source: doc/papers/general/Paper.tex @ 6dba9f99

aaron-thesisarm-ehcleanup-dtorsdeferred_resndemanglerjacob/cs343-translationjenkins-sandboxnew-astnew-ast-unique-exprnew-envno_listpersistent-indexerresolv-newwith_gc
Last change on this file since 6dba9f99 was 6dba9f99, checked in by Aaron Moss <a3moss@…>, 4 years ago

Update dynamic generic type examples

  • Property mode set to 100644
File size: 152.3 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{enumitem}
9\setlist[itemize]{topsep=3pt,itemsep=2pt,parsep=0pt}% global
10\usepackage[flushmargin]{footmisc}                      % support label/reference in footnote
11\usepackage{rotating}
12\usepackage[usenames]{color}
13\usepackage{pslatex}                                            % reduce size of san serif font
14\usepackage[plainpages=false,pdfpagelabels,pdfpagemode=UseNone,pagebackref=true,breaklinks=true,colorlinks=true,linkcolor=blue,citecolor=blue,urlcolor=blue]{hyperref}
15\urlstyle{sf}
16\usepackage{breakurl}
17
18\setlength{\textheight}{9in}
19%\oddsidemargin 0.0in
20\renewcommand{\topfraction}{0.8}                        % float must be greater than X of the page before it is forced onto its own page
21\renewcommand{\bottomfraction}{0.8}                     % float must be greater than X of the page before it is forced onto its own page
22\renewcommand{\floatpagefraction}{0.8}          % float must be greater than X of the page before it is forced onto its own page
23\renewcommand{\textfraction}{0.0}                       % the entire page maybe devoted to floats with no text on the page at all
24
25\lefthyphenmin=4                                                        % hyphen only after 4 characters
26\righthyphenmin=4
27
28% Names used in the document.
29
30\newcommand{\CFAIcon}{\textsf{C}\raisebox{\depth}{\rotatebox{180}{\textsf{A}}}\xspace} % Cforall symbolic name
31\newcommand{\CFA}{\protect\CFAIcon}             % safe for section/caption
32\newcommand{\CFL}{\textrm{Cforall}\xspace}      % Cforall symbolic name
33\newcommand{\Celeven}{\textrm{C11}\xspace}      % C11 symbolic name
34\newcommand{\CC}{\textrm{C}\kern-.1em\hbox{+\kern-.25em+}\xspace} % C++ symbolic name
35\newcommand{\CCeleven}{\textrm{C}\kern-.1em\hbox{+\kern-.25em+}11\xspace} % C++11 symbolic name
36\newcommand{\CCfourteen}{\textrm{C}\kern-.1em\hbox{+\kern-.25em+}14\xspace} % C++14 symbolic name
37\newcommand{\CCseventeen}{\textrm{C}\kern-.1em\hbox{+\kern-.25em+}17\xspace} % C++17 symbolic name
38\newcommand{\CCtwenty}{\textrm{C}\kern-.1em\hbox{+\kern-.25em+}20\xspace} % C++20 symbolic name
39\newcommand{\CCV}{\rm C\kern-.1em\hbox{+\kern-.25em+}obj\xspace} % C++ virtual symbolic name
40\newcommand{\Csharp}{C\raisebox{-0.7ex}{\Large$^\sharp$}\xspace} % C# symbolic name
41
42%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
43
44\newcommand{\Textbf}[2][red]{{\color{#1}{\textbf{#2}}}}
45%\newcommand{\TODO}[1]{\textbf{TODO}: {\itshape #1}} % TODO included
46\newcommand{\TODO}[1]{} % TODO elided
47
48% Default underscore is too low and wide. Cannot use lstlisting "literate" as replacing underscore
49% removes it as a variable-name character so keywords in variables are highlighted. MUST APPEAR
50% AFTER HYPERREF.
51%\DeclareTextCommandDefault{\textunderscore}{\leavevmode\makebox[1.2ex][c]{\rule{1ex}{0.1ex}}}
52\renewcommand{\textunderscore}{\leavevmode\makebox[1.2ex][c]{\rule{1ex}{0.075ex}}}
53
54\makeatletter
55% parindent is relative, i.e., toggled on/off in environments like itemize, so store the value for
56% use rather than use \parident directly.
57\newlength{\parindentlnth}
58\setlength{\parindentlnth}{\parindent}
59
60\newcommand{\LstBasicStyle}[1]{{\lst@basicstyle{\lst@basicstyle{#1}}}}
61\newcommand{\LstKeywordStyle}[1]{{\lst@basicstyle{\lst@keywordstyle{#1}}}}
62\newcommand{\LstCommentStyle}[1]{{\lst@basicstyle{\lst@commentstyle{#1}}}}
63
64\newlength{\gcolumnposn}                                        % temporary hack because lstlisting does not handle tabs correctly
65\newlength{\columnposn}
66\setlength{\gcolumnposn}{2.75in}
67\setlength{\columnposn}{\gcolumnposn}
68\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}}}}
69\newcommand{\CRT}{\global\columnposn=\gcolumnposn}
70
71% Denote newterms in particular font and index them without particular font and in lowercase, e.g., \newterm{abc}.
72% The option parameter provides an index term different from the new term, e.g., \newterm[\texttt{abc}]{abc}
73% The star version does not lowercase the index information, e.g., \newterm*{IBM}.
74\newcommand{\newtermFontInline}{\emph}
75\newcommand{\newterm}{\@ifstar\@snewterm\@newterm}
76\newcommand{\@newterm}[2][\@empty]{\lowercase{\def\temp{#2}}{\newtermFontInline{#2}}\ifx#1\@empty\index{\temp}\else\index{#1@{\protect#2}}\fi}
77\newcommand{\@snewterm}[2][\@empty]{{\newtermFontInline{#2}}\ifx#1\@empty\index{#2}\else\index{#1@{\protect#2}}\fi}
78
79% Latin abbreviation
80\newcommand{\abbrevFont}{\textit}                       % set empty for no italics
81\newcommand{\EG}{\abbrevFont{e}.\abbrevFont{g}.}
82\newcommand*{\eg}{%
83        \@ifnextchar{,}{\EG}%
84                {\@ifnextchar{:}{\EG}%
85                        {\EG,\xspace}}%
86}%
87\newcommand{\IE}{\abbrevFont{i}.\abbrevFont{e}.}
88\newcommand*{\ie}{%
89        \@ifnextchar{,}{\IE}%
90                {\@ifnextchar{:}{\IE}%
91                        {\IE,\xspace}}%
92}%
93\newcommand{\ETC}{\abbrevFont{etc}}
94\newcommand*{\etc}{%
95        \@ifnextchar{.}{\ETC}%
96        {\ETC.\xspace}%
97}%
98\newcommand{\ETAL}{\abbrevFont{et}~\abbrevFont{al}}
99\newcommand*{\etal}{%
100        \@ifnextchar{.}{\protect\ETAL}%
101                {\protect\ETAL.\xspace}%
102}%
103\newcommand{\VIZ}{\abbrevFont{viz}}
104\newcommand*{\viz}{%
105        \@ifnextchar{.}{\VIZ}%
106                {\VIZ.\xspace}%
107}%
108\makeatother
109
110\newenvironment{cquote}{%
111        \list{}{\lstset{resetmargins=true,aboveskip=0pt,belowskip=0pt}\topsep=3pt\parsep=0pt\leftmargin=\parindentlnth\rightmargin\leftmargin}%
112        \item\relax
113}{%
114        \endlist
115}% cquote
116
117% CFA programming language, based on ANSI C (with some gcc additions)
118\lstdefinelanguage{CFA}[ANSI]{C}{
119        morekeywords={
120                _Alignas, _Alignof, __alignof, __alignof__, asm, __asm, __asm__, _At, __attribute,
121                __attribute__, auto, _Bool, catch, catchResume, choose, _Complex, __complex, __complex__,
122                __const, __const__, disable, dtype, enable, exception, __extension__, fallthrough, fallthru,
123                finally, forall, ftype, _Generic, _Imaginary, inline, __label__, lvalue, _Noreturn, one_t,
124                otype, restrict, _Static_assert, throw, throwResume, trait, try, ttype, typeof, __typeof,
125                __typeof__, virtual, with, zero_t},
126        morekeywords=[2]{
127                _Atomic, coroutine, is_coroutine, is_monitor, is_thread, monitor, mutex, nomutex, or,
128                resume, suspend, thread, _Thread_local, waitfor, when, yield},
129        moredirectives={defined,include_next}%
130}
131
132\lstset{
133language=CFA,
134columns=fullflexible,
135basicstyle=\linespread{0.9}\sf,                                                 % reduce line spacing and use sanserif font
136stringstyle=\tt,                                                                                % use typewriter font
137tabsize=5,                                                                                              % N space tabbing
138xleftmargin=\parindentlnth,                                                             % indent code to paragraph indentation
139%mathescape=true,                                                                               % LaTeX math escape in CFA code $...$
140escapechar=\$,                                                                                  % LaTeX escape in CFA code
141keepspaces=true,                                                                                %
142showstringspaces=false,                                                                 % do not show spaces with cup
143showlines=true,                                                                                 % show blank lines at end of code
144aboveskip=4pt,                                                                                  % spacing above/below code block
145belowskip=3pt,
146% replace/adjust listing characters that look bad in sanserif
147literate={-}{\makebox[1ex][c]{\raisebox{0.4ex}{\rule{0.8ex}{0.1ex}}}}1 {^}{\raisebox{0.6ex}{$\scriptscriptstyle\land\,$}}1
148        {~}{\raisebox{0.3ex}{$\scriptstyle\sim\,$}}1 {@}{\small{@}}1 % {`}{\ttfamily\upshape\hspace*{-0.1ex}`}1
149        {<-}{$\leftarrow$}2 {=>}{$\Rightarrow$}2 {->}{\makebox[1ex][c]{\raisebox{0.4ex}{\rule{0.8ex}{0.075ex}}}\kern-0.2ex\textgreater}2,
150moredelim=**[is][\color{red}]{`}{`},
151}% lstset
152
153% inline code @...@
154\lstMakeShortInline@%
155
156\lstnewenvironment{cfa}[1][]
157{\lstset{#1}}
158{}
159\lstnewenvironment{C++}[1][]                            % use C++ style
160{\lstset{language=C++,moredelim=**[is][\protect\color{red}]{`}{`},#1}\lstset{#1}}
161{}
162
163
164\title{\protect\CFA : Adding Modern Programming Language Features to C}
165
166\author{Aaron Moss, Robert Schluntz, Peter Buhr}
167% \email{a3moss@uwaterloo.ca}
168% \email{rschlunt@uwaterloo.ca}
169% \email{pabuhr@uwaterloo.ca}
170% \affiliation{%
171%       \institution{University of Waterloo}
172%       \department{David R. Cheriton School of Computer Science}
173%       \streetaddress{Davis Centre, University of Waterloo}
174%       \city{Waterloo}
175%       \state{ON}
176%       \postcode{N2L 3G1}
177%       \country{Canada}
178% }
179
180%\terms{generic, tuple, variadic, types}
181%\keywords{generic types, tuple types, variadic types, polymorphic functions, C, Cforall}
182
183\begin{document}
184\maketitle
185
186
187\begin{abstract}
188The 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.
189This installation base and the programmers producing it represent a massive software-engineering investment spanning decades and likely to continue for decades more.
190Nevertheless, C, first standardized over thirty years ago, lacks many features that make programming in more modern languages safer and more productive.
191
192The 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.
193Prior 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.
194Specifically, \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 programmers.
195This paper presents a quick tour of \CFA features showing how their design avoids shortcomings of similar features in C and other C-like languages.
196Finally, experimental results are presented to validate several of the new features.
197\end{abstract}
198
199
200\section{Introduction}
201
202The 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.
203This installation base and the programmers producing it represent a massive software-engineering investment spanning decades and likely to continue for decades more.
204The TIOBE~\cite{TIOBE} ranks the top 5 most \emph{popular} programming languages as: Java 15\%, \Textbf{C 12\%}, \Textbf{\CC 5.5\%}, Python 5\%, \Csharp 4.5\% = 42\%, where the next 50 languages are less than 4\% each with a long tail.
205The top 3 rankings over the past 30 years are:
206\begin{center}
207\setlength{\tabcolsep}{10pt}
208\lstDeleteShortInline@%
209\begin{tabular}{@{}rccccccc@{}}
210                & 2018  & 2013  & 2008  & 2003  & 1998  & 1993  & 1988  \\ \hline
211Java    & 1             & 2             & 1             & 1             & 18    & -             & -             \\
212\Textbf{C}& \Textbf{2} & \Textbf{1} & \Textbf{2} & \Textbf{2} & \Textbf{1} & \Textbf{1} & \Textbf{1} \\
213\CC             & 3             & 4             & 3             & 3             & 2             & 2             & 5             \\
214\end{tabular}
215\lstMakeShortInline@%
216\end{center}
217Love it or hate it, C is extremely popular, highly used, and one of the few systems languages.
218In many cases, \CC is often used solely as a better C.
219Nonetheless, C, first standardized over thirty years ago, lacks many features that make programming in more modern languages safer and more productive.
220
221\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.
222The four key design goals for \CFA~\cite{Bilson03} are:
223(1) The behaviour of standard C code must remain the same when translated by a \CFA compiler as when translated by a C compiler;
224(2) Standard C code must be as fast and as small when translated by a \CFA compiler as when translated by a C compiler;
225(3) \CFA code must be at least as portable as standard C code;
226(4) Extensions introduced by \CFA must be translated in the most efficient way possible.
227These 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.
228\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.
229
230\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).
231Ultimately, a compiler is necessary for advanced features and optimal performance.
232All of the features discussed in this paper are working, unless a feature states it is a future feature for completion.
233
234Finally, it is impossible to describe a programming language without usages before definitions.
235Therefore, syntax and semantics appear before explanations;
236hence, patience is necessary until details are presented.
237
238
239\section{Polymorphic Functions}
240
241\CFA introduces both ad-hoc and parametric polymorphism to C, with a design originally formalized by Ditchfield~\cite{Ditchfield92}, and first implemented by Bilson~\cite{Bilson03}.
242Shortcomings are identified in existing approaches to generic and variadic data types in C-like languages and how these shortcomings are avoided in \CFA.
243Specifically, 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.
244The new constructs are empirically compared with C and \CC approaches via performance experiments in Section~\ref{sec:eval}.
245
246
247\subsection{Name Overloading}
248\label{s:NameOverloading}
249
250\begin{quote}
251There are only two hard things in Computer Science: cache invalidation and \emph{naming things} -- Phil Karlton
252\end{quote}
253C already has a limited form of ad-hoc polymorphism in the form of its basic arithmetic operators, which apply to a variety of different types using identical syntax.
254\CFA extends the built-in operator overloading by allowing users to define overloads for any function, not just operators, and even any variable;
255Section~\ref{sec:libraries} includes a number of examples of how this overloading simplifies \CFA programming relative to C.
256Code generation for these overloaded functions and variables is implemented by the usual approach of mangling the identifier names to include a representation of their type, while \CFA decides which overload to apply based on the same ``usual arithmetic conversions'' used in C to disambiguate operator overloads.
257As an example:
258
259\begin{cfa}
260int max = 2147483647;                                           $\C[3.75in]{// (1)}$
261double max = 1.7976931348623157E+308;   $\C{// (2)}$
262int max( int a, int b ) { return a < b ? b : a; }  $\C{// (3)}$
263double max( double a, double b ) { return a < b ? b : a; }  $\C{// (4)}\CRT$
264max( 7, -max );                                                         $\C{// uses (3) and (1), by matching int from constant 7}$
265max( max, 3.14 );                                                       $\C{// uses (4) and (2), by matching double from constant 3.14}$
266max( max, -max );                                                       $\C{// ERROR: ambiguous}$
267int m = max( max, -max );                                       $\C{// uses (3) and (1) twice, by matching return type}$
268\end{cfa}
269\CFA maximizes the ability to reuse names to aggressively address the naming problem.
270In some cases, hundreds of names can be reduced to tens, resulting in a significant cognitive reduction.
271In the above, the name @max@ has a consistent meaning, and a programmer only needs to remember the single concept: maximum.
272To prevent significant ambiguities, \CFA uses the return type in selecting overloads, \eg in the assignment to @m@, the compiler use @m@'s type to unambiguously select the most appropriate call to function @max@ (as does Ada).
273As is shown later, there are a number of situations where \CFA takes advantage of available type information to disambiguate, where other programming languages generate ambiguities.
274
275\Celeven added @_Generic@ expressions, which is used in preprocessor macros to provide a form of ad-hoc polymorphism;
276however, this polymorphism is both functionally and ergonomically inferior to \CFA name overloading.
277The macro wrapping the generic expression imposes some limitations;
278\eg, it cannot implement the example above, because the variables @max@ are ambiguous with the functions @max@.
279Ergonomic limitations of @_Generic@ include the necessity to put a fixed list of supported types in a single place and manually dispatch to appropriate overloads, as well as possible namespace pollution from the dispatch functions, which must all have distinct names.
280For backwards compatibility, \CFA supports @_Generic@ expressions, but it is an unnecessary mechanism. \TODO{actually implement that}
281
282% http://fanf.livejournal.com/144696.html
283% http://www.robertgamble.net/2012/01/c11-generic-selections.html
284% https://abissell.com/2014/01/16/c11s-_generic-keyword-macro-applications-and-performance-impacts/
285
286
287\subsection{\texorpdfstring{\LstKeywordStyle{forall} Functions}{forall Functions}}
288\label{sec:poly-fns}
289
290The 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):
291\begin{cfa}
292`forall( otype T )` T identity( T val ) { return val; }
293int forty_two = identity( 42 );                         $\C{// T is bound to int, forty\_two == 42}$
294\end{cfa}
295This @identity@ function can be applied to any complete \newterm{object type} (or @otype@).
296The 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.
297The \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.
298If this extra information is not needed, \eg for a pointer, the type parameter can be declared as a \newterm{data type} (or @dtype@).
299
300In \CFA, the polymorphic runtime-cost is spread over each polymorphic call, because more arguments are passed to polymorphic functions;
301the experiments in Section~\ref{sec:eval} show this overhead is similar to \CC virtual-function calls.
302A design advantage is that, unlike \CC template-functions, \CFA polymorphic-functions are compatible with C \emph{separate compilation}, preventing compilation and code bloat.
303
304Since bare polymorphic-types provide a restricted set of available operations, \CFA provides a \newterm{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.
