source: doc/papers/general/Paper.tex @ 387c9a1

ADTaaron-thesisarm-ehast-experimentalcleanup-dtorsdeferred_resndemanglerenumforall-pointer-decayjacob/cs343-translationjenkins-sandboxnew-astnew-ast-unique-exprnew-envno_listpersistent-indexerpthread-emulationqualifiedEnumwith_gc
Last change on this file since 387c9a1 was 387c9a1, checked in by Peter A. Buhr <pabuhr@…>, 6 years ago

corrections for referee responses

  • Property mode set to 100644
File size: 159.4 KB
Line 
1\documentclass[AMA,STIX1COL]{WileyNJD-v2}
2
3\articletype{RESEARCH ARTICLE}%
4
5\received{26 April 2016}
6\revised{6 June 2016}
7\accepted{6 June 2016}
8
9\raggedbottom
10
11%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
12
13% Latex packages used in the document.
14
15\usepackage{epic,eepic}
16\usepackage{xspace}
17\usepackage{comment}
18\usepackage{upquote}                                            % switch curled `'" to straight
19\usepackage{listings}                                           % format program code
20\captionsetup{justification=raggedright,singlelinecheck=false}
21%\usepackage{enumitem}
22%\setlist[itemize]{topsep=3pt,itemsep=2pt,parsep=0pt}% global
23%\usepackage{rotating}
24
25\hypersetup{breaklinks=true}
26\definecolor{ForestGreen}{cmyk}{1, 0, 0.99995, 0}
27
28\usepackage[pagewise]{lineno}
29\renewcommand{\linenumberfont}{\scriptsize\sffamily}
30
31\lefthyphenmin=3                                                        % hyphen only after 4 characters
32\righthyphenmin=3
33
34%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
35
36% Names used in the document.
37
38\newcommand{\CFAIcon}{\textsf{C}\raisebox{\depth}{\rotatebox{180}{\textsf{A}}}\xspace} % Cforall symbolic name
39\newcommand{\CFA}{\protect\CFAIcon}             % safe for section/caption
40\newcommand{\CFL}{\textrm{Cforall}\xspace}      % Cforall symbolic name
41\newcommand{\Celeven}{\textrm{C11}\xspace}      % C11 symbolic name
42\newcommand{\CC}{\textrm{C}\kern-.1em\hbox{+\kern-.25em+}\xspace} % C++ symbolic name
43\newcommand{\CCeleven}{\textrm{C}\kern-.1em\hbox{+\kern-.25em+}11\xspace} % C++11 symbolic name
44\newcommand{\CCfourteen}{\textrm{C}\kern-.1em\hbox{+\kern-.25em+}14\xspace} % C++14 symbolic name
45\newcommand{\CCseventeen}{\textrm{C}\kern-.1em\hbox{+\kern-.25em+}17\xspace} % C++17 symbolic name
46\newcommand{\CCtwenty}{\textrm{C}\kern-.1em\hbox{+\kern-.25em+}20\xspace} % C++20 symbolic name
47\newcommand{\CCV}{\rm C\kern-.1em\hbox{+\kern-.25em+}obj\xspace} % C++ virtual symbolic name
48\newcommand{\Csharp}{C\raisebox{-0.7ex}{\Large$^\sharp$}\xspace} % C# symbolic name
49
50%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
51
52\newcommand{\Textbf}[2][red]{{\color{#1}{\textbf{#2}}}}
53%\newcommand{\TODO}[1]{\textbf{TODO}: {\itshape #1}} % TODO included
54\newcommand{\TODO}[1]{} % TODO elided
55
56% Default underscore is too low and wide. Cannot use lstlisting "literate" as replacing underscore
57% removes it as a variable-name character so keywords in variables are highlighted. MUST APPEAR
58% AFTER HYPERREF.
59%\DeclareTextCommandDefault{\textunderscore}{\leavevmode\makebox[1.2ex][c]{\rule{1ex}{0.1ex}}}
60\renewcommand{\textunderscore}{\leavevmode\makebox[1.2ex][c]{\rule{1ex}{0.075ex}}}
61
62\renewcommand*{\thefootnote}{\alph{footnote}} % hack because fnsymbol does not work
63%\renewcommand*{\thefootnote}{\fnsymbol{footnote}}
64
65\makeatletter
66% parindent is relative, i.e., toggled on/off in environments like itemize, so store the value for
67% use rather than use \parident directly.
