics}
1749
1750Many programming languages have some form of tuple construct and/or variadic functions, \eg SETL, C, KW-C, \CC, D, Go, Java, ML, and Scala.
1751SETL~\cite{SETL} is a high-level mathematical programming language, with tuples being one of the primary data types.
1752Tuples in SETL allow subscripting, dynamic expansion, and multiple assignment.
1753C provides variadic functions through @va_list@ objects, but the programmer is responsible for managing the number of arguments and their types, so the mechanism is type unsafe.
1754KW-C~\cite{Buhr94a}, a predecessor of \CFA, introduced tuples to C as an extension of the C syntax, taking much of its inspiration from SETL.
1755The main contributions of that work were adding MRVF, tuple mass and multiple assignment, and record-field access.
1756\CCeleven introduced @std::tuple@ as a library variadic template structure.
1757Tuples are a generalization of @std::pair@, in that they allow for arbitrary length, fixed-size aggregation of heterogeneous values.
1758Operations include @std::get<N>@ to extract values, @std::tie@ to create a tuple of references used for assignment, and lexicographic comparisons.
1759\CCseventeen proposes \emph{structured bindings}~\cite{Sutter15} to eliminate pre-declaring variables and use of @std::tie@ for binding the results.
1760This extension requires the use of @auto@ to infer the types of the new variables, so complicated expressions with a non-obvious type must be documented with some other mechanism.
1761Furthermore, structured bindings are not a full replacement for @std::tie@, as it always declares new variables.
1762Like \CC, D provides tuples through a library variadic-template structure.
1763Go does not have tuples but supports MRVF.
1764Java's variadic functions appear similar to C's but are type-safe using homogeneous arrays, which are less useful than \CFA's heterogeneously-typed variadic functions.
1765Tuples are a fundamental abstraction in most functional programming languages, such as Standard ML~\cite{sml} and~\cite{Scala}, which decompose tuples using pattern matching.
1766
1767
1768\section{Conclusion and Future Work}
1769
1770The goal of \CFA is to provide an evolutionary pathway for large C development-environments to be more productive and safer, while respecting the talent and skill of C programmers.
1771While other programming languages purport to be a better C, they are in fact new and interesting languages in their own right, but not C extensions.
1772The purpose of this paper is to introduce \CFA, and showcase language features that illustrate the \CFA type-system and approaches taken to achieve the goal of evolutionary C extension.
1773The contributions are a powerful type-system using parametric polymorphism and overloading, generic types, and tuples, which all have complex interactions.
1774The work is a challenging design, engineering, and implementation exercise.
1775On the surface, the project may appear as a rehash of similar mechanisms in \CC.
1776However, every \CFA feature is different than its \CC counterpart, often with extended functionality, better integration with C and its programmers, and always supporting separate compilation.
1777All of these new features are being used by the \CFA development-team to build the \CFA runtime-system.
1778Finally, we demonstrate that \CFA performance for some idiomatic cases is better than C and close to \CC, showing the design is practically applicable.
1779
1780There is ongoing work on a wide range of \CFA feature extensions, including reference types, arrays with size, exceptions, concurrent primitives and modules.
1781(While all examples in the paper compile and run, a public beta-release of \CFA will take another 8--12 months to finalize these additional extensions.)
1782In addition, there are interesting future directions for the polymorphism design.
1783Notably, \CC template functions trade compile time and code bloat for optimal runtime of individual instantiations of polymorphic functions.
1784\CFA polymorphic functions use dynamic virtual-dispatch;
1785the runtime overhead of this approach is low, but not as low as inlining, and it may be beneficial to provide a mechanism for performance-sensitive code.
1786Two promising approaches are an @inline@ annotation at polymorphic function call sites to create a template-specialization of the function (provided the code is visible) or placing an @inline@ annotation on polymorphic function-definitions to instantiate a specialized version for some set of types (\CC template specialization).
1787These approaches are not mutually exclusive and allow performance optimizations to be applied only when necessary, without suffering global code-bloat.