305For example, the function @twice@ can be defined using the \CFA syntax for operator overloading:
306\begin{cfa}
307forall( otype T `| { T ?+?(T, T); }` ) T twice( T x ) { return x `+` x; }       $\C{// ? denotes operands}$
308int val = twice( twice( 3.7 ) );
309\end{cfa}
310which works for any type @T@ with a matching addition operator.
311The 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@.
312There 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.
313The first approach has a late conversion from @double@ to @int@ on the final assignment, while the second has an eager conversion to @int@.
314\CFA minimizes the number of conversions and their potential to lose information, so it selects the first approach, which corresponds with C-programmer intuition.
315
316Crucial to the design of a new programming language are the libraries to access thousands of external software features.
317Like \CC, \CFA inherits a massive compatible library-base, where other programming languages must rewrite or provide fragile inter-language communication with C.
318A simple example is leveraging the existing type-unsafe (@void *@) C @bsearch@ to binary search a sorted float array:
319\begin{cfa}
320void * bsearch( const void * key, const void * base, size_t nmemb, size_t size,
321                                         int (* compar)( const void *, const void * ));
322int comp( const void * t1, const void * t2 ) {
323         return *(double *)t1 < *(double *)t2 ? -1 : *(double *)t2 < *(double *)t1 ? 1 : 0;
324}
325double key = 5.0, vals[10] = { /* 10 sorted float values */ };
326double * val = (double *)bsearch( &key, vals, 10, sizeof(vals[0]), comp );      $\C{// search sorted array}$
327\end{cfa}
328which can be augmented simply with a generalized, type-safe, \CFA-overloaded wrappers:
329\begin{cfa}
330forall( otype T | { int ?<?( T, T ); } ) T * bsearch( T key, const T * arr, size_t size ) {
331        int comp( const void * t1, const void * t2 ) { /* as above with double changed to T */ }
332        return (T *)bsearch( &key, arr, size, sizeof(T), comp );
333}
334forall( otype T | { int ?<?( T, T ); } ) unsigned int bsearch( T key, const T * arr, size_t size ) {
335        T * result = bsearch( key, arr, size ); $\C{// call first version}$
336        return result ? result - arr : size;    $\C{// pointer subtraction includes sizeof(T)}$
337}
338double * val = bsearch( 5.0, vals, 10 );        $\C{// selection based on return type}$
339int posn = bsearch( 5.0, vals, 10 );
340\end{cfa}
341The nested function @comp@ provides the hidden interface from typed \CFA to untyped (@void *@) C, plus the cast of the result.
342Providing a hidden @comp@ function in \CC is awkward as lambdas do not use C calling-conventions and template declarations cannot appear at block scope.
343As well, an alternate kind of return is made available: position versus pointer to found element.
344\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 template @bsearch@.
345
346\CFA has replacement libraries condensing hundreds of existing C functions into tens of \CFA overloaded functions, all without rewriting the actual computations (see Section~\ref{sec:libraries}).
347For example, it is possible to write a type-safe \CFA wrapper @malloc@ based on the C @malloc@:
348\begin{cfa}
349forall( dtype T | sized(T) ) T * malloc( void ) { return (T *)malloc( sizeof(T) ); }
350int * ip = malloc();                                            $\C{// select type and size from left-hand side}$
351double * dp = malloc();
352struct S {...} * sp = malloc();
353\end{cfa}
354where the return type supplies the type/size of the allocation, which is impossible in most type systems.
355
356Call-site inferencing and nested functions provide a localized form of inheritance.
357For example, the \CFA @qsort@ only sorts in ascending order using @<@.
358However, it is trivial to locally change this behaviour:
359\begin{cfa}
360forall( otype T | { int ?<?( T, T ); } ) void qsort( const T * arr, size_t size ) { /* use C qsort */ }
361{
362        int ?<?( double x, double y ) { return x `>` y; }       $\C{// locally override behaviour}$
363        qsort( vals, size );                                    $\C{// descending sort}$
364}
365\end{cfa}
366Within the block, the nested version of @?<?@ performs @?>?@ and this local version overrides the built-in @?<?@ so it is passed to @qsort@.
367Hence, programmers can easily form local environments, adding and modifying appropriate functions, to maximize reuse of other existing functions and types.
368
369
370\subsection{Traits}
371
372\CFA provides \newterm{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:
373\begin{cfa}
374trait `summable`( otype T ) {
375        void ?{}( T *, zero_t );                                $\C{// constructor from 0 literal}$
376        T ?+?( T, T );                                                  $\C{// assortment of additions}$
377        T ?+=?( T *, T );
378        T ++?( T * );
379        T ?++( T * ); };
380
381forall( otype T `| summable( T )` ) T sum( T a[$\,$], size_t size ) {  // use trait
382        `T` total = { `0` };                                    $\C{// instantiate T from 0 by calling its constructor}$
383        for ( unsigned int i = 0; i < size; i += 1 ) total `+=` a[i]; $\C{// select appropriate +}$
384        return total; }
385\end{cfa}
386
387In 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:
388\begin{cfa}
389trait otype( dtype T | sized(T) ) {  // sized is a pseudo-trait for types with known size and alignment
390        void ?{}( T & );                                                $\C{// default constructor}$
391        void ?{}( T &, T );                                             $\C{// copy constructor}$
392        void ?=?( T &, T );                                             $\C{// assignment operator}$
393        void ^?{}( T & ); };                                    $\C{// destructor}$
394\end{cfa}
395Given 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.
396
397In summation, the \CFA type-system uses \newterm{nominal typing} for concrete types, matching with the C type-system, and \newterm{structural typing} for polymorphic types.
398Hence, trait names play no part in type equivalence;
399the names are simply macros for a list of polymorphic assertions, which are expanded at usage sites.
400Nevertheless, trait names form a logical subtype-hierarchy with @dtype@ at the top, where traits often contain overlapping assertions, \eg operator @+@.
401Traits are used like interfaces in Java or abstract base-classes in \CC, but without the nominal inheritance-relationships.
402Instead, 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.
403Hence, 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.
404(Nominal inheritance can be approximated with traits using marker variables or functions, as is done in Go.)
405
406% Nominal inheritance can be simulated with traits using marker variables or functions:
407% \begin{cfa}
408% trait nominal(otype T) {
409%     T is_nominal;
410% };
411% int is_nominal;                                                               $\C{// int now satisfies the nominal trait}$
412% \end{cfa}
413%
414% 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:
415% \begin{cfa}
416% trait pointer_like(otype Ptr, otype El) {
417%     lvalue El *?(Ptr);                                                $\C{// Ptr can be dereferenced into a modifiable value of type El}$
418% }
419% struct list {
420%     int value;
421%     list * next;                                                              $\C{// may omit "struct" on type names as in \CC}$
422% };
423% typedef list * list_iterator;
424%
425% lvalue int *?( list_iterator it ) { return it->value; }
426% \end{cfa}
427% 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@).
428% 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.
429
430
431\section{Generic Types}
432
433A significant shortcoming of standard C is the lack of reusable type-safe abstractions for generic data structures and algorithms.
434Broadly speaking, there are three approaches to implement abstract data-structures in C.
435One approach is to write bespoke data-structures for each context in which they are needed.
436While 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.
437A second approach is to use @void *@--based polymorphism, \eg the C standard-library functions @bsearch@ and @qsort@, which allow reuse of code with common functionality.
438However, 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 is not otherwise needed.
439A 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.
440Furthermore, writing and using preprocessor macros is unnatural and inflexible.
441
442\CC, Java, and other languages use \newterm{generic types} to produce type-safe abstract data-types.
443\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.
444However, for known concrete parameters, the generic-type definition can be inlined, like \CC templates.
445
446A 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:
447\begin{cfa}
448forall( otype R, otype S ) struct pair {
449        R first;
450        S second;
451};
452forall( otype T ) T value( pair( const char *, T ) p ) { return p.second; } $\C{// dynamic}$
453forall( dtype F, otype T ) T value( pair( F *, T * ) p ) { return *p.second; } $\C{// dtype-static (concrete)}$
454
455pair( const char *, int ) p = { "magic", 42 }; $\C{// concrete}$
456int i = value( p );
457pair( void *, int * ) q = { 0, &p.second }; $\C{// concrete}$
458i = value( q );
459double d = 1.0;
460pair( double *, double * ) r = { &d, &d }; $\C{// concrete}$
461d = value( r );
462\end{cfa}
463
464\CFA classifies generic types as either \newterm{concrete} or \newterm{dynamic}.
465Concrete types have a fixed memory layout regardless of type parameters, while dynamic types vary in memory layout depending on their type parameters.
466A \newterm{dtype-static} type has polymorphic parameters but is still concrete.
467Polymorphic 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.
468
469\CFA generic types also allow checked argument-constraints.
470For example, the following declaration of a sorted set-type ensures the set key supports equality and relational comparison:
471\begin{cfa}
472forall( otype Key | { _Bool ?==?(Key, Key); _Bool ?<?(Key, Key); } ) struct sorted_set;
473\end{cfa}
474
475
476\subsection{Concrete Generic-Types}
477
478The \CFA translator template-expands concrete generic-types into new structure types, affording maximal inlining.
479To 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.
480A 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.
481For example, the concrete instantiation for @pair( const char *, int )@ is:
482\begin{cfa}
483struct _pair_conc0 {
484        const char * first;
485        int second;
486};
487\end{cfa}
488
489A concrete generic-type with dtype-static parameters is also expanded to a structure type, but this type is used for all matching instantiations.
490In the above example, the @pair( F *, T * )@ parameter to @value@ 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:
491\begin{cfa}
492struct _pair_conc1 {
493        void * first;
494        void * second;
495};
496\end{cfa}
497
498
499\subsection{Dynamic Generic-Types}
500
501Though \CFA implements concrete generic-types efficiently, it also has a fully general system for dynamic generic types.
502As 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.
503Dynamic generic-types also have an \newterm{offset array} containing structure-member offsets.
504A dynamic generic-@union@ needs no such offset array, as all members are at offset 0, but size and alignment are still necessary.
505Access 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.
506
507The offset arrays are statically generated where possible.
508If 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;
509if 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.
510As an example, the body of the second @value@ function is implemented like this:
511\begin{cfa}
512_assign_T(_retval, p + _offsetof_pair[1]); $\C{// return *p.second}$
513\end{cfa}
514@_assign_T@ is passed in as an implicit parameter from @otype T@, and takes two @T*@ (@void*@ in the generated code), a destination and a source; @_retval@ is the pointer to a caller-allocated buffer for the return value, the usual \CFA method to handle dynamically-sized return types.
515@_offsetof_pair@ is the offset array passed into @value@; this array is generated at the call site as:
516\begin{cfa}
517size_t _offsetof_pair[] = { offsetof(_pair_conc0, first), offsetof(_pair_conc0, second) }
518\end{cfa}
519
520In some cases the offset arrays cannot be statically generated.
521For 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.
522\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.
523The \CFA translator automatically generates \newterm{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.
524These 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).
525Results of these layout functions are cached so that they are only computed once per type per function. %, as in the example below for @pair@.
526Layout functions also allow generic types to be used in a function definition without reflecting them in the function signature.
527For 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.
528This function could acquire the layout for @set(T)@ by calling its layout function with the layout of @T@ implicitly passed into the function.
529
530Whether a type is concrete, dtype-static, or dynamic is decided solely on the @forall@'s type parameters.
531This 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.
532If 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.
533
534
535\subsection{Applications}
536\label{sec:generic-apps}
537
538The reuse of dtype-static structure instantiations enables useful programming patterns at zero runtime cost.
539The 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@:
540\begin{cfa}
541forall(dtype T) int lexcmp( pair( T *, T * ) * a, pair( T *, T * ) * b, int (* cmp)( T *, T * ) ) {
542        return cmp( a->first, b->first ) ? : cmp( a->second, b->second );
543}
544\end{cfa}
545Since @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.
546
547Another useful pattern enabled by reused dtype-static type instantiations is zero-cost \newterm{tag-structures}.
548Sometimes information is only used for type-checking and can be omitted at runtime, \eg:
549\begin{cfa}
550forall(dtype Unit) struct scalar { unsigned long value; };
551struct metres {};
552struct litres {};
553
554forall(dtype U) scalar(U) ?+?( scalar(U) a, scalar(U) b ) {
555        return (scalar(U)){ a.value + b.value };
556}
557scalar(metres) half_marathon = { 21093 };
558scalar(litres) swimming_pool = { 2500000 };
559scalar(metres) marathon = half_marathon + half_marathon;
560scalar(litres) two_pools = swimming_pool + swimming_pool;
561marathon + swimming_pool;                                       $\C{// compilation ERROR}$
562\end{cfa}
563@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 @?+?@.
564These implementations may even be separately compiled, unlike \CC template functions.
565However, the \CFA type-checker ensures matching types are used by all calls to @?+?@, preventing nonsensical computations like adding a length to a volume.
566
567
568\section{Tuples}
569\label{sec:tuples}
570
571In many languages, functions can return at most one value;
572however, many operations have multiple outcomes, some exceptional.
573Consider C's @div@ and @remquo@ functions, which return the quotient and remainder for a division of integer and float values, respectively.
574\begin{cfa}
575typedef struct { int quo, rem; } div_t;         $\C{// from include stdlib.h}$
576div_t div( int num, int den );
577double remquo( double num, double den, int * quo );
578div_t qr = div( 13, 5 );                                        $\C{// return quotient/remainder aggregate}$
579int q;
580double r = remquo( 13.5, 5.2, &q );                     $\C{// return remainder, alias quotient}$
581\end{cfa}
582@div@ aggregates the quotient/remainder in a structure, while @remquo@ aliases a parameter to an argument.
583Both approaches are awkward.
584Alternatively, a programming language can directly support returning multiple values, \eg in \CFA:
585\begin{cfa}
586[ int, int ] div( int num, int den );           $\C{// return two integers}$
587[ double, double ] div( double num, double den ); $\C{// return two doubles}$
588int q, r;                                                                       $\C{// overloaded variable names}$
589double q, r;
590[ q, r ] = div( 13, 5 );                                        $\C{// select appropriate div and q, r}$
591[ q, r ] = div( 13.5, 5.2 );                            $\C{// assign into tuple}$
592\end{cfa}
593Clearly, this approach is straightforward to understand and use;
594therefore, why do few programming languages support this obvious feature or provide it awkwardly?
595The answer is that there are complex consequences that cascade through multiple aspects of the language, especially the type-system.
596This section show these consequences and how \CFA handles them.
597
598
599\subsection{Tuple Expressions}
600
601The addition of multiple-return-value functions (MRVF) are \emph{useless} without a syntax for accepting multiple values at the call-site.
602The simplest mechanism for capturing the return values is variable assignment, allowing the values to be retrieved directly.
603As such, \CFA allows assigning multiple values from a function into multiple variables, using a square-bracketed list of lvalue expressions (as above), called a \newterm{tuple}.
604
605However, functions also use \newterm{composition} (nested calls), with the direct consequence that MRVFs must also support composition to be orthogonal with single-returning-value functions (SRVF), \eg:
606\begin{cfa}
607printf( "%d %d\n", div( 13, 5 ) );                      $\C{// return values seperated into arguments}$
608\end{cfa}
609Here, the values returned by @div@ are composed with the call to @printf@ by flattening the tuple into separate arguments.
610However, the \CFA type-system must support significantly more complex composition:
611\begin{cfa}
612[ int, int ] foo$\(_1\)$( int );                        $\C{// overloaded foo functions}$
613[ double ] foo$\(_2\)$( int );
614void bar( int, double, double );
615`bar`( foo( 3 ), foo( 3 ) );
616\end{cfa}
617The type-resolver only has the tuple return-types to resolve the call to @bar@ as the @foo@ parameters are identical, which involves unifying the possible @foo@ functions with @bar@'s parameter list.
618No combination of @foo@s are an exact match with @bar@'s parameters, so the resolver applies C conversions.
619The minimal cost is @bar( foo@$_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.
620
621
622\subsection{Tuple Variables}
623
624An important observation from function composition is that new variable names are not required to initialize parameters from an MRVF.
625\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:
626\begin{cfa}
627[ int, int ] qr = div( 13, 5 );                         $\C{// tuple-variable declaration and initialization}$
628[ double, double ] qr = div( 13.5, 5.2 );
629\end{cfa}
630where the tuple variable-name serves the same purpose as the parameter name(s).
631Tuple variables can be composed of any types, except for array types, since array sizes are generally unknown in C.
632
633One way to access the tuple-variable components is with assignment or composition:
634\begin{cfa}
635[ q, r ] = qr;                                                          $\C{// access tuple-variable components}$
636printf( "%d %d\n", qr );
637\end{cfa}
638\CFA also supports \newterm{tuple indexing} to access single components of a tuple expression:
639\begin{cfa}
640[int, int] * p = &qr;                                           $\C{// tuple pointer}$
641int rem = qr`.1`;                                                       $\C{// access remainder}$
642int quo = div( 13, 5 )`.0`;                                     $\C{// access quotient}$
643p`->0` = 5;                                                                     $\C{// change quotient}$
644bar( qr`.1`, qr );                                                      $\C{// pass remainder and quotient/remainder}$
645rem = [div( 13, 5 ), 42]`.0.1`;                         $\C{// access 2nd component of 1st component of tuple expression}$
646\end{cfa}
647
648
649\subsection{Flattening and Restructuring}
650
651In function call contexts, tuples support implicit flattening and restructuring conversions.
652Tuple flattening recursively expands a tuple into the list of its basic components.
653Tuple structuring packages a list of expressions into a value of tuple type, \eg:
654%\lstDeleteShortInline@%
655%\par\smallskip
656%\begin{tabular}{@{}l@{\hspace{1.5\parindent}}||@{\hspace{1.5\parindent}}l@{}}
657\begin{cfa}
658int f( int, int );
659int g( [int, int] );
660int h( int, [int, int] );
661[int, int] x;
662int y;
663f( x );                 $\C{// flatten}$
664g( y, 10 );             $\C{// structure}$
665h( x, y );              $\C{// flatten and structure}$
666\end{cfa}
667%\end{cfa}
668%&
669%\begin{cfa}
670%\end{tabular}
671%\smallskip\par\noindent
672%\lstMakeShortInline@%
673In the call to @f@, @x@ is implicitly flattened so the components of @x@ are passed as the two arguments.
674In 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@.
675Finally, 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]@.
676The 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.
677
678
679\subsection{Tuple Assignment}
680
681An assignment where the left side is a tuple type is called \newterm{tuple assignment}.