68\newlength{\parindentlnth}
69\setlength{\parindentlnth}{\parindent}
70
71\newcommand{\LstBasicStyle}[1]{{\lst@basicstyle{\lst@basicstyle{#1}}}}
72\newcommand{\LstKeywordStyle}[1]{{\lst@basicstyle{\lst@keywordstyle{#1}}}}
73\newcommand{\LstCommentStyle}[1]{{\lst@basicstyle{\lst@commentstyle{#1}}}}
74
75\newlength{\gcolumnposn}                                        % temporary hack because lstlisting does not handle tabs correctly
76\newlength{\columnposn}
77\setlength{\gcolumnposn}{3.5in}
78\setlength{\columnposn}{\gcolumnposn}
79
80\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}}}}
81\newcommand{\CRT}{\global\columnposn=\gcolumnposn}
82
83% Denote newterms in particular font and index them without particular font and in lowercase, e.g., \newterm{abc}.
84% The option parameter provides an index term different from the new term, e.g., \newterm[\texttt{abc}]{abc}
85% The star version does not lowercase the index information, e.g., \newterm*{IBM}.
86\newcommand{\newtermFontInline}{\emph}
87\newcommand{\newterm}{\@ifstar\@snewterm\@newterm}
88\newcommand{\@newterm}[2][\@empty]{\lowercase{\def\temp{#2}}{\newtermFontInline{#2}}\ifx#1\@empty\index{\temp}\else\index{#1@{\protect#2}}\fi}
89\newcommand{\@snewterm}[2][\@empty]{{\newtermFontInline{#2}}\ifx#1\@empty\index{#2}\else\index{#1@{\protect#2}}\fi}
90
91% Latin abbreviation
92\newcommand{\abbrevFont}{\textit}                       % set empty for no italics
93\@ifundefined{eg}{
94\newcommand{\EG}{\abbrevFont{e}\abbrevFont{g}}
95\newcommand*{\eg}{%
96        \@ifnextchar{,}{\EG}%
97                {\@ifnextchar{:}{\EG}%
98                        {\EG,\xspace}}%
99}}{}%
100\@ifundefined{ie}{
101\newcommand{\IE}{\abbrevFont{i}\abbrevFont{e}}
102\newcommand*{\ie}{%
103        \@ifnextchar{,}{\IE}%
104                {\@ifnextchar{:}{\IE}%
105                        {\IE,\xspace}}%
106}}{}%
107\@ifundefined{etc}{
108\newcommand{\ETC}{\abbrevFont{etc}}
109\newcommand*{\etc}{%
110        \@ifnextchar{.}{\ETC}%
111        {\ETC.\xspace}%
112}}{}%
113\@ifundefined{etal}{
114\newcommand{\ETAL}{\abbrevFont{et}~\abbrevFont{al}}
115\newcommand*{\etal}{%
116        \@ifnextchar{.}{\protect\ETAL}%
117                {\protect\ETAL.\xspace}%
118}}{}%
119\@ifundefined{viz}{
120\newcommand{\VIZ}{\abbrevFont{viz}}
121\newcommand*{\viz}{%
122        \@ifnextchar{.}{\VIZ}%
123                {\VIZ.\xspace}%
124}}{}%
125\makeatother
126
127\newenvironment{cquote}{%
128        \list{}{\lstset{resetmargins=true,aboveskip=0pt,belowskip=0pt}\topsep=3pt\parsep=0pt\leftmargin=\parindentlnth\rightmargin\leftmargin}%
129        \item\relax
130}{%
131        \endlist
132}% cquote
133
134% CFA programming language, based on ANSI C (with some gcc additions)
135\lstdefinelanguage{CFA}[ANSI]{C}{
136        morekeywords={
137                _Alignas, _Alignof, __alignof, __alignof__, asm, __asm, __asm__, __attribute, __attribute__,
138                auto, _Bool, catch, catchResume, choose, _Complex, __complex, __complex__, __const, __const__,
139                coroutine, disable, dtype, enable, exception, __extension__, fallthrough, fallthru, finally,
140                __float80, float80, __float128, float128, forall, ftype, _Generic, _Imaginary, __imag, __imag__,
141                inline, __inline, __inline__, __int128, int128, __label__, monitor, mutex, _Noreturn, one_t, or,
142                otype, restrict, __restrict, __restrict__, __signed, __signed__, _Static_assert, thread,
143                _Thread_local, throw, throwResume, timeout, trait, try, ttype, typeof, __typeof, __typeof__,
144                virtual, __volatile, __volatile__, waitfor, when, with, zero_t},