1788In general, we believe separate compilation, producing smaller code, works well with loaded hardware-caches, which may offset the benefit of larger inlined-code.
1789
1790
1791\section{Acknowledgments}
1792
1793The authors would like to recognize the design assistance of Glen Ditchfield, Richard Bilson, and Thierry Delisle on the features described in this paper, and thank Magnus Madsen and the three anonymous reviewers for valuable feedback.
1794%This work is supported in part by a corporate partnership with \grantsponsor{Huawei}{Huawei Ltd.}{http://www.huawei.com}, and Aaron Moss and Peter Buhr are funded by the \grantsponsor{Natural Sciences and Engineering Research Council} of Canada.
1795% the first author's \grantsponsor{NSERC-PGS}{NSERC PGS D}{http://www.nserc-crsng.gc.ca/Students-Etudiants/PG-CS/BellandPostgrad-BelletSuperieures_eng.asp} scholarship.
1796
1797
1798\bibliographystyle{plain}
1799\bibliography{pl}
1800
1801
1802\appendix
1803
1804\section{Benchmark Stack Implementation}
1805\label{sec:BenchmarkStackImplementation}
1806
1807\lstset{basicstyle=\linespread{0.9}\sf\small}
1808
1809Throughout, @/***/@ designates a counted redundant type annotation.
1810
1811\smallskip\noindent
1812\CFA
1813\begin{lstlisting}[xleftmargin=2\parindentlnth,aboveskip=0pt,belowskip=0pt]
1814forall(otype T) struct stack_node {
1815        T value;
1816        stack_node(T) * next;
1817};
1818forall(otype T) void ?{}(stack(T) * s) { (&s->head){ 0 }; }
1819forall(otype T) void ?{}(stack(T) * s, stack(T) t) {
1820        stack_node(T) ** crnt = &s->head;
1821        for ( stack_node(T) * next = t.head; next; next = next->next ) {
1822                *crnt = ((stack_node(T) *)malloc()){ next->value }; /***/
1823                stack_node(T) * acrnt = *crnt;
1824                crnt = &acrnt->next;
1825        }
1826        *crnt = 0;
1827}
1828forall(otype T) stack(T) ?=?(stack(T) * s, stack(T) t) {
1829        if ( s->head == t.head ) return *s;
1830        clear(s);
1831        s{ t };
1832        return *s;
1833}
1834forall(otype T) void ^?{}(stack(T) * s) { clear(s); }
1835forall(otype T) _Bool empty(const stack(T) * s) { return s->head == 0; }
1836forall(otype T) void push(stack(T) * s, T value) {
1837        s->head = ((stack_node(T) *)malloc()){ value, s->head }; /***/
1838}
1839forall(otype T) T pop(stack(T) * s) {
1840        stack_node(T) * n = s->head;
1841        s->head = n->next;
1842        T x = n->value;
1843        ^n{};
1844        free(n);
1845        return x;
1846}
1847forall(otype T) void clear(stack(T) * s) {
1848        for ( stack_node(T) * next = s->head; next; ) {
1849                stack_node(T) * crnt = next;
1850                next = crnt->next;
1851                delete(crnt);
1852        }
1853        s->head = 0;
1854}
1855\end{lstlisting}
1856
1857\medskip\noindent
1858\CC
1859\begin{lstlisting}[xleftmargin=2\parindentlnth,aboveskip=0pt,belowskip=0pt]
1860template<typename T> class stack {
1861        struct node {
1862                T value;
1863                node * next;
1864                node( const T & v, node * n = nullptr ) : value(v), next(n) {}
1865        };
1866        node * head;
1867        void copy(const stack<T>& o) {
1868                node ** crnt = &head;
1869                for ( node * next = o.head;; next; next = next->next ) {
1870                        *crnt = new node{ next->value }; /***/
1871                        crnt = &(*crnt)->next;
1872                }
1873                *crnt = nullptr;
1874        }
1875  public:
1876        stack() : head(nullptr) {}
1877        stack(const stack<T>& o) { copy(o); }
1878        stack(stack<T> && o) : head(o.head) { o.head = nullptr; }
1879        ~stack() { clear(); }
1880        stack & operator= (const stack<T>& o) {
1881                if ( this == &o ) return *this;
1882                clear();
1883                copy(o);
1884                return *this;
1885        }
1886        stack & operator= (stack<T> && o) {
1887                if ( this == &o ) return *this;
1888                head = o.head;