682There 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 \newterm{multiple} and \newterm{mass assignment}, respectively.
683%\lstDeleteShortInline@%
684%\par\smallskip
685%\begin{tabular}{@{}l@{\hspace{1.5\parindent}}||@{\hspace{1.5\parindent}}l@{}}
686\begin{cfa}
687int x = 10;
688double y = 3.5;
689[int, double] z;
690z = [x, y];                                                                     $\C{// multiple assignment}$
691[x, y] = z;                                                                     $\C{// multiple assignment}$
692z = 10;                                                                         $\C{// mass assignment}$
693[y, x] = 3.14;                                                          $\C{// mass assignment}$
694\end{cfa}
695%\end{cfa}
696%&
697%\begin{cfa}
698%\end{tabular}
699%\smallskip\par\noindent
700%\lstMakeShortInline@%
701Both kinds of tuple assignment have parallel semantics, so that each value on the left and right side is evaluated before any assignments occur.
702As 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]@.
703This 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.
704For example, @[y, x] = 3.14@ performs the assignments @y = 3.14@ and @x = 3.14@, yielding @y == 3.14@ and @x == 3@;
705whereas, C cascading assignment @y = x = 3.14@ performs the assignments @x = 3.14@ and @y = x@, yielding @3@ in @y@ and @x@.
706Finally, 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.
707This example shows mass, multiple, and cascading assignment used in one expression:
708\begin{cfa}
709void f( [int, int] );
710f( [x, y] = z = 1.5 );                                          $\C{// assignments in parameter list}$
711\end{cfa}
712
713
714\subsection{Member Access}
715
716It is also possible to access multiple fields from a single expression using a \newterm{member-access}.
717The result is a single tuple-valued expression whose type is the tuple of the types of the members, \eg:
718\begin{cfa}
719struct S { int x; double y; char * z; } s;
720s.[x, y, z] = 0;
721\end{cfa}
722Here, the mass assignment sets all members of @s@ to zero.
723Since 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).
724%\lstDeleteShortInline@%
725%\par\smallskip
726%\begin{tabular}{@{}l@{\hspace{1.5\parindent}}||@{\hspace{1.5\parindent}}l@{}}
727\begin{cfa}
728[int, int, long, double] x;
729void f( double, long );
730x.[0, 1] = x.[1, 0];                                            $\C{// rearrange: [x.0, x.1] = [x.1, x.0]}$
731f( x.[0, 3] );                                                          $\C{// drop: f(x.0, x.3)}$
732[int, int, int] y = x.[2, 0, 2];                        $\C{// duplicate: [y.0, y.1, y.2] = [x.2, x.0.x.2]}$
733\end{cfa}
734%\end{cfa}
735%&
736%\begin{cfa}
737%\end{tabular}
738%\smallskip\par\noindent
739%\lstMakeShortInline@%
740It is also possible for a member access to contain other member accesses, \eg:
741\begin{cfa}
742struct A { double i; int j; };
743struct B { int * k; short l; };
744struct C { int x; A y; B z; } v;
745v.[x, y.[i, j], z.k];                                           $\C{// [v.x, [v.y.i, v.y.j], v.z.k]}$
746\end{cfa}
747
748
749\begin{comment}
750\subsection{Casting}
751
752In C, the cast operator is used to explicitly convert between types.
753In \CFA, the cast operator has a secondary use as type ascription.
754That 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:
755\begin{cfa}
756int f();     // (1)
757double f()// (2)
758
759f();       // ambiguous - (1),(2) both equally viable
760(int)f()// choose (2)
761\end{cfa}
762
763Since casting is a fundamental operation in \CFA, casts should be given a meaningful interpretation in the context of tuples.
764Taking a look at standard C provides some guidance with respect to the way casts should work with tuples:
765\begin{cfa}
766int f();
767void g();
768
769(void)f()// (1)
770(int)g()// (2)
771\end{cfa}
772In C, (1) is a valid cast, which calls @f@ and discards its result.
773On the other hand, (2) is invalid, because @g@ does not produce a result, so requesting an @int@ to materialize from nothing is nonsensical.
774Generalizing 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.
775
776Formally, 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 $T_i$.
777Excess 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.
778This approach follows naturally from the way that a cast to @void@ works in C.
779
780For example, in
781\begin{cfa}
782[int, int, int] f();
783[int, [int, int], int] g();
784
785([int, double])f();           $\C{// (1)}$
786([int, int, int])g();         $\C{// (2)}$
787([void, [int, int]])g();      $\C{// (3)}$
788([int, int, int, int])g();    $\C{// (4)}$
789([int, [int, int, int]])g()$\C{// (5)}$
790\end{cfa}
791
792(1) discards the last element of the return value and converts the second element to @double@.
793Since @int@ is effectively a 1-element tuple, (2) discards the second component of the second element of the return value of @g@.
794If @g@ is free of side effects, this expression is equivalent to @[(int)(g().0), (int)(g().1.0), (int)(g().2)]@.
795Since @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)]@).
796
797Note 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.}.
798As such, (4) is invalid because the cast target type contains 4 components, while the source type contains only 3.
799Similarly, (5) is invalid because the cast @([int, int, int])(g().1)@ is invalid.
800That is, it is invalid to cast @[int, int]@ to @[int, int, int]@.
801\end{comment}
802
803
804\subsection{Polymorphism}
805
806Tuples also integrate with \CFA polymorphism as a kind of generic type.
807Due 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:
808\begin{cfa}
809forall(otype T, dtype U) void f( T x, U * y );
810f( [5, "hello"] );
811\end{cfa}
812where @[5, "hello"]@ is flattened, giving argument list @5, "hello"@, and @T@ binds to @int@ and @U@ binds to @const char@.
813Tuples, however, may contain polymorphic components.
814For example, a plus operator can be written to add two triples together.
815\begin{cfa}
816forall(otype T | { T ?+?( T, T ); }) [T, T, T] ?+?( [T, T, T] x, [T, T, T] y ) {
817        return [x.0 + y.0, x.1 + y.1, x.2 + y.2];
818}
819[int, int, int] x;
820int i1, i2, i3;
821[i1, i2, i3] = x + ([10, 20, 30]);
822\end{cfa}
823
824Flattening and restructuring conversions are also applied to tuple types in polymorphic type assertions.
825\begin{cfa}
826int f( [int, double], double );
827forall(otype T, otype U | { T f( T, U, U ); }) void g( T, U );
828g( 5, 10.21 );
829\end{cfa}
830Hence, function parameter and return lists are flattened for the purposes of type unification allowing the example to pass expression resolution.
831This relaxation is possible by extending the thunk scheme described by Bilson~\cite{Bilson03}.
832Whenever 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:
833\begin{cfa}
834int _thunk( int _p0, double _p1, double _p2 ) { return f( [_p0, _p1], _p2 ); }
835\end{cfa}
836so the thunk provides flattening and structuring conversions to inferred functions, improving the compatibility of tuples and polymorphism.
837These thunks take advantage of gcc C nested-functions to produce closures that have the usual function-pointer signature WHAT DOES THIS MEAN???.
838
839
840\subsection{Variadic Tuples}
841\label{sec:variadic-tuples}
842
843To define variadic functions, \CFA adds a new kind of type parameter, @ttype@ (tuple type).
844Matching against a @ttype@ parameter consumes all remaining argument components and packages them into a tuple, binding to the resulting tuple of types.
845In 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.
846As such, @ttype@ variables are also called \newterm{argument packs}.
847
848Like variadic templates, the main way to manipulate @ttype@ polymorphic functions is via recursion.
849Since nothing is known about a parameter pack by default, assertion parameters are key to doing anything meaningful.
850Unlike variadic templates, @ttype@ polymorphic functions can be separately compiled.
851For example, a generalized @sum@ function written using @ttype@:
852\begin{cfa}
853int sum$\(_0\)$() { return 0; }
854forall(ttype Params | { int sum( Params ); } ) int sum$\(_1\)$( int x, Params rest ) {
855        return x + sum( rest );
856}
857sum( 10, 20, 30 );
858\end{cfa}
859Since @sum@\(_0\) does not accept any arguments, it is not a valid candidate function for the call @sum(10, 20, 30)@.
860In 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]@.
861The process continues until @Params@ is bound to @[]@, requiring an assertion @int sum()@, which matches @sum@\(_0\) and terminates the recursion.
862Effectively, 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))@.
863
864It is reasonable to take the @sum@ function a step further to enforce a minimum number of arguments:
865\begin{cfa}
866int sum( int x, int y ) { return x + y; }
867forall(ttype Params | { int sum( int, Params ); } ) int sum( int x, int y, Params rest ) {
868        return sum( x + y, rest );
869}
870\end{cfa}
871One more step permits the summation of any summable type with all arguments of the same type:
872\begin{cfa}
873trait summable(otype T) {
874        T ?+?( T, T );
875};
876forall(otype R | summable( R ) ) R sum( R x, R y ) {
877        return x + y;
878}
879forall(otype R, ttype Params | summable(R) | { R sum(R, Params); } ) R sum(R x, R y, Params rest) {
880        return sum( x + y, rest );
881}
882\end{cfa}
883Unlike C variadic functions, it is unnecessary to hard code the number and expected types.
884Furthermore, this code is extendable for any user-defined type with a @?+?@ operator.
885Summing arbitrary heterogeneous lists is possible with similar code by adding the appropriate type variables and addition operators.
886
887It is also possible to write a type-safe variadic print function to replace @printf@:
888\begin{cfa}
889struct S { int x, y; };
890forall(otype T, ttype Params | { void print(T); void print(Params); }) void print(T arg, Params rest) {
891        print(arg);  print(rest);
892}
893void print( const char * x ) { printf( "%s", x ); }
894void print( int x ) { printf( "%d", x ); }
895void print( S s ) { print( "{ ", s.x, ",", s.y, " }" ); }
896print( "s = ", (S){ 1, 2 }, "\n" );
897\end{cfa}
898This example showcases a variadic-template-like decomposition of the provided argument list.
899The individual @print@ functions allow printing a single element of a type.
900The polymorphic @print@ allows printing any list of types, where as each individual type has a @print@ function.
901The individual print functions can be used to build up more complicated @print@ functions, such as @S@, which cannot be done with @printf@ in C.
902This mechanism is used to seamlessly print tuples in the \CFA I/O library (see Section~\ref{s:IOLibrary}).
903
904Finally, it is possible to use @ttype@ polymorphism to provide arbitrary argument forwarding functions.
905For example, it is possible to write @new@ as a library function:
906\begin{cfa}
907forall( otype R, otype S ) void ?{}( pair(R, S) *, R, S );
908forall( dtype T, ttype Params | sized(T) | { void ?{}( T *, Params ); } ) T * new( Params p ) {
909        return ((T *)malloc()){ p };                    $\C{// construct into result of malloc}$
910}
911pair( int, char ) * x = new( 42, '!' );
912\end{cfa}
913The @new@ function provides the combination of type-safe @malloc@ with a \CFA constructor call, making it impossible to forget constructing dynamically allocated objects.
914This function provides the type-safety of @new@ in \CC, without the need to specify the allocated type again, thanks to return-type inference.
915
916
917\subsection{Implementation}
918
919Tuples are implemented in the \CFA translator via a transformation into \newterm{generic types}.
920For each $N$, the first time an $N$-tuple is seen in a scope a generic type with $N$ type parameters is generated, \eg:
921\begin{cfa}
922[int, int] f() {
923        [double, double] x;
924        [int, double, int] y;
925}
926\end{cfa}
927is transformed into:
928\begin{cfa}
929forall(dtype T0, dtype T1 | sized(T0) | sized(T1)) struct _tuple2 {
930        T0 field_0;                                                             $\C{// generated before the first 2-tuple}$
931        T1 field_1;
932};
933_tuple2(int, int) f() {
934        _tuple2(double, double) x;
935        forall(dtype T0, dtype T1, dtype T2 | sized(T0) | sized(T1) | sized(T2)) struct _tuple3 {
936                T0 field_0;                                                     $\C{// generated before the first 3-tuple}$
937                T1 field_1;
938                T2 field_2;
939        };
940        _tuple3(int, double, int) y;
941}
942\end{cfa}
943\begin{sloppypar}
944Tuple expressions are then simply converted directly into compound literals, \eg @[5, 'x', 1.24]@ becomes @(_tuple3(int, char, double)){ 5, 'x', 1.24 }@.
945\end{sloppypar}
946
947\begin{comment}
948Since tuples are essentially structures, tuple indexing expressions are just field accesses:
949\begin{cfa}
950void f(int, [double, char]);
951[int, double] x;
952
953x.0+x.1;
954printf("%d %g\n", x);
955f(x, 'z');
956\end{cfa}
957Is transformed into:
958\begin{cfa}
959void f(int, _tuple2(double, char));
960_tuple2(int, double) x;
961
962x.field_0+x.field_1;
963printf("%d %g\n", x.field_0, x.field_1);
964f(x.field_0, (_tuple2){ x.field_1, 'z' });
965\end{cfa}
966Note that due to flattening, @x@ used in the argument position is converted into the list of its fields.
967In the call to @f@, the second and third argument components are structured into a tuple argument.
968Similarly, tuple member expressions are recursively expanded into a list of member access expressions.
969
970Expressions that may contain side effects are made into \newterm{unique expressions} before being expanded by the flattening conversion.
971Each unique expression is assigned an identifier and is guaranteed to be executed exactly once:
972\begin{cfa}
973void g(int, double);
974[int, double] h();
975g(h());
976\end{cfa}
977Internally, this expression is converted to two variables and an expression:
978\begin{cfa}
979void g(int, double);
980[int, double] h();
981
982_Bool _unq0_finished_ = 0;
983[int, double] _unq0;
984g(
985        (_unq0_finished_ ? _unq0 : (_unq0 = f(), _unq0_finished_ = 1, _unq0)).0,
986        (_unq0_finished_ ? _unq0 : (_unq0 = f(), _unq0_finished_ = 1, _unq0)).1,
987);
988\end{cfa}
989Since argument evaluation order is not specified by the C programming language, this scheme is built to work regardless of evaluation order.
990The first time a unique expression is executed, the actual expression is evaluated and the accompanying boolean is set to true.
991Every subsequent evaluation of the unique expression then results in an access to the stored result of the actual expression.
992Tuple member expressions also take advantage of unique expressions in the case of possible impurity.
993
994Currently, the \CFA translator has a very broad, imprecise definition of impurity, where any function call is assumed to be impure.
995This 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.
996
997The various kinds of tuple assignment, constructors, and destructors generate GNU C statement expressions.
998A 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.
999The 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.
1000However, 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.
1001\end{comment}
1002
1003
1004\section{Control Structures}
1005
1006\CFA identifies inconsistent, problematic, and missing control structures in C, and extends, modifies, and adds to control structures to increase functionality and safety.
1007
1008
1009\subsection{\texorpdfstring{\LstKeywordStyle{if} Statement}{if Statement}}
1010
1011The @if@ expression allows declarations, similar to @for@ declaration expression:
1012\begin{cfa}
1013if ( int x = f() ) ...                                          $\C{// x != 0}$
1014if ( int x = f(), y = g() ) ...                         $\C{// x != 0 \&\& y != 0}$
1015if ( int x = f(), y = g(); `x < y` ) ...        $\C{// relational expression}$
1016\end{cfa}
1017Unless a relational expression is specified, each variable is compared not equal to 0, which is the standard semantics for the @if@ expression, and the results are combined using the logical @&&@ operator.\footnote{\CC only provides a single declaration always compared not equal to 0.}
1018The scope of the declaration(s) is local to the @if@ statement but exist within both the ``then'' and ``else'' clauses.
1019
1020
1021\subsection{\texorpdfstring{\LstKeywordStyle{switch} Statement}{switch Statement}}
1022
1023There are a number of deficiencies with the C @switch@ statements: enumerating @case@ lists, placement of @case@ clauses, scope of the switch body, and fall through between case clauses.
1024
1025C has no shorthand for specifying a list of case values, whether the list is non-contiguous or contiguous\footnote{C provides this mechanism via fall through.}.
1026\CFA provides a shorthand for a non-contiguous list:
1027\begin{cquote}
1028\lstDeleteShortInline@%
1029\begin{tabular}{@{}l@{\hspace{\parindentlnth}}l@{}}
1030\multicolumn{1}{c@{\hspace{\parindentlnth}}}{\textbf{\CFA}}     & \multicolumn{1}{c}{\textbf{C}}        \\
1031\begin{cfa}
1032case 2, 10, 34, 42:
1033\end{cfa}
1034&
1035\begin{cfa}
1036case 2: case 10: case 34: case 42:
1037\end{cfa}
1038\end{tabular}
1039\lstMakeShortInline@%
1040\end{cquote}
1041for a contiguous list:\footnote{gcc provides the same mechanism with awkward syntax, \lstinline@2 ... 42@, where spaces are required around the ellipse.}
1042\begin{cquote}
1043\lstDeleteShortInline@%
1044\begin{tabular}{@{}l@{\hspace{\parindentlnth}}l@{}}
1045\multicolumn{1}{c@{\hspace{\parindentlnth}}}{\textbf{\CFA}}     & \multicolumn{1}{c}{\textbf{C}}        \\
1046\begin{cfa}
1047case 2~42:
1048\end{cfa}
1049&
1050\begin{cfa}
1051case 2: case 3: ... case 41: case 42:
1052\end{cfa}
1053\end{tabular}
1054\lstMakeShortInline@%
1055\end{cquote}
1056and a combination:
1057\begin{cfa}
1058case -12~-4, -1~5, 14~21, 34~42:
1059\end{cfa}
1060
1061C allows placement of @case@ clauses \emph{within} statements nested in the @switch@ body (see Duff's device~\cite{Duff83});
1062\begin{cfa}
1063switch ( i ) {
1064  case 0:
1065        for ( int i = 0; i < 10; i += 1 ) {
1066                ...
1067  `case 1:`             // no initialization of loop index
1068                ...
1069        }
1070}
1071\end{cfa}
1072\CFA precludes this form of transfer into a control structure because it causes undefined behaviour, especially with respect to missed initialization, and provides very limited functionality.
1073
1074C allows placement of declaration within the @switch@ body and unreachable code at the start, resulting in undefined behaviour:
1075\begin{cfa}
1076switch ( x ) {
1077        `int y = 1;`                                                    $\C{// unreachable initialization}$
1078        `x = 7;`                                                                $\C{// unreachable code without label/branch}$
1079  case 0:
1080        ...