145        moredirectives={defined,include_next}%
146}
147
148\lstset{
149language=CFA,
150columns=fullflexible,
151basicstyle=\linespread{0.9}\sf,                                                 % reduce line spacing and use sanserif font
152stringstyle=\tt,                                                                                % use typewriter font
153tabsize=5,                                                                                              % N space tabbing
154xleftmargin=\parindentlnth,                                                             % indent code to paragraph indentation
155%mathescape=true,                                                                               % LaTeX math escape in CFA code $...$
156escapechar=\$,                                                                                  % LaTeX escape in CFA code
157keepspaces=true,                                                                                %
158showstringspaces=false,                                                                 % do not show spaces with cup
159showlines=true,                                                                                 % show blank lines at end of code
160aboveskip=4pt,                                                                                  % spacing above/below code block
161belowskip=3pt,
162% replace/adjust listing characters that look bad in sanserif
163literate={-}{\makebox[1ex][c]{\raisebox{0.4ex}{\rule{0.8ex}{0.1ex}}}}1 {^}{\raisebox{0.6ex}{$\scriptstyle\land\,$}}1
164        {~}{\raisebox{0.3ex}{$\scriptstyle\sim\,$}}1 % {`}{\ttfamily\upshape\hspace*{-0.1ex}`}1
165        {<-}{$\leftarrow$}2 {=>}{$\Rightarrow$}2 {->}{\makebox[1ex][c]{\raisebox{0.4ex}{\rule{0.8ex}{0.075ex}}}\kern-0.2ex{\textgreater}}2,
166moredelim=**[is][\color{red}]{`}{`},
167}% lstset
168
169\lstnewenvironment{cfa}[1][]
170{\lstset{#1}}
171{}
172\lstnewenvironment{C++}[1][]                            % use C++ style
173{\lstset{language=C++,moredelim=**[is][\protect\color{red}]{`}{`},#1}\lstset{#1}}
174{}
175
176% inline code @...@
177\lstMakeShortInline@%
178
179\let\OLDthebibliography\thebibliography
180\renewcommand\thebibliography[1]{
181  \OLDthebibliography{#1}
182  \setlength{\parskip}{0pt}
183  \setlength{\itemsep}{4pt plus 0.3ex}
184}
185
186\title{\texorpdfstring{\protect\CFA : Adding Modern Programming Language Features to C}{Cforall : Adding Modern Programming Language Features to C}}
187
188\author[1]{Aaron Moss}
189\author[1]{Robert Schluntz}
190\author[1]{Peter A. Buhr*}
191\authormark{MOSS \textsc{et al}}
192
193\address[1]{\orgdiv{Cheriton School of Computer Science}, \orgname{University of Waterloo}, \orgaddress{\state{Waterloo, ON}, \country{Canada}}}
194
195\corres{*Peter A. Buhr, Cheriton School of Computer Science, University of Waterloo, 200 University Avenue West, Waterloo, ON, N2L 3G1, Canada. \email{pabuhr{\char`\@}uwaterloo.ca}}
196
197\fundingInfo{Natural Sciences and Engineering Research Council of Canada}
198
199\abstract[Summary]{
200The 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.
201This installation base and the programmers producing it represent a massive software-engineering investment spanning decades and likely to continue for decades more.
202Nevertheless, C, first standardized almost forty years ago, lacks many features that make programming in more modern languages safer and more productive.
203
204The goal of the \CFA project (pronounced ``C-for-all'') 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.
205Prior 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.
206Specifically, \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.