1889                o.head = nullptr;
1890                return *this;
1891        }
1892        bool empty() const { return head == nullptr; }
1893        void push(const T & value) { head = new node{ value, head };  /***/ }
1894        T pop() {
1895                node * n = head;
1896                head = n->next;
1897                T x = std::move(n->value);
1898                delete n;
1899                return x;
1900        }
1901        void clear() {
1902                for ( node * next = head; next; ) {
1903                        node * crnt = next;
1904                        next = crnt->next;
1905                        delete crnt;
1906                }
1907                head = nullptr;
1908        }
1909};
1910\end{lstlisting}
1911
1912\medskip\noindent
1913C
1914\begin{lstlisting}[xleftmargin=2\parindentlnth,aboveskip=0pt,belowskip=0pt]
1915struct stack_node {
1916        void * value;
1917        struct stack_node * next;
1918};
1919struct stack new_stack() { return (struct stack){ NULL }; /***/ }
1920void copy_stack(struct stack * s, const struct stack * t, void * (*copy)(const void *)) {
1921        struct stack_node ** crnt = &s->head;
1922        for ( struct stack_node * next = t->head; next; next = next->next ) {
1923                *crnt = malloc(sizeof(struct stack_node)); /***/
1924                **crnt = (struct stack_node){ copy(next->value) }; /***/
1925                crnt = &(*crnt)->next;
1926        }
1927        *crnt = 0;
1928}
1929_Bool stack_empty(const struct stack * s) { return s->head == NULL; }
1930void push_stack(struct stack * s, void * value) {
1931        struct stack_node * n = malloc(sizeof(struct stack_node)); /***/
1932        *n = (struct stack_node){ value, s->head }; /***/
1933        s->head = n;
1934}
1935void * pop_stack(struct stack * s) {
1936        struct stack_node * n = s->head;
1937        s->head = n->next;
1938        void * x = n->value;
1939        free(n);
1940        return x;
1941}
1942void clear_stack(struct stack * s, void (*free_el)(void *)) {
1943        for ( struct stack_node * next = s->head; next; ) {
1944                struct stack_node * crnt = next;
1945                next = crnt->next;
1946                free_el(crnt->value);
1947                free(crnt);
1948        }
1949        s->head = NULL;
1950}
1951\end{lstlisting}
1952
1953\medskip\noindent
1954\CCV
1955\begin{lstlisting}[xleftmargin=2\parindentlnth,aboveskip=0pt,belowskip=0pt]
1956stack::node::node( const object & v, node * n ) : value( v.new_copy() ), next( n ) {}
1957void stack::copy(const stack & o) {
1958        node ** crnt = &head;
1959        for ( node * next = o.head; next; next = next->next ) {
1960                *crnt = new node{ *next->value };
1961                crnt = &(*crnt)->next;
1962        }
1963        *crnt = nullptr;
1964}
1965stack::stack() : head(nullptr) {}
1966stack::stack(const stack & o) { copy(o); }
1967stack::stack(stack && o) : head(o.head) { o.head = nullptr; }
1968stack::~stack() { clear(); }
1969stack & stack::operator= (const stack & o) {
1970        if ( this == &o ) return *this;
1971        clear();
1972        copy(o);
1973        return *this;
1974}
1975stack & stack::operator= (stack && o) {
1976        if ( this == &o ) return *this;
1977        head = o.head;
1978        o.head = nullptr;
1979        return *this;
1980}
1981bool stack::empty() const { return head == nullptr; }
1982void stack::push(const object & value) { head = new node{ value, head }; /***/ }
1983ptr<object> stack::pop() {
1984        node * n = head;
1985        head = n->next;
1986        ptr<object> x = std::move(n->value);
1987        delete n;
1988        return x;
1989}
1990void stack::clear() {
1991        for ( node * next = head; next; ) {
1992                node * crnt = next;
1993                next = crnt->next;
1994                delete crnt;
1995        }
1996        head = nullptr;
1997}
1998\end{lstlisting}
1999
2000
2001\begin{comment}
2002
2003\subsubsection{bench.h}
2004(\texttt{bench.hpp} is similar.)