1081        `int z = 0;`                                                    $\C{// unreachable initialization, cannot appear after case}$
1082        z = 2;
1083  case 1:
1084        `x = z;`                                                                $\C{// without fall through, z is undefined}$
1085}
1086\end{cfa}
1087\CFA allows the declaration of local variables, \eg @y@, at the start of the @switch@ with scope across the entire @switch@ body, \ie all @case@ clauses.
1088\CFA disallows the declaration of local variable, \eg @z@, directly within the @switch@ body, because a declaration cannot occur immediately after a @case@ since a label can only be attached to a statement, and the use of @z@ is undefined in @case 1@ as neither storage allocation nor initialization may have occurred.
1089
1090C @switch@ provides multiple entry points into the statement body, but once an entry point is selected, control continues across \emph{all} @case@ clauses until the end of the @switch@ body, called \newterm{fall through};
1091@case@ clauses are made disjoint by the @break@ statement.
1092While the ability to fall through \emph{is} a useful form of control flow, it does not match well with programmer intuition, resulting in many errors from missing @break@ statements.
1093For backwards compatibility, \CFA provides a \emph{new} control structure, @choose@, which mimics @switch@, but reverses the meaning of fall through (see Figure~\ref{f:ChooseSwitchStatements}).
1094
1095Collectively, these enhancements reduce programmer burden and increase readability and safety.
1096
1097\begin{figure}
1098\centering
1099\lstDeleteShortInline@%
1100\begin{tabular}{@{}l@{\hspace{2\parindentlnth}}l@{}}
1101\multicolumn{1}{c@{\hspace{2\parindentlnth}}}{\textbf{\CFA}}    & \multicolumn{1}{c}{\textbf{C}}        \\
1102\begin{cfa}
1103`choose` ( day ) {
1104  case Mon~Thu:  // program
1105
1106  case Fri:  // program
1107        wallet += pay;
1108        `fallthrough;`
1109  case Sat:  // party
1110        wallet -= party;
1111
1112  case Sun:  // rest
1113
1114  default:  // error
1115}
1116\end{cfa}
1117&
1118\begin{cfa}
1119switch ( day ) {
1120  case Mon: case Tue: case Wed: case Thu:  // program
1121        `break;`
1122  case Fri:  // program
1123        wallet += pay;
1124
1125  case Sat:  // party
1126        wallet -= party;
1127        `break;`
1128  case Sun:  // rest
1129        `break;`
1130  default:  // error
1131}
1132\end{cfa}
1133\end{tabular}
1134\lstMakeShortInline@%
1135\caption{\lstinline|choose| versus \lstinline|switch| Statements}
1136\label{f:ChooseSwitchStatements}
1137\end{figure}
1138
1139\begin{comment}
1140Forgotten @break@ statements at the end of @switch@ cases are a persistent sort of programmer error in C, and the @break@ statements themselves introduce visual clutter and an un-C-like keyword-based block delimiter.
1141\CFA addresses this error by introducing a @choose@ statement, which works identically to a @switch@ except that its default end-of-case behaviour is to break rather than to fall through for all non-empty cases.
1142Since empty cases like @case 7:@ in @case 7: case 11:@ still have fall-through semantics and explicit @break@ is still allowed at the end of a @choose@ case, many idiomatic uses of @switch@ in standard C can be converted to @choose@ statements by simply changing the keyword.
1143Where fall-through is desired for a non-empty case, it can be specified with the new @fallthrough@ statement, making @choose@ equivalently powerful to @switch@, but more concise in the common case where most non-empty cases end with a @break@ statement, as in the example below:
1144
1145\begin{cfa}
1146choose( i ) {
1147        case 2:
1148                printf("even ");
1149                fallthrough;
1150        case 3: case 5: case 7:
1151                printf("small prime\n");
1152        case 4,6,8,9:
1153                printf("small composite\n");
1154        case 13~19:
1155                printf("teen\n");
1156        default:
1157                printf("something else\n");
1158}
1159\end{cfa}
1160\end{comment}
1161
1162
1163\subsection{\texorpdfstring{Labelled \LstKeywordStyle{continue} / \LstKeywordStyle{break}}{Labelled continue / break}}
1164
1165While C provides @continue@ and @break@ statements for altering control flow, both are restricted to one level of nesting for a particular control structure.
1166Unfortunately, this restriction forces programmers to use @goto@ to achieve the equivalent control-flow for more than one level of nesting.
1167To 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.
1168For both @continue@ and @break@, the target label must be directly associated with a @for@, @while@ or @do@ statement;
1169for @break@, the target label can also be associated with a @switch@, @if@ or compound (@{}@) statement.
1170Figure~\ref{f:MultiLevelExit} shows @continue@ and @break@ indicating the specific control structure, and the corresponding C program using only @goto@ and labels.
1171The innermost loop has 7 exit points, which cause continuation or termination of one or more of the 7 nested control-structures.
1172
1173\begin{figure}
1174\lstDeleteShortInline@%
1175\begin{tabular}{@{\hspace{\parindentlnth}}l@{\hspace{\parindentlnth}}l@{\hspace{\parindentlnth}}l@{}}
1176\multicolumn{1}{@{\hspace{\parindentlnth}}c@{\hspace{\parindentlnth}}}{\textbf{\CFA}}   & \multicolumn{1}{@{\hspace{\parindentlnth}}c}{\textbf{C}}      \\
1177\begin{cfa}
1178`LC:` {
1179        ... $declarations$ ...
1180        `LS:` switch ( ... ) {
1181          case 3:
1182                `LIF:` if ( ... ) {
1183                        `LF:` for ( ... ) {
1184                                `LW:` while ( ... ) {
1185                                        ... break `LC`; ...
1186                                        ... break `LS`; ...
1187                                        ... break `LIF`; ...
1188                                        ... continue `LF;` ...
1189                                        ... break `LF`; ...
1190                                        ... continue `LW`; ...
1191                                        ... break `LW`; ...
1192                                } // while
1193                        } // for
1194                } else {
1195                        ... break `LIF`; ...
1196                } // if
1197        } // switch
1198} // compound
1199\end{cfa}
1200&
1201\begin{cfa}
1202{
1203        ... $declarations$ ...
1204        switch ( ... ) {
1205          case 3:
1206                if ( ... ) {
1207                        for ( ... ) {
1208                                while ( ... ) {
1209                                        ... goto `LC`; ...
1210                                        ... goto `LS`; ...
1211                                        ... goto `LIF`; ...
1212                                        ... goto `LFC`; ...
1213                                        ... goto `LFB`; ...
1214                                        ... goto `LWC`; ...
1215                                        ... goto `LWB`; ...
1216                                  `LWC`: ; } `LWB:` ;
1217                          `LFC:` ; } `LFB:` ;
1218                } else {
1219                        ... goto `LIF`; ...
1220                } `LIF:` ;
1221        } `LS:` ;
1222} `LC:` ;
1223\end{cfa}
1224&
1225\begin{cfa}
1226
1227
1228
1229
1230
1231
1232
1233// terminate compound
1234// terminate switch
1235// terminate if
1236// continue loop
1237// terminate loop
1238// continue loop
1239// terminate loop
1240
1241
1242
1243// terminate if
1244
1245
1246
1247\end{cfa}
1248\end{tabular}
1249\lstMakeShortInline@%
1250\caption{Multi-level Exit}
1251\label{f:MultiLevelExit}
1252\end{figure}
1253
1254With respect to safety, both labelled @continue@ and @break@ are a @goto@ restricted in the following ways:
1255\begin{itemize}
1256\item
1257They cannot create a loop, which means only the looping constructs cause looping.
1258This restriction means all situations resulting in repeated execution are clearly delineated.
1259\item
1260They cannot branch into a control structure.
1261This restriction prevents missing declarations and/or initializations at the start of a control structure resulting in undefined behaviour.
1262\end{itemize}
1263The 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.
1264Furthermore, 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.
1265With @goto@, the label is at the end of the control structure, which fails to convey this important clue early enough to the reader.
1266Finally, using an explicit target for the transfer instead of an implicit target allows new constructs to be added or removed without affecting existing constructs.
1267Otherwise, the implicit targets of the current @continue@ and @break@, \ie the closest enclosing loop or @switch@, change as certain constructs are added or removed.
1268
1269
1270\subsection{Exception Handling}
1271
1272The following framework for \CFA exception handling is in place, excluding some run-time type-information and dynamic casts.
1273\CFA provides two forms of exception handling: \newterm{fix-up} and \newterm{recovery} (see Figure~\ref{f:CFAExceptionHandling})~\cite{Buhr92b,Buhr00a}.
1274Both mechanisms provide dynamic call to a handler using dynamic name-lookup, where fix-up has dynamic return and recovery has static return from the handler.
1275\CFA restricts exception types to those defined by aggregate type @exception@.
1276The form of the raise dictates the set of handlers examined during propagation: \newterm{resumption propagation} (@resume@) only examines resumption handlers (@catchResume@); \newterm{terminating propagation} (@throw@) only examines termination handlers (@catch@).
1277If @resume@ or @throw@ have no exception type, it is a reresume/rethrow, meaning the currently exception continues propagation.
1278If there is no current exception, the reresume/rethrow results in a runtime error.
1279
1280\begin{figure}
1281\begin{cquote}
1282\lstDeleteShortInline@%
1283\begin{tabular}{@{}l@{\hspace{2\parindentlnth}}l@{}}
1284\multicolumn{1}{c@{\hspace{2\parindentlnth}}}{\textbf{Resumption}}      & \multicolumn{1}{c}{\textbf{Termination}}      \\
1285\begin{cfa}
1286`exception R { int fix; };`
1287void f() {
1288        R r;
1289        ... `resume( r );` ...
1290        ... r.fix // control does return here after handler
1291}
1292`try` {
1293        ... f(); ...
1294} `catchResume( R r )` {
1295        ... r.fix = ...; // return correction to raise
1296} // dynamic return to _Resume
1297\end{cfa}
1298&
1299\begin{cfa}
1300`exception T {};`
1301void f() {
1302
1303        ... `throw( T{} );` ...
1304        // control does NOT return here after handler
1305}
1306`try` {
1307        ... f(); ...
1308} `catch( T t )` {
1309        ... // recover and continue
1310} // static return to next statement
1311\end{cfa}
1312\end{tabular}
1313\lstMakeShortInline@%
1314\end{cquote}
1315\caption{\CFA Exception Handling}
1316\label{f:CFAExceptionHandling}
1317\end{figure}
1318
1319The set of exception types in a list of catch clause may include both a resumption and termination handler:
1320\begin{cfa}
1321try {
1322        ... resume( `R{}` ); ...
1323} catchResume( `R` r ) { ... throw( R{} ); ... } $\C{\color{red}// H1}$
1324   catch( `R` r ) { ... }                                       $\C{\color{red}// H2}$
1325
1326\end{cfa}
1327The resumption propagation raises @R@ and the stack is not unwound;
1328the exception is caught by the @catchResume@ clause and handler H1 is invoked.
1329The termination propagation in handler H1 raises @R@ and the stack is unwound;
1330the exception is caught by the @catch@ clause and handler H2 is invoked.
1331The termination handler is available because the resumption propagation did not unwind the stack.
1332
1333An additional feature is conditional matching in a catch clause:
1334\begin{cfa}
1335try {
1336        ... write( `datafile`, ... ); ...               $\C{// may throw IOError}$
1337        ... write( `logfile`, ... ); ...
1338} catch ( IOError err; `err.file == datafile` ) { ... } $\C{// handle datafile error}$
1339   catch ( IOError err; `err.file == logfile` ) { ... } $\C{// handle logfile error}$
1340   catch ( IOError err ) { ... }                        $\C{// handler error from other files}$
1341\end{cfa}
1342where the throw inserts the failing file-handle in the I/O exception.
1343Conditional catch cannot be trivially mimicked by other mechanisms because once an exception is caught, handler clauses in that @try@ statement are no longer eligible..
1344
1345The resumption raise can specify an alternate stack on which to raise an exception, called a \newterm{nonlocal raise}:
1346\begin{cfa}
1347resume( $\emph{exception-type}$, $\emph{alternate-stack}$ )
1348resume( $\emph{alternate-stack}$ )
1349\end{cfa}
1350These overloads of @resume@ raise the specified exception or the currently propagating exception (reresume) at another \CFA coroutine or task~\cite{Delisle18}.\footnote{\CFA coroutine and concurrency features are discussed in a separately submitted paper.}
1351Nonlocal raise is restricted to resumption to provide the exception handler the greatest flexibility because processing the exception does not unwind its stack, allowing it to continue after the handle returns.
1352
1353To facilitate nonlocal exception, \CFA provides dynamic enabling and disabling of nonlocal exception-propagation.
1354The constructs for controlling propagation of nonlocal exceptions are the @enable@ and the @disable@ blocks:
1355\begin{cquote}
1356\lstDeleteShortInline@%
1357\begin{tabular}{@{}l@{\hspace{2\parindentlnth}}l@{}}
1358\begin{cfa}
1359enable $\emph{exception-type-list}$ {
1360        // allow non-local resumption
1361}
1362\end{cfa}
1363&
1364\begin{cfa}
1365disable $\emph{exception-type-list}$ {
1366        // disallow non-local resumption
1367}
1368\end{cfa}
1369\end{tabular}
1370\lstMakeShortInline@%
1371\end{cquote}
1372The arguments for @enable@/@disable@ specify the exception types allowed to be propagated or postponed, respectively.
1373Specifying no exception type is shorthand for specifying all exception types.
1374Both @enable@ and @disable@ blocks can be nested, turning propagation on/off on entry, and on exit, the specified exception types are restored to their prior state.
1375Coroutines and tasks start with non-local exceptions disabled, allowing handlers to be put in place, before non-local exceptions are explicitly enabled.
1376\begin{cfa}
1377void main( mytask & c ) {
1378        try {
1379                enable {                        $\C{// now allow non-local exception delivery}$
1380                        // task body
1381                }
1382        // appropriate catchResume/catch
1383        }
1384}
1385\end{cfa}
1386
1387Finally, \CFA provides a Java like  @finally@ clause after the catch clauses:
1388\begin{cfa}
1389try {
1390        ... f(); ...
1391// catchResume or catch clauses
1392} `finally` {
1393        // house keeping
1394}
1395\end{cfa}
1396The finally clause is always executed, i.e., if the try block ends normally or if an exception is raised.
1397If an exception is raised and caught, the handler is run before the finally clause.
1398Like a destructor (see Section~\ref{s:ConstructorsDestructors}), a finally clause can raise an exception but not if there is an exception being propagated.
1399Mimicking the @finally@ clause with mechanisms like RAII is non-trivially when there are multiple types and local accesses.
1400
1401
1402\subsection{\texorpdfstring{\LstKeywordStyle{with} Clause / Statement}{with Clause / Statement}}
1403\label{s:WithClauseStatement}
1404
1405Grouping heterogeneous 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:
1406\begin{cfa}
1407struct S {                                                                      $\C{// aggregate}$
1408        char c;                                                                 $\C{// fields}$
1409        int i;
1410        double d;
1411};
1412S s, as[10];
1413\end{cfa}
1414However, functions manipulating aggregates must repeat the aggregate name to access its containing fields:
1415\begin{cfa}
1416void f( S s ) {
1417        `s.`c; `s.`i; `s.`d;                                    $\C{// access containing fields}$
1418}
1419\end{cfa}
1420which extends to multiple levels of qualification for nested aggregates.
1421A similar situation occurs in object-oriented programming, \eg \CC:
1422\begin{C++}
1423class C {
1424        char c;                                                                 $\C{// fields}$
1425        int i;
1426        double d;
1427        int f() {                                                               $\C{// implicit ``this'' aggregate}$
1428                `this->`c; `this->`i; `this->`d;        $\C{// access containing fields}$
1429        }
1430}
1431\end{C++}
1432Object-oriented nesting of member functions in a \lstinline[language=C++]@class@ allows eliding \lstinline[language=C++]@this->@ because of lexical scoping.
1433However, for other aggregate parameters, qualification is necessary:
1434\begin{cfa}
1435struct T { double m, n; };
1436int C::f( T & t ) {                                                     $\C{// multiple aggregate parameters}$
1437        c; i; d;                                                                $\C{\color{red}// this-\textgreater.c, this-\textgreater.i, this-\textgreater.d}$
1438        `t.`m; `t.`n;                                                   $\C{// must qualify}$
1439}
1440\end{cfa}
1441
1442To simplify the programmer experience, \CFA provides a @with@ statement (see Pascal~\cite[\S~4.F]{Pascal}) to elide aggregate qualification to fields by opening a scope containing the field identifiers.
1443Hence, the qualified fields become variables with the side-effect that it is easier to optimizing field references in a block.
1444\begin{cfa}
1445void f( S & this ) `with ( this )` {            $\C{// with statement}$
1446        c; i; d;                                                                $\C{\color{red}// this.c, this.i, this.d}$
1447}
1448\end{cfa}
1449with the generality of opening multiple aggregate-parameters:
1450\begin{cfa}
1451int f( S & s, T & t ) `with ( s, t )` {         $\C{// multiple aggregate parameters}$
1452        c; i; d;                                                                $\C{\color{red}// s.c, s.i, s.d}$
1453        m; n;                                                                   $\C{\color{red}// t.m, t.n}$
1454}
1455\end{cfa}
1456
1457In detail, the @with@ statement has the form:
1458\begin{cfa}
1459$\emph{with-statement}$:
1460        'with' '(' $\emph{expression-list}$ ')' $\emph{compound-statement}$
1461\end{cfa}
1462and may appear as the body of a function or nested within a function body.
1463Each expression in the expression-list provides a type and object.
1464The type must be an aggregate type.
1465(Enumerations are already opened.)
1466The object is the implicit qualifier for the open structure-fields.
1467
1468All expressions in the expression list are open in parallel within the compound statement.
1469This semantic is different from Pascal, which nests the openings from left to right.
1470The difference between parallel and nesting occurs for fields with the same name and type:
1471\begin{cfa}
1472struct S { int `i`; int j; double m; } s, w;
1473struct T { int `i`; int k; int m; } t, w;
1474with ( s, t ) {
1475        j + k;                                                                  $\C{// unambiguous, s.j + t.k}$
1476        m = 5.0;                                                                $\C{// unambiguous, t.m = 5.0}$
1477        m = 1;                                                                  $\C{// unambiguous, s.m = 1}$
1478        int a = m;                                                              $\C{// unambiguous, a = s.i }$
1479        double b = m;                                                   $\C{// unambiguous, b = t.m}$
1480        int c = s.i + t.i;                                              $\C{// unambiguous, qualification}$
1481        (double)m;                                                              $\C{// unambiguous, cast}$
1482}
1483\end{cfa}
1484For parallel semantics, both @s.i@ and @t.i@ are visible, so @i@ is ambiguous without qualification;
1485for nested semantics, @t.i@ hides @s.i@, so @i@ implies @t.i@.