207This paper presents a quick tour of \CFA features showing how their design avoids shortcomings of similar features in C and other C-like languages.
208Finally, experimental results are presented to validate several of the new features.
209}%
210
211\keywords{generic types, tuple types, variadic types, polymorphic functions, C, Cforall}
212
213
214\begin{document}
215\linenumbers                                            % comment out to turn off line numbering
216
217\maketitle
218
219
220\section{Introduction}
221The 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.
222This installation base and the programmers producing it represent a massive software-engineering investment spanning decades and likely to continue for decades more.
223The 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.
224The top 3 rankings over the past 30 years are:
225\begin{center}
226\setlength{\tabcolsep}{10pt}
227\lstDeleteShortInline@%
228\begin{tabular}{@{}rccccccc@{}}
229                & 2018  & 2013  & 2008  & 2003  & 1998  & 1993  & 1988  \\ \hline
230Java    & 1             & 2             & 1             & 1             & 18    & -             & -             \\
231\Textbf{C}& \Textbf{2} & \Textbf{1} & \Textbf{2} & \Textbf{2} & \Textbf{1} & \Textbf{1} & \Textbf{1} \\
232\CC             & 3             & 4             & 3             & 3             & 2             & 2             & 5             \\
233\end{tabular}
234\lstMakeShortInline@%
235\end{center}
236Love it or hate it, C is extremely popular, highly used, and one of the few systems languages.
237In many cases, \CC is often used solely as a better C.
238Nevertheless, C, first standardized almost forty years ago~\cite{ANSI89:C}, lacks many features that make programming in more modern languages safer and more productive.
239
240\CFA (pronounced ``C-for-all'', and written \CFA or Cforall) is an evolutionary extension of the C programming language that adds modern language-features to C, while maintaining both source and runtime compatibility with C and a familiar programming model for programmers.
241The four key design goals for \CFA~\cite{Bilson03} are:
242(1) The behaviour of standard C code must remain the same when translated by a \CFA compiler as when translated by a C compiler;
243(2) Standard C code must be as fast and as small when translated by a \CFA compiler as when translated by a C compiler;
244(3) \CFA code must be at least as portable as standard C code;
245(4) Extensions introduced by \CFA must be translated in the most efficient way possible.
246These 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.
247\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.
248
249All languages features discussed in this paper are working, except some advanced exception-handling features.
250Not discussed in this paper are the integrated concurrency-constructs and user-level threading-library~\cite{Delisle18}.
251\CFA is an \emph{open-source} project implemented as an 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).
252Ultimately, a compiler is necessary for advanced features and optimal performance.
253% @plg2[9]% cd cfa-cc/src; cloc ArgTweak CodeGen CodeTools Common Concurrency ControlStruct Designators GenPoly InitTweak MakeLibCfa.cc MakeLibCfa.h Parser ResolvExpr SymTab SynTree Tuples driver prelude main.cc
254% -------------------------------------------------------------------------------
255% Language                     files          blank        comment           code
256% -------------------------------------------------------------------------------
257% C++                            108           5420           5232          34961
258% C/C++ Header                    86           2379           2450           8464
259% Teamcenter def                   2            115             65           1387
260% make                             5            168             87           1052
261% C                               20            109            403            488
262% awk                              1             12             26            121
263% sed                              1              0              0              6
264% -------------------------------------------------------------------------------
265% SUM:                           223           8203           8263          46479
266% -------------------------------------------------------------------------------
267The \CFA translator is 200+ files and 46,000+ lines of code written in C/\CC.
268Starting with a translator versus a compiler makes it easier and faster to generate and debug C object-code rather than intermediate, assembler or machine code.
269The translator design is based on the \emph{visitor pattern}, allowing multiple passes over the abstract code-tree, which works well for incrementally adding new feature through additional visitor passes.
270At the heart of the translator is the type resolver, which handles the polymorphic routine/type overload-resolution.