2005
2006\lstinputlisting{evaluation/bench.h}
2007
2008\subsection{C}
2009
2010\subsubsection{c-stack.h} ~
2011
2012\lstinputlisting{evaluation/c-stack.h}
2013
2014\subsubsection{c-stack.c} ~
2015
2016\lstinputlisting{evaluation/c-stack.c}
2017
2018\subsubsection{c-pair.h} ~
2019
2020\lstinputlisting{evaluation/c-pair.h}
2021
2022\subsubsection{c-pair.c} ~
2023
2024\lstinputlisting{evaluation/c-pair.c}
2025
2026\subsubsection{c-print.h} ~
2027
2028\lstinputlisting{evaluation/c-print.h}
2029
2030\subsubsection{c-print.c} ~
2031
2032\lstinputlisting{evaluation/c-print.c}
2033
2034\subsubsection{c-bench.c} ~
2035
2036\lstinputlisting{evaluation/c-bench.c}
2037
2038\subsection{\CFA}
2039
2040\subsubsection{cfa-stack.h} ~
2041
2042\lstinputlisting{evaluation/cfa-stack.h}
2043
2044\subsubsection{cfa-stack.c} ~
2045
2046\lstinputlisting{evaluation/cfa-stack.c}
2047
2048\subsubsection{cfa-print.h} ~
2049
2050\lstinputlisting{evaluation/cfa-print.h}
2051
2052\subsubsection{cfa-print.c} ~
2053
2054\lstinputlisting{evaluation/cfa-print.c}
2055
2056\subsubsection{cfa-bench.c} ~
2057
2058\lstinputlisting{evaluation/cfa-bench.c}
2059
2060\subsection{\CC}
2061
2062\subsubsection{cpp-stack.hpp} ~
2063
2064\lstinputlisting[language=c++]{evaluation/cpp-stack.hpp}
2065
2066\subsubsection{cpp-print.hpp} ~
2067
2068\lstinputlisting[language=c++]{evaluation/cpp-print.hpp}
2069
2070\subsubsection{cpp-bench.cpp} ~
2071
2072\lstinputlisting[language=c++]{evaluation/cpp-bench.cpp}
2073
2074\subsection{\CCV}
2075
2076\subsubsection{object.hpp} ~
2077
2078\lstinputlisting[language=c++]{evaluation/object.hpp}
2079
2080\subsubsection{cpp-vstack.hpp} ~
2081
2082\lstinputlisting[language=c++]{evaluation/cpp-vstack.hpp}
2083
2084\subsubsection{cpp-vstack.cpp} ~
2085
2086\lstinputlisting[language=c++]{evaluation/cpp-vstack.cpp}
2087
2088\subsubsection{cpp-vprint.hpp} ~
2089
2090\lstinputlisting[language=c++]{evaluation/cpp-vprint.hpp}
2091
2092\subsubsection{cpp-vbench.cpp} ~
2093
2094\lstinputlisting[language=c++]{evaluation/cpp-vbench.cpp}
2095\end{comment}
2096
2097\end{document}
2098
2099% Local Variables: %
2100% tab-width: 4 %
2101% compile-command: "make" %
2102% End: %
Note: See TracBrowser for help on using the repository browser.