1486\CFA's ability to overload variables means fields with the same name but different types are automatically disambiguated, eliminating most qualification when opening multiple aggregates.
1487Qualification or a cast is used to disambiguate.
1488
1489There is an interesting problem between parameters and the function-body @with@, \eg:
1490\begin{cfa}
1491void ?{}( S & s, int i ) with ( s ) {           $\C{// constructor}$
1492        `s.i = i;`  j = 3;  m = 5.5;                    $\C{// initialize fields}$
1493}
1494\end{cfa}
1495Here, the assignment @s.i = i@ means @s.i = s.i@, which is meaningless, and there is no mechanism to qualify the parameter @i@, making the assignment impossible using the function-body @with@.
1496To solve this problem, parameters are treated like an initialized aggregate:
1497\begin{cfa}
1498struct Params {
1499        S & s;
1500        int i;
1501} params;
1502\end{cfa}
1503and implicitly opened \emph{after} a function-body open, to give them higher priority:
1504\begin{cfa}
1505void ?{}( S & s, int `i` ) with ( s ) `with( $\emph{\color{red}params}$ )` {
1506        s.i = `i`; j = 3; m = 5.5;
1507}
1508\end{cfa}
1509Finally, a cast may be used to disambiguate among overload variables in a @with@ expression:
1510\begin{cfa}
1511with ( w ) { ... }                                                      $\C{// ambiguous, same name and no context}$
1512with ( (S)w ) { ... }                                           $\C{// unambiguous, cast}$
1513\end{cfa}
1514and @with@ expressions may be complex expressions with type reference (see Section~\ref{s:References}) to aggregate:
1515\begin{cfa}
1516struct S { int i, j; } sv;
1517with ( sv ) {                                                           $\C{implicit reference}$
1518        S & sr = sv;
1519        with ( sr ) {                                                   $\C{explicit reference}$
1520                S * sp = &sv;
1521                with ( *sp ) {                                          $\C{computed reference}$
1522                        i = 3; j = 4;                                   $\C{\color{red}// sp-{\textgreater}i, sp-{\textgreater}j}$
1523                }
1524                i = 2; j = 3;                                           $\C{\color{red}// sr.i, sr.j}$
1525        }
1526        i = 1; j = 2;                                                   $\C{\color{red}// sv.i, sv.j}$
1527}
1528\end{cfa}
1529
1530
1531\section{Declarations}
1532
1533Declarations in C have weaknesses and omissions.
1534\CFA attempts to correct and add to C declarations, while ensuring \CFA subjectively ``feels like'' C.
1535An important part of this subjective feel is maintaining C's syntax and procedural paradigm, as opposed to functional and object-oriented approaches in other systems languages such as \CC and Rust.
1536Maintaining the C approach means that C coding-patterns remain not only useable but idiomatic in \CFA, reducing the mental burden of retraining C programmers and switching between C and \CFA development.
1537Nevertheless, some features from other approaches are undeniably convenient;
1538\CFA attempts to adapt these features to the C paradigm.
1539
1540
1541\subsection{Alternative Declaration Syntax}
1542
1543C declaration syntax is notoriously confusing and error prone.
1544For example, many C programmers are confused by a declaration as simple as:
1545\begin{cquote}
1546\lstDeleteShortInline@%
1547\begin{tabular}{@{}ll@{}}
1548\begin{cfa}
1549int * x[5]
1550\end{cfa}
1551&
1552\raisebox{-0.75\totalheight}{\input{Cdecl}}
1553\end{tabular}
1554\lstMakeShortInline@%
1555\end{cquote}
1556Is this an array of 5 pointers to integers or a pointer to an array of 5 integers?
1557If there is any doubt, it implies productivity and safety issues even for basic programs.
1558Another example of confusion results from the fact that a function name and its parameters are embedded within the return type, mimicking the way the return value is used at the function's call site.
1559For example, a function returning a pointer to an array of integers is defined and used in the following way:
1560\begin{cfa}
1561int `(*`f`())[`5`]` {...};                                      $\C{// definition}$
1562 ... `(*`f`())[`3`]` += 1;                                      $\C{// usage}$
1563\end{cfa}
1564Essentially, the return type is wrapped around the function name in successive layers (like an onion).
1565While attempting to make the two contexts consistent is a laudable goal, it has not worked out in practice.
1566
1567\CFA provides its own type, variable and function declarations, using a different syntax~\cite[pp.~856--859]{Buhr94a}.
1568The new declarations place qualifiers to the left of the base type, while C declarations place qualifiers to the right.
1569The qualifiers have the same meaning but are ordered left to right to specify a variable's type.
1570\begin{cquote}
1571\lstDeleteShortInline@%
1572\begin{tabular}{@{}l@{\hspace{\parindentlnth}}l@{\hspace{\parindentlnth}}l@{}}
1573\multicolumn{1}{c@{\hspace{\parindentlnth}}}{\textbf{\CFA}}     & \multicolumn{1}{c}{\textbf{C}}        \\
1574\begin{cfa}
1575`[5] *` int x1;
1576`* [5]` int x2;
1577`[* [5] int]` f( int p );
1578\end{cfa}
1579&
1580\begin{cfa}
1581int `*` x1 `[5]`;
1582int `(*`x2`)[5]`;
1583`int (*`f( int p )`)[5]`;
1584\end{cfa}
1585&
1586\begin{cfa}
1587// array of 5 pointers to int
1588// pointer to an array of 5 int
1589// function returning pointer to an array of 5 int and taking an int
1590\end{cfa}
1591\end{tabular}
1592\lstMakeShortInline@%
1593\end{cquote}
1594The only exception is bit field specification, which always appear to the right of the base type.
1595% 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 function parameter.
1596However, unlike C, \CFA type declaration tokens are distributed across all variables in the declaration list.
1597For instance, variables @x@ and @y@ of type pointer to integer are defined in \CFA as follows:
1598\begin{cquote}
1599\lstDeleteShortInline@%
1600\begin{tabular}{@{}l@{\hspace{\parindentlnth}}l@{}}
1601\multicolumn{1}{c@{\hspace{\parindentlnth}}}{\textbf{\CFA}}     & \multicolumn{1}{c}{\textbf{C}}        \\
1602\begin{cfa}
1603`*` int x, y;
1604int y;
1605\end{cfa}
1606&
1607\begin{cfa}
1608int `*`x, `*`y;
1609
1610\end{cfa}
1611\end{tabular}
1612\lstMakeShortInline@%
1613\end{cquote}
1614The downside of the \CFA semantics is the need to separate regular and pointer declarations.
1615
1616\begin{comment}
1617Other examples are:
1618\begin{cquote}
1619\lstDeleteShortInline@%
1620\begin{tabular}{@{}l@{\hspace{\parindentlnth}}l@{\hspace{\parindentlnth}}l@{}}
1621\multicolumn{1}{c@{\hspace{\parindentlnth}}}{\textbf{\CFA}}     & \multicolumn{1}{c@{\hspace{\parindentlnth}}}{\textbf{C}}      \\
1622\begin{cfa}
1623[ 5 ] int z;
1624[ 5 ] * char w;
1625* [ 5 ] double v;
1626struct s {
1627        int f0:3;
1628        * int f1;
1629        [ 5 ] * int f2;
1630};
1631\end{cfa}
1632&
1633\begin{cfa}
1634int z[ 5 ];
1635char * w[ 5 ];
1636double (* v)[ 5 ];
1637struct s {
1638        int f0:3;
1639        int * f1;
1640        int * f2[ 5 ]
1641};
1642\end{cfa}
1643&
1644\begin{cfa}
1645// array of 5 integers
1646// array of 5 pointers to char
1647// pointer to array of 5 doubles
1648
1649// common bit field syntax
1650
1651
1652
1653\end{cfa}
1654\end{tabular}
1655\lstMakeShortInline@%
1656\end{cquote}
1657\end{comment}
1658
1659All specifiers (@extern@, @static@, \etc) and qualifiers (@const@, @volatile@, \etc) are used in the normal way with the new declarations and also appear left to right, \eg:
1660\begin{cquote}
1661\lstDeleteShortInline@%
1662\begin{tabular}{@{}l@{\hspace{1em}}l@{\hspace{1em}}l@{}}
1663\multicolumn{1}{c@{\hspace{1em}}}{\textbf{\CFA}}        & \multicolumn{1}{c@{\hspace{1em}}}{\textbf{C}} \\
1664\begin{cfa}
1665extern const * const int x;
1666static const * [ 5 ] const int y;
1667\end{cfa}
1668&
1669\begin{cfa}
1670int extern const * const x;
1671static const int (* const y)[ 5 ]
1672\end{cfa}
1673&
1674\begin{cfa}
1675// external const pointer to const int
1676// internal const pointer to array of 5 const int
1677\end{cfa}
1678\end{tabular}
1679\lstMakeShortInline@%
1680\end{cquote}
1681Specifiers must appear at the start of a \CFA function declaration\footnote{\label{StorageClassSpecifier}
1682The 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}}.
1683
1684The new declaration syntax can be used in other contexts where types are required, \eg casts and the pseudo-function @sizeof@:
1685\begin{cquote}
1686\lstDeleteShortInline@%
1687\begin{tabular}{@{}l@{\hspace{\parindentlnth}}l@{}}
1688\multicolumn{1}{c@{\hspace{\parindentlnth}}}{\textbf{\CFA}}     & \multicolumn{1}{c}{\textbf{C}}        \\
1689\begin{cfa}
1690y = (* int)x;
1691i = sizeof([ 5 ] * int);
1692\end{cfa}
1693&
1694\begin{cfa}
1695y = (int *)x;
1696i = sizeof(int * [ 5 ]);
1697\end{cfa}
1698\end{tabular}
1699\lstMakeShortInline@%
1700\end{cquote}
1701
1702The syntax of the new function-prototype declaration follows directly from the new function-definition syntax;
1703as well, parameter names are optional, \eg:
1704\begin{cfa}
1705[ int x ] f ( /* void */ );                                     $\C{// returning int with no parameters}$
1706[ int x ] f (...);                                                      $\C{// returning int with unknown parameters}$
1707[ * int ] g ( int y );                                          $\C{// returning pointer to int with int parameter}$
1708[ void ] h ( int, char );                                       $\C{// returning no result with int and char parameters}$
1709[ * int, int ] j ( int );                                       $\C{// returning pointer to int and int, with int parameter}$
1710\end{cfa}
1711This syntax allows a prototype declaration to be created by cutting and pasting source text from the function-definition header (or vice versa).
1712Like C, it is possible to declare multiple function-prototypes in a single declaration, where the return type is distributed across \emph{all} function names in the declaration list, \eg:
1713\begin{cquote}
1714\lstDeleteShortInline@%
1715\begin{tabular}{@{}l@{\hspace{\parindentlnth}}l@{}}
1716\multicolumn{1}{c@{\hspace{\parindentlnth}}}{\textbf{\CFA}}     & \multicolumn{1}{c}{\textbf{C}}        \\
1717\begin{cfa}
1718[double] foo(), foo( int ), foo( double ) {...}
1719\end{cfa}
1720&
1721\begin{cfa}
1722double foo1( void ), foo2( int ), foo3( double );
1723\end{cfa}
1724\end{tabular}
1725\lstMakeShortInline@%
1726\end{cquote}
1727where \CFA allows the last function in the list to define its body.
1728
1729The syntax for pointers to \CFA functions specifies the pointer name on the right, \eg:
1730\begin{cfa}
1731* [ int x ] () fp;                                                      $\C{// pointer to function returning int with no parameters}$
1732* [ * int ] ( int y ) gp;                                       $\C{// pointer to function returning pointer to int with int parameter}$
1733* [ ] ( int, char ) hp;                                         $\C{// pointer to function returning no result with int and char parameters}$
1734* [ * int, int ] ( int ) jp;                            $\C{// pointer to function returning pointer to int and int, with int parameter}$
1735\end{cfa}
1736Note, a function name cannot be specified:
1737\begin{cfa}
1738* [ int x ] f () fp;                                            $\C{// function name "f" is disallowed}$
1739\end{cfa}
1740
1741Finally, new \CFA declarations may appear together with C declarations in the same program block, but cannot be mixed within a specific declaration.
1742Therefore, a programmer has the option of either continuing to use traditional C declarations or take advantage of the new style.
1743Clearly, both styles need to be supported for some time due to existing C-style header-files, particularly for UNIX-like systems.
1744
1745
1746\subsection{References}
1747\label{s:References}
1748
1749All variables in C have an \newterm{address}, a \newterm{value}, and a \newterm{type};
1750at 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.
1751The C type-system does not always track the relationship between a value and its address;
1752a value that does not have a corresponding address is called a \newterm{rvalue} (for ``right-hand value''), while a value that does have an address is called a \newterm{lvalue} (for ``left-hand value'').
1753For example, 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.
1754Despite the nomenclature of ``left-hand'' and ``right-hand'', an expression's classification as lvalue or rvalue is entirely dependent on whether it has an address or not; in imperative programming, the address of a value is used for both reading and writing (mutating) a value, and as such, lvalues can be converted to rvalues and read from, but rvalues cannot be mutated because they lack a location to store the updated value.
1755
1756Within a lexical scope, lvalue expressions have an \newterm{address interpretation} for writing a value or a \newterm{value interpretation} to read a value.
1757For example, in @x = y@, @x@ has an address interpretation, while @y@ has a value interpretation.
1758While this duality of interpretation is useful, C lacks a direct mechanism to pass lvalues between contexts, instead relying on \newterm{pointer types} to serve a similar purpose.
1759In C, for any type @T@ there is a pointer type @T *@, the value of which is the address of a value of type @T@.
1760A pointer rvalue can be explicitly \newterm{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 @&?@.
1761
1762\begin{cfa}
1763int x = 1, y = 2, * p1, * p2, ** p3;
1764p1 = &x;                                                                        $\C{// p1 points to x}$
1765p2 = &y;                                                                        $\C{// p2 points to y}$
1766p3 = &p1;                                                                       $\C{// p3 points to p1}$
1767*p2 = ((*p1 + *p2) * (**p3 - *p1)) / (**p3 - 15);
1768\end{cfa}
1769
1770Unfortunately, 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 because of pointer arithmetic.
1771For both brevity and clarity, it is desirable for the compiler to figure out how to elide the dereference operators in a complex expression such as the assignment to @*p2@ above.
1772However, since C defines a number of forms of \newterm{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 programmer may write one when they mean the other, and precluding any simple algorithm for elision of dereference operators.
1773To solve these problems, \CFA introduces reference types @T &@;
1774a @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:
1775
1776\begin{cfa}
1777int x = 1, y = 2, & r1, & r2, && r3;
1778&r1 = &x;                                                                       $\C{// r1 points to x}$
1779&r2 = &y;                                                                       $\C{// r2 points to y}$
1780&&r3 = &&r1;                                                            $\C{// r3 points to r2}$
1781r2 = ((r1 + r2) * (r3 - r1)) / (r3 - 15);       $\C{// implicit dereferencing}$
1782\end{cfa}
1783
1784Except for auto-dereferencing by the compiler, this reference example is exactly the same as the previous pointer example.
1785Hence, 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.
1786One 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:
1787
1788\begin{cfa}
1789`*`r2 = ((`*`r1 + `*`r2) * (`**`r3 - `*`r1)) / (`**`r3 - 15);
1790\end{cfa}
1791
1792References in \CFA are similar to those in \CC, with important improvements, which can be seen in the example above.
1793Firstly, \CFA does not forbid references to references.
1794This provides a much more orthogonal design for library implementors, obviating the need for workarounds such as @std::reference_wrapper@.
1795Secondly, \CFA references are rebindable, whereas \CC references have a fixed address.
1796\newsavebox{\LstBox}
1797\begin{lrbox}{\LstBox}
1798\lstset{basicstyle=\footnotesize\linespread{0.9}\sf}
1799\begin{cfa}
1800int & r = *new( int );
1801...                                                                                     $\C{// non-null reference}$
1802delete &r;
1803r += 1;                                                                         $\C{// undefined reference}$
1804\end{cfa}
1805\end{lrbox}
1806Rebinding allows \CFA references to be default-initialized (\eg to a null pointer\footnote{
1807While effort has been made into non-null reference checking in \CC and Java, the exercise seems moot for any non-managed languages (C/\CC), given that it only handles one of many different error situations:
1808\begin{cquote}
1809\usebox{\LstBox}
1810\end{cquote}
1811}%
1812) and point to different addresses throughout their lifetime, like pointers.
1813Rebinding is accomplished by extending the existing syntax and semantics of the address-of operator in C.
1814
1815In 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.
1816In \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.
1817The 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.
1818This rebinding occurs to an arbitrary depth of reference nesting;
1819loosely speaking, nested address-of operators produce a nested lvalue pointer up to the depth of the reference.
1820These 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@.
1821More precisely:
1822\begin{itemize}
1823\item
1824if @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).
1825       
1826\item
1827if @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).
1828\end{itemize}
1829Since 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 is made solely on convenience, \eg many pointer or value accesses.
1830
1831By analogy to pointers, \CFA references also allow cv-qualifiers such as @const@:
1832\begin{cfa}
1833const int cx = 5;                                                       $\C{// cannot change cx}$
1834const int & cr = cx;                                            $\C{// cannot change cr's referred value}$
1835&cr = &cx;                                                                      $\C{// rebinding cr allowed}$
1836cr = 7;                                                                         $\C{// ERROR, cannot change cr}$
1837int & const rc = x;                                                     $\C{// must be initialized, like in \CC}$
1838&rc = &x;                                                                       $\C{// ERROR, cannot rebind rc}$
1839rc = 7;                                                                         $\C{// x now equal to 7}$
1840\end{cfa}
1841Given that a reference is meant to represent a lvalue, \CFA provides some syntactic shortcuts when initializing references.
1842There are three initialization contexts in \CFA: declaration initialization, argument/parameter binding, and return/temporary binding.
1843In each of these contexts, the address-of operator on the target lvalue is elided.
1844The syntactic motivation 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@.