271% @plg2[8]% cd cfa-cc/src; cloc libcfa
272% -------------------------------------------------------------------------------
273% Language                     files          blank        comment           code
274% -------------------------------------------------------------------------------
275% C                               35           1256           1240           9116
276% C/C++ Header                    54            358           1106           1198
277% make                             2            201            325           1167
278% C++                              3             18             17            124
279% Assembly                         3             56             97            111
280% Bourne Shell                     2              2              0             25
281% awk                              1              4              0             22
282% -------------------------------------------------------------------------------
283% SUM:                           100           1895           2785          11763
284% -------------------------------------------------------------------------------
285The \CFA runtime system is 100+ files and 11,000+ lines of code, written in \CFA.
286Currently, the \CFA runtime is the largest \emph{user} of \CFA providing a vehicle to test the language features and implementation.
287% @plg2[6]% cd cfa-cc/src; cloc tests examples benchmark
288% -------------------------------------------------------------------------------
289% Language                     files          blank        comment           code
290% -------------------------------------------------------------------------------
291% C                              237          12260           2869          23286
292% make                             8            464            245           2838
293% C/C++ Header                    22            225            175            785
294% Python                           5            131             93            420
295% C++                             10             48              5            201
296% Lua                              2             31              4            126
297% Java                             4              5              0             80
298% Go                               2             11              9             40
299% -------------------------------------------------------------------------------
300% SUM:                           290          13175           3400          27776
301% -------------------------------------------------------------------------------
302The \CFA tests are 290+ files and 27,000+ lines of code.
303The tests illustrate syntactic and semantic features in \CFA, plus a growing number of runtime benchmarks.
304The tests check for correctness and are used for daily regression testing of 3800+ commits.
305
306Finally, it is impossible to describe a programming language without usages before definitions.
307Therefore, syntax and semantics appear before explanations;
308hence, patience is necessary until details are presented.
309
310
311\section{Polymorphic Functions}
312
313\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}.
314Shortcomings are identified in existing approaches to generic and variadic data types in C-like languages and how these shortcomings are avoided in \CFA.
315Specifically, 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.
316The new constructs are empirically compared with C and \CC approaches via performance experiments in Section~\ref{sec:eval}.
317
318
319\subsection{Name Overloading}
320\label{s:NameOverloading}
321
322\begin{quote}
323There are only two hard things in Computer Science: cache invalidation and \emph{naming things} -- Phil Karlton
324\end{quote}
325\vspace{-10pt}
326C 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.
327\CFA extends the built-in operator overloading by allowing users to define overloads for any function, not just operators, and even any variable;
328Section~\ref{sec:libraries} includes a number of examples of how this overloading simplifies \CFA programming relative to C.
329Code 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.
330As an example:
331\begin{cfa}
332int max = 2147483647;                                           $\C[4in]{// (1)}$
333double max = 1.7976931348623157E+308;           $\C{// (2)}$
334int max( int a, int b ) { return a < b ? b : a; }  $\C{// (3)}$
335double max( double a, double b ) { return a < b ? b : a; }  $\C{// (4)}\CRT$
336max( 7, -max );                                         $\C{// uses (3) and (1), by matching int from constant 7}$
337max( max, 3.14 );                                       $\C{// uses (4) and (2), by matching double from constant 3.14}$
338max( max, -max );                                       $\C{// ERROR, ambiguous}$
339int m = max( max, -max );                       $\C{// uses (3) and (1) twice, by matching return type}$
340\end{cfa}
341
342\CFA maximizes the ability to reuse names to aggressively address the naming problem.
343In some cases, hundreds of names can be reduced to tens, resulting in a significant cognitive reduction.
344In the above, the name @max@ has a consistent meaning, and a programmer only needs to remember the single concept: maximum.
345To 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).
346As 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.
347
348\Celeven added @_Generic@ expressions~\cite[\S~6.5.1.1]{C11}, which is used with preprocessor macros to provide ad-hoc polymorphism;
349however, this polymorphism is both functionally and ergonomically inferior to \CFA name overloading.
350The macro wrapping the generic expression imposes some limitations;
351\eg, it cannot implement the example above, because the variables @max@ are ambiguous with the functions @max@.
352Ergonomic 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.