1845
1846More 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;
1847this conversion is used in any context in \CFA where an implicit conversion is allowed.
1848Similarly, 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, analogous to the @T *@ to @const T *@ conversion in standard C.
1849The final reference conversion included in \CFA is ``rvalue-to-reference'' conversion, implemented by means of an implicit temporary.
1850When 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.
1851\begin{cfa}
1852struct S { double x, y; };
1853int i, j;
1854void f( int & i, int & j, S & s, int v[] );
1855f( 3, i + j, (S){ 1.0, 7.0 }, (int [3]){ 1, 2, 3 } );   $\C{// pass rvalue to lvalue \(\Rightarrow\) implicit temporary}$
1856\end{cfa}
1857This 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.
1858\CC allows a similar binding, but only for @const@ references; the more general semantics of \CFA are an attempt to avoid the \newterm{const hell} problem, in which addition of a @const@ qualifier to one reference requires a cascading chain of added qualifiers.
1859
1860
1861\subsection{Type Nesting}
1862
1863Nested types provide a mechanism to organize associated types and refactor a subset of fields into a named aggregate (\eg sub-aggregates @name@, @address@, @department@, within aggregate @employe@).
1864Java nested types are dynamic (apply to objects), \CC are static (apply to the \lstinline[language=C++]@class@), and C hoists (refactors) nested types into the enclosing scope, meaning there is no need for type qualification.
1865Since \CFA in not object-oriented, adopting dynamic scoping does not make sense;
1866instead \CFA adopts \CC static nesting, using the field-selection operator ``@.@'' for type qualification, as does Java, rather than the \CC type-selection operator ``@::@'' (see Figure~\ref{f:TypeNestingQualification}).
1867\begin{figure}
1868\centering
1869\lstDeleteShortInline@%
1870\begin{tabular}{@{}l@{\hspace{3em}}l|l@{}}
1871\multicolumn{1}{c@{\hspace{3em}}}{\textbf{C Type Nesting}}      & \multicolumn{1}{c}{\textbf{C Implicit Hoisting}}      & \multicolumn{1}{|c}{\textbf{\CFA}}    \\
1872\hline
1873\begin{cfa}
1874struct S {
1875        enum C { R, G, B };
1876        struct T {
1877                union U { int i, j; };
1878                enum C c;
1879                short int i, j;
1880        };
1881        struct T t;
1882} s;
1883
1884int rtn() {
1885        s.t.c = R;
1886        struct T t = { R, 1, 2 };
1887        enum C c;
1888        union U u;
1889}
1890\end{cfa}
1891&
1892\begin{cfa}
1893enum C { R, G, B };
1894union U { int i, j; };
1895struct T {
1896        enum C c;
1897        short int i, j;
1898};
1899struct S {
1900        struct T t;
1901} s;
1902       
1903
1904
1905
1906
1907
1908
1909\end{cfa}
1910&
1911\begin{cfa}
1912struct S {
1913        enum C { R, G, B };
1914        struct T {
1915                union U { int i, j; };
1916                enum C c;
1917                short int i, j;
1918        };
1919        struct T t;
1920} s;
1921
1922int rtn() {
1923        s.t.c = `S.`R;  // type qualification
1924        struct `S.`T t = { `S.`R, 1, 2 };
1925        enum `S.`C c;
1926        union `S.T.`U u;
1927}
1928\end{cfa}
1929\end{tabular}
1930\lstMakeShortInline@%
1931\caption{Type Nesting / Qualification}
1932\label{f:TypeNestingQualification}
1933\end{figure}
1934In the C left example, types @C@, @U@ and @T@ are implicitly hoisted outside of type @S@ into the containing block scope.
1935In the \CFA right example, the types are not hoisted and accessible.
1936
1937
1938\subsection{Constructors and Destructors}
1939\label{s:ConstructorsDestructors}
1940
1941One of the strengths (and weaknesses) of C is memory-management control, allowing resource release to be precisely specified versus unknown release with garbage-collected memory-management.
1942However, this manual approach is often verbose, and it is useful to manage resources other than memory (\eg file handles) using the same mechanism as memory.
1943\CC addresses these issues using Resource Aquisition Is Initialization (RAII), implemented by means of \newterm{constructor} and \newterm{destructor} functions;
1944\CFA adopts constructors and destructors (and @finally@) to facilitate RAII.
1945While constructors and destructors are a common feature of object-oriented programming-languages, they are an independent capability allowing \CFA to adopt them while retaining a procedural paradigm.
1946Specifically, \CFA constructors and destructors are denoted by name and first parameter-type versus name and nesting in an aggregate type.
1947Constructor calls seamlessly integrate with existing C initialization syntax, providing a simple and familiar syntax to C programmers and allowing constructor calls to be inserted into legacy C code with minimal code changes.
1948
1949In \CFA, a constructor is named @?{}@ and a destructor is named @^?{}@.
1950The name @{}@ comes from the syntax for the initializer: @struct S { int i, j; } s = `{` 2, 3 `}`@\footnote{%
1951The symbol \lstinline+^+ is used for the destructor name because it was the last binary operator that could be used in a unary context.}.
1952Like other \CFA operators, these names represent the syntax used to call the constructor or destructor, \eg @?{}(x, ...)@ or @^{}(x, ...)@.
1953The constructor and destructor have return type @void@, and the first parameter is a reference to the object type to be constructed or destructed.
1954While the first parameter is informally called the @this@ parameter, as in object-oriented languages, any variable name may be used.
1955Both constructors and destructors allow additional parametes after the @this@ parameter for specifying values for initialization/de-initialization\footnote{
1956Destruction parameters are useful for specifying storage-management actions, such as de-initialize but not deallocate.}.
1957\begin{cfa}
1958struct VLA { int len, * data; };
1959void ?{}( VLA & vla ) with ( vla ) { len = 10;  data = alloc( len ); }  $\C{// default constructor}$
1960void ^?{}( VLA & vla ) with ( vla ) { free( data ); } $\C{// destructor}$
1961{
1962        VLA x;                                                                  $\C{// implicit:  ?\{\}( x );}$
1963}                                                                                       $\C{// implicit:  ?\^{}\{\}( x );}$
1964\end{cfa}
1965@VLA@ is a \newterm{managed type}\footnote{
1966A managed type affects the runtime environment versus a self-contained type.}: a type requiring a non-trivial constructor or destructor, or with a field of a managed type.
1967A managed type is implicitly constructed at allocation and destructed at deallocation to ensure proper interaction with runtime resources, in this case, the @data@ array in the heap.
1968For details of the code-generation placement of implicit constructor and destructor calls among complex executable statements see~\cite[\S~2.2]{Schluntz17}.
1969
1970\CFA also provides syntax for \newterm{initialization} and \newterm{copy}:
1971\begin{cfa}
1972void ?{}( VLA & vla, int size, char fill ) with ( vla ) {  $\C{// initialization}$
1973        len = size;  data = alloc( len, fill );
1974}
1975void ?{}( VLA & vla, VLA other ) {                      $\C{// copy, shallow}$
1976        vla.len = other.len;  vla.data = other.data;
1977}
1978\end{cfa}
1979(Note, the example is purposely simplified using shallow-copy semantics.)
1980An initialization constructor-call has the same syntax as a C initializer, except the initialization values are passed as arguments to a matching constructor (number and type of paremeters).
1981\begin{cfa}
1982VLA va = `{` 20, 0 `}`,  * arr = alloc()`{` 5, 0 `}`;
1983\end{cfa}
1984Note, the use of a \newterm{constructor expression} to initialize the storage from the dynamic storage-allocation.
1985Like \CC, the copy constructor has two parameters, the second of which is a value parameter with the same type as the first parameter;
1986appropriate care is taken to not recursively call the copy constructor when initializing the second parameter.
1987
1988\CFA constructors may be explicitly call, like Java, and destructors may be explicitly called, like \CC.
1989Explicit calls to constructors double as a \CC-style \emph{placement syntax}, useful for construction of member fields in user-defined constructors and reuse of existing storage allocations.
1990While existing call syntax works for explicit calls to constructors and destructors, \CFA also provides a more concise \newterm{operator syntax} for both:
1991\begin{cfa}
1992{
1993        VLA  x,            y = { 20, 0x01 },     z = y; $\C{// z points to y}$
1994        //      ?{}( x );  ?{}( y, 20, 0x01 );  ?{}( z, y );
1995        ^x{};                                                                   $\C{// deallocate x}$
1996        x{};                                                                    $\C{// reallocate x}$
1997        z{ 5, 0xff };                                                   $\C{// reallocate z, not pointing to y}$
1998        ^y{};                                                                   $\C{// deallocate y}$
1999        y{ x };                                                                 $\C{// reallocate y, points to x}$
2000        x{};                                                                    $\C{// reallocate x, not pointing to y}$
2001        // ^?{}(z);  ^?{}(y);  ^?{}(x);
2002}
2003\end{cfa}
2004
2005To 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.
2006These default functions can be overridden by user-generated versions.
2007For 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;
2008if default zero-initialization is desired, the default constructors can be overridden.
2009For user-generated types, the four functions are also automatically generated.
2010@enum@ types are handled the same as their underlying integral type, and unions are also bitwise copied and no-op initialized and destructed.
2011For compatibility with C, a copy constructor from the first union member type is also defined.
2012For @struct@ types, each of the four functions are implicitly defined to call their corresponding functions on each member of the struct.
2013To better simulate the behaviour of C initializers, a set of \newterm{field constructors} is also generated for structures.
2014A constructor is generated for each non-empty prefix of a structure's member-list to copy-construct the members passed as parameters and default-construct the remaining members.
2015To 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;
2016similarly, the generated default constructor is hidden upon declaration of any constructor.
2017These semantics closely mirror the rule for implicit declaration of constructors in \CC\cite[p.~186]{ANSI98:C++}.
2018
2019In some circumstance programmers may not wish to have constructor and destructor calls.
2020In these cases, \CFA provides the initialization syntax \lstinline|S x @= {}|, and the object becomes unmanaged, so implicit constructor and destructor calls are not generated.
2021Any C initializer can be the right-hand side of an \lstinline|@=| initializer, \eg \lstinline|VLA a @= { 0, 0x0 }|, with the usual C initialization semantics.
2022The point of \lstinline|@=| is to provide a migration path from legacy C code to \CFA, by providing a mechanism to incrementally convert to implicit initialization.
2023
2024
2025% \subsection{Default Parameters}
2026
2027
2028\section{Literals}
2029
2030C already includes limited polymorphism for literals -- @0@ can be either an integer or a pointer literal, depending on context, while the syntactic forms of literals of the various integer and float types are very similar, differing from each other only in suffix.
2031In keeping with the general \CFA approach of adding features while respecting the ``C-style'' of doing things, C's polymorphic constants and typed literal syntax are extended to interoperate with user-defined types, while maintaining a backwards-compatible semantics.
2032
2033A simple example is allowing the underscore, as in Ada, to separate prefixes, digits, and suffixes in all \CFA constants, \eg @0x`_`1.ffff`_`ffff`_`p`_`128`_`l@, where the underscore is also the standard separator in C identifiers.
2034\CC uses a single quote as a separator but it is restricted among digits, precluding its use in the literal prefix or suffix, \eg @0x1.ffff@@`'@@ffffp128l@, and causes problems with most IDEs, which must be extended to deal with this alternate use of the single quote.
2035
2036
2037\subsection{Integral Suffixes}
2038
2039Additional integral suffixes are added to cover all the integral types and lengths.
2040\begin{cquote}
2041\lstDeleteShortInline@%
2042\begin{tabular}{@{}l@{\hspace{\parindentlnth}}l@{\hspace{\parindentlnth}}l@{}}
2043\begin{cfa}
204420_`hh`     // signed char
204521_`hhu`   // unsigned char
204622_`h`       // signed short int
204723_`uh`     // unsigned short int
204824_`z`       // size_t
2049\end{cfa}
2050&
2051\begin{cfa}
205220_`L8`      // int8_t
205321_`ul8`     // uint8_t
205422_`l16`     // int16_t
205523_`ul16`   // uint16_t
205624_`l32`     // int32_t
2057\end{cfa}
2058&
2059\begin{cfa}
206025_`ul32`      // uint32_t
206126_`l64`        // int64_t
206227_`l64u`      // uint64_t
206326_`L128`     // int128
206427_`L128u`   // unsigned int128
2065\end{cfa}
2066\end{tabular}
2067\lstMakeShortInline@%
2068\end{cquote}
2069
2070
2071\subsection{0/1}
2072
2073In C, @0@ has the special property that it is the only ``false'' value; by the standard, any value which compares equal to @0@ is false, while any value that compares unequal to @0@ is true.
2074As such, an expression @x@ in any boolean context (such as the condition of an @if@ or @while@ statement, or the arguments to @&&@, @||@, or @?:@) can be rewritten as @x != 0@ without changing its semantics.
2075The operator overloading feature of \CFA provides a natural means to implement this truth value comparison for arbitrary types, but the C type system is not precise enough to distinguish an equality comparison with @0@ from an equality comparison with an arbitrary integer or pointer.
2076To provide this precision, \CFA introduces a new type @zero_t@ as type type of literal @0@ (somewhat analagous to @nullptr_t@ and @nullptr@ in \CCeleven); @zero_t@ can only take the value @0@, but has implicit conversions to the integer and pointer types so that C code involving @0@ continues to work properly.
2077With this addition, the \CFA compiler rewrites @if (x)@ and similar expressions to @if ((x) != 0)@ or the appropriate analogue, and any type @T@ can be made ``truthy'' by defining an operator overload @int ?!=?(T, zero_t)@.
2078\CC makes types truthy by adding a conversion to @bool@; prior to the addition of explicit cast operators in \CCeleven this approach had the pitfall of making truthy types transitively convertable to any numeric type; our design for \CFA avoids this issue.
2079
2080\CFA also includes a special type for @1@, @one_t@; like @zero_t@, @one_t@ has built-in implicit conversions to the various integral types so that @1@ maintains its expected semantics in legacy code.
2081The addition of @one_t@ allows generic algorithms to handle the unit value uniformly for types where that is meaningful.
2082\TODO{Make this sentence true} In particular, polymorphic functions in the \CFA prelude define @++x@ and @x++@ in terms of @x += 1@, allowing users to idiomatically define all forms of increment for a type @T@ by defining the single function @T & ?+=(T &, one_t)@; analogous overloads for the decrement operators are present as well.
2083
2084
2085\subsection{User Literals}
2086
2087For readability, it is useful to associate units to scale literals, \eg weight (stone, pound, kilogram) or time (seconds, minutes, hours).
2088The left of Figure~\ref{f:UserLiteral} shows the \CFA alternative call-syntax (literal argument before function name), using the backquote, to convert basic literals into user literals.
2089The backquote is a small character, making the unit (function name) predominate.
2090For examples, the multi-precision integers in Section~\ref{s:MultiPrecisionIntegers} make use of user literals:
2091{\lstset{language=CFA,moredelim=**[is][\color{red}]{|}{|},deletedelim=**[is][]{`}{`}}
2092\begin{cfa}
2093y = 9223372036854775807L|`mp| * 18446744073709551615UL|`mp|;
2094y = "12345678901234567890123456789"|`mp| + "12345678901234567890123456789"|`mp|;
2095\end{cfa}
2096Because \CFA uses a standard function, all types and literals are applicable, as well as overloading and conversions.
2097}%
2098
2099The right of Figure~\ref{f:UserLiteral} shows the equivalent \CC version using the underscore for the call-syntax.
2100However, \CC restricts the types, \eg @unsigned long long int@ and @long double@ to represent integral and floating literals.
2101After which, user literals must match (no conversions);
2102hence, it is necessary to overload the unit with all appropriate types.
2103
2104\begin{figure}
2105\centering
2106\lstset{language=CFA,moredelim=**[is][\color{red}]{|}{|},deletedelim=**[is][]{`}{`}}
2107\lstDeleteShortInline@%
2108\begin{tabular}{@{}l@{\hspace{\parindentlnth}}l@{}}
2109\multicolumn{1}{c@{\hspace{\parindentlnth}}}{\textbf{\CFA}}     & \multicolumn{1}{c}{\textbf{\CC}}      \\
2110\begin{cfa}
2111struct W {
2112        double stones;
2113};
2114void ?{}( W & w ) { w.stones = 0; }
2115void ?{}( W & w, double w ) { w.stones = w; }
2116W ?+?( W l, W r ) {
2117        return (W){ l.stones + r.stones };
2118}
2119W |?`st|( double w ) { return (W){ w }; }
2120W |?`lb|( double w ) { return (W){ w / 14.0 }; }
2121W |?`kg|( double w ) { return (W) { w * 0.16 }; }
2122
2123
2124
2125int main() {
2126        W w, heavy = { 20 };
2127        w = 155|`lb|;
2128        w = 0b_1111|`st|;
2129        w = 0_233|`lb|;
2130        w = 0x_9b_u|`kg|;
2131        w = 5.5|`st| + 8|`kg| + 25.01|`lb| + heavy;
2132}
2133\end{cfa}
2134&
2135\begin{cfa}
2136struct W {
2137    double stones;
2138    W() { stones = 0.0; }
2139    W( double w ) { stones = w; }
2140};
2141W operator+( W l, W r ) {
2142        return W( l.stones + r.stones );
2143}
2144W |operator"" _st|( unsigned long long int w ) { return W( w ); }
2145W |operator"" _lb|( unsigned long long int w ) { return W( w / 14.0 ); }
2146W |operator"" _kg|( unsigned long long int w ) { return W( w * 0.16 ); }
2147W |operator"" _st|( long double w ) { return W( w ); }
2148W |operator"" _lb|( long double w ) { return W( w / 14.0 ); }
2149W |operator"" _kg|( long double w ) { return W( w * 0.16 ); }
2150int main() {
2151        W w, heavy = { 20 };
2152        w = 155|_lb|;
2153        w = 0b1111|_lb|;       // error, binary unsupported
2154        w = 0${\color{red}\LstBasicStyle{'}}$233|_lb|;          // quote separator
2155        w = 0x9b|_kg|;
2156        w = 5.5d|_st| + 8|_kg| + 25.01|_lb| + heavy;
2157}
2158\end{cfa}
2159\end{tabular}
2160\lstMakeShortInline@%
2161\caption{User Literal}
2162\label{f:UserLiteral}
2163\end{figure}
2164
2165
2166\section{Libraries}
2167\label{sec:libraries}
2168
2169As stated in Section~\ref{sec:poly-fns}, \CFA inherits a large corpus of library code, where other programming languages must rewrite or provide fragile inter-language communication with C.
2170\CFA has replacement libraries condensing hundreds of existing C names into tens of \CFA overloaded names, all without rewriting the actual computations.
2171In many cases, the interface is an inline wrapper providing overloading during compilation but zero cost at runtime.
2172The following sections give a glimpse of the interface reduction to many C libraries.
2173In many cases, @signed@/@unsigned@ @char@, @short@, and @_Complex@ functions are available (but not shown) to ensure expression computations remain in a single type, as conversions can distort results.
2174
2175
2176\subsection{Limits}
2177
2178C library @limits.h@ provides lower and upper bound constants for the basic types.
2179\CFA name overloading is used to condense these typed constants, \eg:
2180\begin{cquote}
2181\lstDeleteShortInline@%
2182\begin{tabular}{@{}l@{\hspace{\parindentlnth}}l@{}}
2183\multicolumn{1}{c@{\hspace{\parindentlnth}}}{\textbf{Definition}}       & \multicolumn{1}{c}{\textbf{Usage}}    \\
2184\begin{cfa}
2185const short int `MIN` = -32768;
2186const int `MIN` = -2147483648;
2187const long int `MIN` = -9223372036854775808L;
2188\end{cfa}
2189&
2190\begin{cfa}
2191short int si = `MIN`;
2192int i = `MIN`;
2193long int li = `MIN`;
2194\end{cfa}
2195\end{tabular}
2196\lstMakeShortInline@%
2197\end{cquote}
2198The result is a significant reduction in names to access typed constants, \eg:
2199\begin{cquote}
2200\lstDeleteShortInline@%
2201\lstset{basicstyle=\linespread{0.9}\sf\small}
2202\begin{tabular}{@{}l@{\hspace{0.5\parindentlnth}}l@{}}
2203\multicolumn{1}{c@{\hspace{0.5\parindentlnth}}}{\textbf{\CFA}}  & \multicolumn{1}{c}{\textbf{C}}        \\
2204\begin{cfa}
2205MIN
2206MAX
2207PI
2208E
2209\end{cfa}
2210&
2211\begin{cfa}
2212SCHAR_MIN, CHAR_MIN, SHRT_MIN, INT_MIN, LONG_MIN, LLONG_MIN, FLT_MIN, DBL_MIN, LDBL_MIN
2213SCHAR_MAX, UCHAR_MAX, SHRT_MAX, INT_MAX, LONG_MAX, LLONG_MAX, FLT_MAX, DBL_MAX, LDBL_MAX
2214M_PI, M_PIl
2215M_E, M_El
2216\end{cfa}
2217\end{tabular}
2218\lstMakeShortInline@%
2219\end{cquote}
2220
2221
2222\subsection{Math}
2223
2224C library @math.h@ provides many mathematical functions.
2225\CFA function overloading is used to condense these mathematical functions, \eg:
2226\begin{cquote}
2227\lstDeleteShortInline@%
2228\begin{tabular}{@{}l@{\hspace{\parindentlnth}}l@{}}
2229\multicolumn{1}{c@{\hspace{\parindentlnth}}}{\textbf{Definition}}       & \multicolumn{1}{c}{\textbf{Usage}}    \\
2230\begin{cfa}
2231float `log`( float x );
2232double `log`( double );
2233double _Complex `log`( double _Complex x );
2234\end{cfa}
2235&
2236\begin{cfa}
2237float f = `log`( 3.5 );
2238double d = `log`( 3.5 );
2239double _Complex dc = `log`( 3.5+0.5I );
2240\end{cfa}
2241\end{tabular}
2242\lstMakeShortInline@%
2243\end{cquote}
2244The result is a significant reduction in names to access math functions, \eg:
2245\begin{cquote}
2246\lstDeleteShortInline@%
2247\begin{tabular}{@{}l@{\hspace{\parindentlnth}}l@{}}
2248\multicolumn{1}{c@{\hspace{\parindentlnth}}}{\textbf{\CFA}}     & \multicolumn{1}{c}{\textbf{C}}        \\
2249\begin{cfa}
2250log
2251sqrt
2252sin
2253\end{cfa}
2254&
2255\begin{cfa}
2256logf, log, logl, clogf, clog, clogl
2257sqrtf, sqrt, sqrtl, csqrtf, csqrt, csqrtl
2258sinf, sin, sinl, csinf, csin, csinl
2259\end{cfa}
2260\end{tabular}
2261\lstMakeShortInline@%
2262\end{cquote}
2263While \Celeven has type-generic math~\cite[\S~7.25]{C11} in @tgmath.h@ to provide a similar mechanism, these macros are limited, matching a function name with a single set of floating type(s).
2264For example, it is impossible to overload @atan@ for both one and two arguments;
2265instead the names @atan@ and @atan2@ are required (see Section~\ref{s:NameOverloading}).
2266The key observation is that only a restricted set of type-generic macros are provided for a limited set of function names, which do not generalize across the type system, as in \CFA.
2267
2268
2269\subsection{Standard}
2270
2271C library @stdlib.h@ provides many general functions.
2272\CFA function overloading is used to condense these utility functions, \eg:
2273\begin{cquote}
2274\lstDeleteShortInline@%
2275\begin{tabular}{@{}l@{\hspace{\parindentlnth}}l@{}}
2276\multicolumn{1}{c@{\hspace{\parindentlnth}}}{\textbf{Definition}}       & \multicolumn{1}{c}{\textbf{Usage}}    \\
2277\begin{cfa}
2278unsigned int `abs`( int );
2279double `abs`( double );
2280double abs( double _Complex );
2281\end{cfa}
2282&
2283\begin{cfa}
2284unsigned int i = `abs`( -1 );
2285double d = `abs`( -1.5 );
2286double d = `abs`( -1.5+0.5I );
2287\end{cfa}
2288\end{tabular}
2289\lstMakeShortInline@%
2290\end{cquote}
2291The result is a significant reduction in names to access utility functions, \eg:
2292\begin{cquote}
2293\lstDeleteShortInline@%
2294\begin{tabular}{@{}l@{\hspace{\parindentlnth}}l@{}}
2295\multicolumn{1}{c@{\hspace{\parindentlnth}}}{\textbf{\CFA}}     & \multicolumn{1}{c}{\textbf{C}}        \\
2296\begin{cfa}
2297abs
2298strto
2299random
2300\end{cfa}
2301&
2302\begin{cfa}
2303abs, labs, llabs, fabsf, fabs, fabsl, cabsf, cabs, cabsl
2304strtol, strtoul, strtoll, strtoull, strtof, strtod, strtold
2305srand48, mrand48, lrand48, drand48
2306\end{cfa}
2307\end{tabular}
2308\lstMakeShortInline@%
2309\end{cquote}
2310In additon, there are polymorphic functions, like @min@ and @max@, which work on any type with operators @?<?@ or @?>?@.
2311
2312The following shows one example where \CFA \emph{extends} an existing standard C interface to reduce complexity and provide safety.
2313C/\Celeven provide a number of complex and overlapping storage-management operation to support the following capabilities:
2314\begin{description}[topsep=3pt,itemsep=2pt,parsep=0pt]
2315\item[fill]
2316an allocation with a specified character.
2317\item[resize]
2318an existing allocation to decreased or increased its size.
2319In either case, new storage may or may not be allocated and, if there is a new allocation, as much data from the existing allocation is copied.
2320For an increase in storage size, new storage after the copied data may be filled.
2321\item[alignment]
2322an allocation on a specified memory boundary, \eg, an address multiple of 64 or 128 for cache-line purposes.
2323\item[array]
2324allocation of the specified number of elements.
2325An array may be filled, resized, or aligned.
2326\end{description}
2327Table~\ref{t:StorageManagementOperations} shows the capabilities provided by C/\Celeven allocation-functions and how all the capabilities can be combined into two \CFA functions.
2328\CFA storage-management functions extend the C equivalents by overloading, providing shallow type-safety, and removing the need to specify the base allocation-size.
2329Figure~\ref{f:StorageAllocation} contrasts \CFA and C storage-allocation performing the same operations with the same type safety.
2330
2331\begin{table}
2332\centering
2333\lstDeleteShortInline@%
2334\lstMakeShortInline~%
2335\begin{tabular}{@{}r|r|l|l|l|l@{}}
2336\multicolumn{1}{c}{}&           & \multicolumn{1}{c|}{fill}     & resize        & alignment     & array \\
2337\hline
2338C               & ~malloc~                      & no                    & no            & no            & no    \\
2339                & ~calloc~                      & yes (0 only)  & no            & no            & yes   \\
2340                & ~realloc~                     & no/copy               & yes           & no            & no    \\
2341                & ~memalign~            & no                    & no            & yes           & no    \\
2342                & ~posix_memalign~      & no                    & no            & yes           & no    \\
2343\hline
2344C11             & ~aligned_alloc~       & no                    & no            & yes           & no    \\
2345\hline
2346\CFA    & ~alloc~                       & yes/copy              & no/yes        & no            & yes   \\
2347                & ~align_alloc~         & yes                   & no            & yes           & yes   \\
2348\end{tabular}
2349\lstDeleteShortInline~%
2350\lstMakeShortInline@%
2351\caption{Storage-Management Operations}
2352\label{t:StorageManagementOperations}
2353\end{table}
2354
2355\begin{figure}
2356\centering
2357\begin{cquote}
2358\begin{cfa}[aboveskip=0pt]
2359size_t  dim = 10;                                                       $\C{// array dimension}$
2360char fill = '\xff';                                                     $\C{// initialization fill value}$
2361int * ip;
2362\end{cfa}
2363\lstDeleteShortInline@%
2364\begin{tabular}{@{}l@{\hspace{\parindentlnth}}l@{}}
2365\multicolumn{1}{c@{\hspace{\parindentlnth}}}{\textbf{\CFA}}     & \multicolumn{1}{c}{\textbf{C}}        \\
2366\begin{cfa}
2367ip = alloc();
2368ip = alloc( fill );
2369ip = alloc( dim );
2370ip = alloc( dim, fill );
2371ip = alloc( ip, 2 * dim );
2372ip = alloc( ip, 4 * dim, fill );
2373
2374ip = align_alloc( 16 );
2375ip = align_alloc( 16, fill );
2376ip = align_alloc( 16, dim );
2377ip = align_alloc( 16, dim, fill );
2378\end{cfa}
2379&
2380\begin{cfa}
2381ip = (int *)malloc( sizeof( int ) );
2382ip = (int *)malloc( sizeof( int ) ); memset( ip, fill, sizeof( int ) );
2383ip = (int *)malloc( dim * sizeof( int ) );
2384ip = (int *)malloc( sizeof( int ) ); memset( ip, fill, dim * sizeof( int ) );
2385ip = (int *)realloc( ip, 2 * dim * sizeof( int ) );
2386ip = (int *)realloc( ip, 4 * dim * sizeof( int ) ); memset( ip, fill, 4 * dim * sizeof( int ) );
2387
2388ip = memalign( 16, sizeof( int ) );
2389ip = memalign( 16, sizeof( int ) ); memset( ip, fill, sizeof( int ) );
2390ip = memalign( 16, dim * sizeof( int ) );
2391ip = memalign( 16, dim * sizeof( int ) ); memset( ip, fill, dim * sizeof( int ) );
2392\end{cfa}
2393\end{tabular}
2394\lstMakeShortInline@%
2395\end{cquote}
2396\caption{\CFA versus C Storage-Allocation}
2397\label{f:StorageAllocation}
2398\end{figure}
2399
2400Variadic @new@ (see Section~\ref{sec:variadic-tuples}) cannot support the same overloading because extra parameters are for initialization.
2401Hence, there are @new@ and @anew@ functions for single and array variables, and the fill value is the arguments to the constructor, \eg:
2402\begin{cfa}
2403struct S { int i, j; };
2404void ?{}( S & s, int i, int j ) { s.i = i; s.j = j; }
2405S * s = new( 2, 3 );                                            $\C{// allocate storage and run constructor}$
2406S * as = anew( dim, 2, 3 );                                     $\C{// each array element initialized to 2, 3}$
2407\end{cfa}
2408Note, \CC can only initialization array elements via the default constructor.
2409
2410Finally, the \CFA memory-allocator has \newterm{sticky properties} for dynamic storage: fill and alignment are remembered with an object's storage in the heap.
2411When a @realloc@ is performed, the sticky properties are respected, so that new storage is correctly aligned and initialized with the fill character.
2412
2413
2414\subsection{I/O}
2415\label{s:IOLibrary}
2416
2417The goal of \CFA I/O is to simplify the common cases, while fully supporting polymorphism and user defined types in a consistent way.
2418The approach combines ideas from \CC and Python.
2419The \CFA header file for the I/O library is @fstream@.
2420
2421The common case is printing out a sequence of variables separated by whitespace.
2422\begin{cquote}
2423\lstDeleteShortInline@%
2424\begin{tabular}{@{}l@{\hspace{\parindentlnth}}l@{}}
2425\multicolumn{1}{c@{\hspace{\parindentlnth}}}{\textbf{\CFA}}     & \multicolumn{1}{c}{\textbf{\CC}}      \\
2426\begin{cfa}
2427int x = 1, y = 2, z = 3;
2428sout | x `|` y `|` z | endl;
2429\end{cfa}
2430&
2431\begin{cfa}
2432
2433cout << x `<< " "` << y `<< " "` << z << endl;
2434\end{cfa}
2435\\
2436\begin{cfa}[showspaces=true,aboveskip=0pt,belowskip=0pt]
24371` `2` `3
2438\end{cfa}
2439&
2440\begin{cfa}[showspaces=true,aboveskip=0pt,belowskip=0pt]
24411 2 3
2442\end{cfa}
2443\end{tabular}
2444\lstMakeShortInline@%
2445\end{cquote}
2446The \CFA form has half the characters of the \CC form, and is similar to Python I/O with respect to implicit separators.
2447Similar simplification occurs for tuple I/O, which prints all tuple values separated by ``\lstinline[showspaces=true]@, @''.
2448\begin{cfa}
2449[int, [ int, int ] ] t1 = [ 1, [ 2, 3 ] ], t2 = [ 4, [ 5, 6 ] ];
2450sout | t1 | t2 | endl;                                  $\C{// print tuples}$
2451\end{cfa}
2452\begin{cfa}[showspaces=true,aboveskip=0pt]
24531`, `2`, `3 4`, `5`, `6
2454\end{cfa}
2455Finally, \CFA uses the logical-or operator for I/O as it is the lowest-priority overloadable operator, other than assignment.
2456Therefore, fewer output expressions require parenthesis.
2457\begin{cquote}
2458\lstDeleteShortInline@%
2459\begin{tabular}{@{}ll@{}}
2460\textbf{\CFA:}
2461&
2462\begin{cfa}
2463sout | x * 3 | y + 1 | z << 2 | x == y | (x | y) | (x || y) | (x > z ? 1 : 2) | endl;
2464\end{cfa}
2465\\
2466\textbf{\CC:}
2467&
2468\begin{cfa}
2469cout << x * 3 << y + 1 << `(`z << 2`)` << `(`x == y`)` << (x | y) << (x || y) << (x > z ? 1 : 2) << endl;
2470\end{cfa}
2471\\
2472&
2473\begin{cfa}[showspaces=true,aboveskip=0pt]
24743 3 12 0 3 1 2
2475\end{cfa}
2476\end{tabular}
2477\lstMakeShortInline@%
2478\end{cquote}
2479There is a weak similarity between the \CFA logical-or operator and the Shell pipe-operator for moving data, where data flows in the correct direction for input but the opposite direction for output.
2480
2481The implicit separator character (space/blank) is a separator not a terminator.
2482The rules for implicitly adding the separator are:
2483\begin{itemize}
2484\item
2485A separator does not appear at the start or end of a line.
2486\item
2487A separator does not appear before or after a character literal or variable.
2488\item
2489A separator does not appear before or after a null (empty) C string, which is a local mechanism to disable insertion of the separator character.
2490\item
2491A separator does not appear before a C string starting with the characters: \lstinline[mathescape=off,basicstyle=\tt]@([{=$@
2492\item
2493A separator does not appear after a C string ending with the characters: \lstinline[basicstyle=\tt]@,.;!?)]}%@
2494\item
2495{\lstset{language=CFA,deletedelim=**[is][]{`}{`}}
2496A separator does not appear before or after a C string beginning/ending with the quote or whitespace characters: \lstinline[basicstyle=\tt,showspaces=true]@`'": \t\v\f\r\n@
2497}%
2498\end{itemize}
2499There are functions to set and get the separator string, and manipulators to toggle separation on and off in the middle of output.
2500
2501
2502\subsection{Multi-precision Integers}
2503\label{s:MultiPrecisionIntegers}
2504
2505\CFA has an interface to the GMP multi-precision signed-integers~\cite{GMP}, similar to the \CC interface provided by GMP.
2506The \CFA interface wraps GMP functions into operator functions to make programming with multi-precision integers identical to using fixed-sized integers.
2507The \CFA type name for multi-precision signed-integers is @Int@ and the header file is @gmp@.
2508The following multi-precision factorial programs contrast using GMP with the \CFA and C interfaces.
2509\begin{cquote}
2510\lstDeleteShortInline@%
2511\begin{tabular}{@{}l@{\hspace{\parindentlnth}}@{\hspace{\parindentlnth}}l@{}}
2512\multicolumn{1}{c@{\hspace{\parindentlnth}}}{\textbf{\CFA}}     & \multicolumn{1}{@{\hspace{\parindentlnth}}c}{\textbf{C}}      \\
2513\begin{cfa}
2514#include <gmp>
2515int main( void ) {
2516        sout | "Factorial Numbers" | endl;
2517        Int fact = 1;
2518        sout | 0 | fact | endl;
2519        for ( unsigned int i = 1; i <= 40; i += 1 ) {
2520                fact *= i;
2521                sout | i | fact | endl;
2522        }
2523}
2524\end{cfa}
2525&
2526\begin{cfa}
2527#include <gmp.h>
2528int main( void ) {
2529        `gmp_printf`( "Factorial Numbers\n" );
2530        `mpz_t` fact;  `mpz_init_set_ui`( fact, 1 );
2531        `gmp_printf`( "%d %Zd\n", 0, fact );
2532        for ( unsigned int i = 1; i <= 40; i += 1 ) {
2533                `mpz_mul_ui`( fact, fact, i );
2534                `gmp_printf`( "%d %Zd\n", i, fact );
2535        }
2536}
2537\end{cfa}
2538\end{tabular}
2539\lstMakeShortInline@%
2540\end{cquote}
2541
2542
2543\section{Evaluation}
2544\label{sec:eval}
2545
2546Though \CFA provides significant added functionality over C, these features have a low runtime penalty.
2547In fact, \CFA's features for generic programming can enable faster runtime execution than idiomatic @void *@-based C code.
2548This 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}).
2549Since 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.
2550A more illustrative benchmark measures the costs of idiomatic usage of each language's features.
2551Figure~\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@ function similar to that in Section~\ref{sec:variadic-tuples}.
2552The benchmark test is similar for C and \CC.
2553The 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.
2554
2555\begin{figure}
2556\begin{cfa}[xleftmargin=3\parindentlnth,aboveskip=0pt,belowskip=0pt]
2557int main( int argc, char * argv[] ) {
2558        int max = 0, val = 42;
2559        stack( int ) si, ti;
2560
2561        REPEAT_TIMED( "push_int", N, push( si, val ); )
2562        TIMED( "copy_int", ti = si; )
2563        TIMED( "clear_int", clear( si ); )
2564        REPEAT_TIMED( "pop_int", N,
2565                int x = pop( ti ); if ( x > max ) max = x; )
2566
2567        pair( _Bool, char ) max = { (_Bool)0, '\0' }, val = { (_Bool)1, 'a' };
2568        stack( pair( _Bool, char ) ) sp, tp;
2569
2570        REPEAT_TIMED( "push_pair", N, push( sp, val ); )
2571        TIMED( "copy_pair", tp = sp; )
2572        TIMED( "clear_pair", clear( sp ); )
2573        REPEAT_TIMED( "pop_pair", N,
2574                pair(_Bool, char) x = pop( tp ); if ( x > max ) max = x; )
2575}
2576\end{cfa}
2577\caption{\protect\CFA Benchmark Test}
2578\label{fig:BenchmarkTest}
2579\end{figure}
2580
2581The 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.
2582The \CCV variant illustrates an alternative object-oriented idiom where all objects inherit from a base @object@ class, mimicking a Java-like interface;
2583hence runtime checks are necessary to safely down-cast objects.
2584The 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.
2585For 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.
2586Note, the C benchmark uses unchecked casts as there is no runtime mechanism to perform such checks, while \CFA and \CC provide type-safety statically.
2587
2588Figure~\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.
2589The graph plots the median of 5 consecutive runs of each program, with an initial warm-up run omitted.
2590All code is compiled at \texttt{-O2} by gcc or g++ 6.2.0, with all \CC code compiled as \CCfourteen.
2591The 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.
2592
2593\begin{figure}
2594\centering
2595\input{timing}
2596\caption{Benchmark Timing Results (smaller is better)}
2597\label{fig:eval}
2598\end{figure}
2599
2600\begin{table}
2601\centering
2602\caption{Properties of benchmark code}
2603\label{tab:eval}
2604\newcommand{\CT}[1]{\multicolumn{1}{c}{#1}}
2605\begin{tabular}{rrrrr}
2606                                                                        & \CT{C}        & \CT{\CFA}     & \CT{\CC}      & \CT{\CCV}             \\ \hline
2607maximum memory usage (MB)                       & 10001         & 2502          & 2503          & 11253                 \\
2608source code size (lines)                        & 247           & 222           & 165           & 339                   \\
2609redundant type annotations (lines)      & 39            & 2                     & 2                     & 15                    \\
2610binary size (KB)                                        & 14            & 229           & 18            & 38                    \\
2611\end{tabular}
2612\end{table}
2613
2614The 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;
2615this inefficiency is exacerbated by the second level of generic types in the pair-based benchmarks.
2616By 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.
2617\CCV is slower than C largely due to the cost of runtime type-checking of down-casts (implemented with @dynamic_cast@);
2618There are two outliers in the graph for \CFA: all prints and pop of @pair@.
2619Both 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.
2620A compiler designed for \CFA could easily perform these optimizations.
2621Finally, the binary size for \CFA is larger because of static linking with the \CFA libraries.
2622
2623\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.
2624On 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.
2625\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;
2626with their omission, the \CCV line count is similar to C.
2627We 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.
2628
2629Raw line-count, however, is a fairly rough measure of code complexity;
2630another important factor is how much type information the programmer must manually specify, especially where that information is not checked by the compiler.
2631Such 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@).
2632To 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.
2633The \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.
2634The 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).
2635These uses are similar to the @new@ expressions in \CC, though the \CFA compiler's type resolver should shortly render even these type casts superfluous.
2636
2637
2638\section{Related Work}
2639
2640
2641\subsection{Polymorphism}
2642
2643\CC is the most similar language to \CFA;
2644both are extensions to C with source and runtime backwards compatibility.
2645The fundamental difference is the engineering approach to maintain C compatibility and programmer expectation.
2646While \CC provides good compatibility with C, it has a steep learning curve for many of its extensions.
2647For example, polymorphism is provided via three disjoint mechanisms: overloading, inheritance, and templates.
2648The 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.
2649In 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.
2650The key mechanism to support separate compilation is \CFA's \emph{explicit} use of assumed type properties.
2651Until \CC concepts~\cite{C++Concepts} are standardized (anticipated for \CCtwenty), \CC provides no way to specify the requirements of a generic function beyond compilation errors during template expansion;
2652furthermore, \CC concepts are restricted to template polymorphism.
2653
2654Cyclone~\cite{Grossman06} also provides capabilities for polymorphic functions and existential types, similar to \CFA's @forall@ functions and generic types.
2655Cyclone 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.
2656Furthermore, 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@.
2657In \CFA terms, all Cyclone polymorphism must be dtype-static.
2658While 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.
2659Smith 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.
2660
2661Objective-C~\cite{obj-c-book} is an industrially successful extension to C.
2662However, Objective-C is a radical departure from C, using an object-oriented model with message-passing.
2663Objective-C did not support type-checked generics until recently \cite{xcode7}, historically using less-efficient runtime checking of object types.
2664The GObject~\cite{GObject} framework also adds object-oriented programming with runtime type-checking and reference-counting garbage-collection to C;
2665these features are more intrusive additions than those provided by \CFA, in addition to the runtime overhead of reference-counting.
2666Vala~\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.
2667Java~\cite{Java8} included generic types in Java~5, which are type-checked at compilation and type-erased at runtime, similar to \CFA's.
2668However, in Java, each object carries its own table of method pointers, while \CFA passes the method pointers separately to maintain a C-compatible layout.
2669Java is also a garbage-collected, object-oriented language, with the associated resource usage and C-interoperability burdens.
2670
2671D~\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.
2672However, 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.
2673D and Go are garbage-collected languages, imposing the associated runtime overhead.
2674The necessity of accounting for data transfer between managed runtimes and the unmanaged C runtime complicates foreign-function interfaces to C.
2675Furthermore, 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.
2676D 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.
2677Rust also possesses much more powerful abstraction capabilities for writing generic code than Go.
2678On 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.
2679\CFA, with its more modest safety features, allows direct ports of C code while maintaining the idiomatic style of the original source.
2680
2681
2682\subsection{Tuples/Variadics}
2683
2684Many programming languages have some form of tuple construct and/or variadic functions, \eg SETL, C, KW-C, \CC, D, Go, Java, ML, and Scala.
2685SETL~\cite{SETL} is a high-level mathematical programming language, with tuples being one of the primary data types.
2686Tuples in SETL allow subscripting, dynamic expansion, and multiple assignment.
2687C 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.
2688KW-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.
2689The main contributions of that work were adding MRVF, tuple mass and multiple assignment, and record-field access.
2690\CCeleven introduced @std::tuple@ as a library variadic template structure.
2691Tuples are a generalization of @std::pair@, in that they allow for arbitrary length, fixed-size aggregation of heterogeneous values.
2692Operations include @std::get<N>@ to extract values, @std::tie@ to create a tuple of references used for assignment, and lexicographic comparisons.
2693\CCseventeen proposes \emph{structured bindings}~\cite{Sutter15} to eliminate pre-declaring variables and use of @std::tie@ for binding the results.
2694This 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.
2695Furthermore, structured bindings are not a full replacement for @std::tie@, as it always declares new variables.
2696Like \CC, D provides tuples through a library variadic-template structure.
2697Go does not have tuples but supports MRVF.
2698Java'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.
2699Tuples are a fundamental abstraction in most functional programming languages, such as Standard ML~\cite{sml} and~\cite{Scala}, which decompose tuples using pattern matching.
2700
2701
2702\subsection{Control Structures / Declarations / Literals}
2703
2704Java has default fall through like C/\CC.
2705Pascal/Ada/Go/Rust do not have default fall through.
2706\Csharp does not have fall through but still requires a break.
2707Python uses dictionary mapping. \\
2708\CFA choose is like Rust match.
2709
2710Java has labelled break/continue. \\
2711Languages with and without exception handling.
2712
2713Alternative C declarations. \\
2714Different references \\
2715Constructors/destructors
2716
27170/1 Literals \\
2718user defined: D, Objective-C
2719
2720\section{Conclusion and Future Work}
2721
2722The 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.
2723While 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.
2724The 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.
2725The contributions are a powerful type-system using parametric polymorphism and overloading, generic types, tuples, advanced control structures, and extended declarations, which all have complex interactions.
2726The work is a challenging design, engineering, and implementation exercise.
2727On the surface, the project may appear as a rehash of similar mechanisms in \CC.
2728However, 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.
2729All of these new features are being used by the \CFA development-team to build the \CFA runtime-system.
2730Finally, we demonstrate that \CFA performance for some idiomatic cases is better than C and close to \CC, showing the design is practically applicable.
2731
2732There is ongoing work on a wide range of \CFA feature extensions, including arrays with size, user-defined conversions, concurrent primitives, and modules.
2733(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.)
2734In addition, there are interesting future directions for the polymorphism design.
2735Notably, \CC template functions trade compile time and code bloat for optimal runtime of individual instantiations of polymorphic functions.
2736\CFA polymorphic functions use dynamic virtual-dispatch;
2737the 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.
2738Two 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).
2739These approaches are not mutually exclusive and allow performance optimizations to be applied only when necessary, without suffering global code-bloat.
2740In general, we believe separate compilation, producing smaller code, works well with loaded hardware-caches, which may offset the benefit of larger inlined-code.
2741
2742
2743\section{Acknowledgments}
2744
2745The authors would like to recognize the design assistance of Glen Ditchfield, Richard Bilson, Thierry Delisle, and Andrew Beach on the features described in this paper, and thank Magnus Madsen for feedback in the writing.
2746This work is supported through a corporate partnership with Huawei Ltd.\ (\url{http://www.huawei.com}), and Aaron Moss and Peter Buhr are partially funded by the Natural Sciences and Engineering Research Council of Canada.
2747
2748% 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.
2749
2750
2751\bibliographystyle{plain}
2752\bibliography{pl}
2753
2754
2755\appendix
2756
2757\section{Benchmark Stack Implementation}
2758\label{sec:BenchmarkStackImplementation}
2759
2760\lstset{basicstyle=\linespread{0.9}\sf\small}
2761
2762Throughout, @/***/@ designates a counted redundant type annotation.
2763
2764\smallskip\noindent
2765\CFA
2766\begin{cfa}[xleftmargin=2\parindentlnth,aboveskip=0pt,belowskip=0pt]
2767forall(otype T) struct stack_node;
2768forall(otype T) struct stack {
2769        stack_node(T) * head;
2770};
2771forall(otype T) struct stack_node {
2772        T value;
2773        stack_node(T) * next;
2774};
2775forall(otype T) void ?{}( stack(T) & s ) { (s.head){ 0 }; }
2776forall(otype T) void ?{}( stack(T) & s, stack(T) t ) {
2777        stack_node(T) ** crnt = &s.head;
2778        for ( stack_node(T) * next = t.head; next; next = next->next ) {
2779                stack_node(T) * new_node = ((stack_node(T)*)malloc());
2780                (*new_node){ next->value }; /***/
2781                *crnt = new_node;
2782                stack_node(T) * acrnt = *crnt;
2783                crnt = &acrnt->next;
2784        }
2785        *crnt = 0;
2786}
2787forall(otype T) stack(T) ?=?( stack(T) & s, stack(T) t ) {
2788        if ( s.head == t.head ) return s;
2789        clear( s );
2790        s{ t };
2791        return s;
2792}
2793forall(otype T) void ^?{}( stack(T) & s) { clear( s ); }
2794forall(otype T) _Bool empty( const stack(T) & s ) { return s.head == 0; }
2795forall(otype T) void push( stack(T) & s, T value ) {
2796        stack_node(T) * new_node = ((stack_node(T)*)malloc());
2797        (*new_node){ value, s.head }; /***/
2798        s.head = new_node;
2799}
2800forall(otype T) T pop( stack(T) & s ) {
2801        stack_node(T) * n = s.head;
2802        s.head = n->next;
2803        T v = n->value;
2804        delete( n );
2805        return v;
2806}
2807forall(otype T) void clear( stack(T) & s ) {
2808        for ( stack_node(T) * next = s.head; next; ) {
2809                stack_node(T) * crnt = next;
2810                next = crnt->next;
2811                delete( crnt );
2812        }
2813        s.head = 0;
2814}
2815\end{cfa}
2816
2817\medskip\noindent
2818\CC
2819\begin{cfa}[xleftmargin=2\parindentlnth,aboveskip=0pt,belowskip=0pt]
2820template<typename T> class stack {
2821        struct node {
2822                T value;
2823                node * next;
2824                node( const T & v, node * n = nullptr ) : value(v), next(n) {}
2825        };
2826        node * head;
2827        void copy(const stack<T>& o) {
2828                node ** crnt = &head;
2829                for ( node * next = o.head;; next; next = next->next ) {
2830                        *crnt = new node{ next->value }; /***/
2831                        crnt = &(*crnt)->next;
2832                }
2833                *crnt = nullptr;
2834        }
2835  public:
2836        stack() : head(nullptr) {}
2837        stack(const stack<T>& o) { copy(o); }
2838        stack(stack<T> && o) : head(o.head) { o.head = nullptr; }
2839        ~stack() { clear(); }
2840        stack & operator= (const stack<T>& o) {
2841                if ( this == &o ) return *this;
2842                clear();
2843                copy(o);
2844                return *this;
2845        }
2846        stack & operator= (stack<T> && o) {
2847                if ( this == &o ) return *this;
2848                head = o.head;
2849                o.head = nullptr;
2850                return *this;
2851        }
2852        bool empty() const { return head == nullptr; }
2853        void push(const T & value) { head = new node{ value, head };  /***/ }
2854        T pop() {
2855                node * n = head;
2856                head = n->next;
2857                T x = std::move(n->value);
2858                delete n;
2859                return x;
2860        }
2861        void clear() {
2862                for ( node * next = head; next; ) {
2863                        node * crnt = next;
2864                        next = crnt->next;
2865                        delete crnt;
2866                }
2867                head = nullptr;
2868        }
2869};
2870\end{cfa}
2871
2872\medskip\noindent
2873C
2874\begin{cfa}[xleftmargin=2\parindentlnth,aboveskip=0pt,belowskip=0pt]
2875struct stack_node {
2876        void * value;
2877        struct stack_node * next;
2878};
2879struct stack new_stack() { return (struct stack){ NULL }; /***/ }
2880void copy_stack(struct stack * s, const struct stack * t, void * (*copy)(const void *)) {
2881        struct stack_node ** crnt = &s->head;
2882        for ( struct stack_node * next = t->head; next; next = next->next ) {
2883                *crnt = malloc(sizeof(struct stack_node)); /***/
2884                **crnt = (struct stack_node){ copy(next->value) }; /***/
2885                crnt = &(*crnt)->next;
2886        }
2887        *crnt = 0;
2888}
2889_Bool stack_empty(const struct stack * s) { return s->head == NULL; }
2890void push_stack(struct stack * s, void * value) {
2891        struct stack_node * n = malloc(sizeof(struct stack_node)); /***/
2892        *n = (struct stack_node){ value, s->head }; /***/
2893        s->head = n;
2894}
2895void * pop_stack(struct stack * s) {
2896        struct stack_node * n = s->head;
2897        s->head = n->next;
2898        void * x = n->value;
2899        free(n);
2900        return x;
2901}
2902void clear_stack(struct stack * s, void (*free_el)(void *)) {
2903        for ( struct stack_node * next = s->head; next; ) {
2904                struct stack_node * crnt = next;
2905                next = crnt->next;
2906                free_el(crnt->value);
2907                free(crnt);
2908        }
2909        s->head = NULL;
2910}
2911\end{cfa}
2912
2913\medskip\noindent
2914\CCV
2915\begin{cfa}[xleftmargin=2\parindentlnth,aboveskip=0pt,belowskip=0pt]
2916stack::node::node( const object & v, node * n ) : value( v.new_copy() ), next( n ) {}
2917void stack::copy(const stack & o) {
2918        node ** crnt = &head;
2919        for ( node * next = o.head; next; next = next->next ) {
2920                *crnt = new node{ *next->value };
2921                crnt = &(*crnt)->next;
2922        }
2923        *crnt = nullptr;
2924}
2925stack::stack() : head(nullptr) {}
2926stack::stack(const stack & o) { copy(o); }
2927stack::stack(stack && o) : head(o.head) { o.head = nullptr; }
2928stack::~stack() { clear(); }
2929stack & stack::operator= (const stack & o) {
2930        if ( this == &o ) return *this;
2931        clear();
2932        copy(o);
2933        return *this;
2934}
2935stack & stack::operator= (stack && o) {
2936        if ( this == &o ) return *this;
2937        head = o.head;
2938        o.head = nullptr;
2939        return *this;
2940}
2941bool stack::empty() const { return head == nullptr; }
2942void stack::push(const object & value) { head = new node{ value, head }; /***/ }
2943ptr<object> stack::pop() {
2944        node * n = head;
2945        head = n->next;
2946        ptr<object> x = std::move(n->value);
2947        delete n;
2948        return x;
2949}
2950void stack::clear() {
2951        for ( node * next = head; next; ) {
2952                node * crnt = next;
2953                next = crnt->next;
2954                delete crnt;
2955        }
2956        head = nullptr;
2957}
2958\end{cfa}
2959
2960
2961\end{document}
2962
2963% Local Variables: %
2964% tab-width: 4 %
2965% compile-command: "make" %
2966% End: %
Note: See TracBrowser for help on using the repository browser.