353\CFA supports @_Generic@ expressions for backwards compatibility, but it is an unnecessary mechanism. \TODO{actually implement that}
354
355% http://fanf.livejournal.com/144696.html
356% http://www.robertgamble.net/2012/01/c11-generic-selections.html
357% https://abissell.com/2014/01/16/c11s-_generic-keyword-macro-applications-and-performance-impacts/
358
359
360\subsection{\texorpdfstring{\protect\lstinline{forall} Functions}{forall Functions}}
361\label{sec:poly-fns}
362
363The 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):
364\begin{cfa}
365`forall( otype T )` T identity( T val ) { return val; }
366int forty_two = identity( 42 );         $\C{// T is bound to int, forty\_two == 42}$
367\end{cfa}
368This @identity@ function can be applied to any complete \newterm{object type} (or @otype@).
369The 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.
370The \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.
371If this extra information is not needed, \eg for a pointer, the type parameter can be declared as a \newterm{data type} (or @dtype@).
372
373In \CFA, the polymorphic runtime-cost is spread over each polymorphic call, because more arguments are passed to polymorphic functions;
374the experiments in Section~\ref{sec:eval} show this overhead is similar to \CC virtual-function calls.
375A design advantage is that, unlike \CC template-functions, \CFA polymorphic-functions are compatible with C \emph{separate compilation}, preventing compilation and code bloat.
376
377Since 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.
378For example, the function @twice@ can be defined using the \CFA syntax for operator overloading:
379\begin{cfa}
380forall( otype T `| { T ?+?(T, T); }` ) T twice( T x ) { return x `+` x; }  $\C{// ? denotes operands}$
381int val = twice( twice( 3.7 ) );  $\C{// val == 14}$
382\end{cfa}
383which works for any type @T@ with a matching addition operator.
384The 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@.
385There 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.
386The first approach has a late conversion from @double@ to @int@ on the final assignment, while the second has an early conversion to @int@.
387\CFA minimizes the number of conversions and their potential to lose information, so it selects the first approach, which corresponds with C-programmer intuition.
388
389Crucial to the design of a new programming language are the libraries to access thousands of external software features.
390Like \CC, \CFA inherits a massive compatible library-base, where other programming languages must rewrite or provide fragile inter-language communication with C.
391A simple example is leveraging the existing type-unsafe (@void *@) C @bsearch@ to binary search a sorted float array:
392\begin{cfa}
393void * bsearch( const void * key, const void * base, size_t nmemb, size_t size,
394                                int (* compar)( const void *, const void * ));
395int comp( const void * t1, const void * t2 ) {
396         return *(double *)t1 < *(double *)t2 ? -1 : *(double *)t2 < *(double *)t1 ? 1 : 0;
397}
398double key = 5.0, vals[10] = { /* 10 sorted float values */ };
399double * val = (double *)bsearch( &key, vals, 10, sizeof(vals[0]), comp ); $\C{// search sorted array}$
400\end{cfa}
401which can be augmented simply with generalized, type-safe, \CFA-overloaded wrappers:
402\begin{cfa}
403forall( otype T | { int ?<?( T, T ); } ) T * bsearch( T key, const T * arr, size_t size ) {
404        int comp( const void * t1, const void * t2 ) { /* as above with double changed to T */ }
405        return (T *)bsearch( &key, arr, size, sizeof(T), comp );
406}
407forall( otype T | { int ?<?( T, T ); } ) unsigned int bsearch( T key, const T * arr, size_t size ) {
408        T * result = bsearch( key, arr, size ); $\C{// call first version}$
409        return result ? result - arr : size; $\C{// pointer subtraction includes sizeof(T)}$
410}
411double * val = bsearch( 5.0, vals, 10 ); $\C{// selection based on return type}$
412int posn = bsearch( 5.0, vals, 10 );
413\end{cfa}
414The nested function @comp@ provides the hidden interface from typed \CFA to untyped (@void *@) C, plus the cast of the result.
415Providing a hidden @comp@ function in \CC is awkward as lambdas do not use C calling-conventions and template declarations cannot appear at block scope.
416As well, an alternate kind of return is made available: position versus pointer to found element.
417\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@.
418
419\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}).
420For example, it is possible to write a type-safe \CFA wrapper @malloc@ based on the C @malloc@, where the return type supplies the type/size of the allocation, which is impossible in most type systems.
421\begin{cfa}
422forall( dtype T | sized(T) ) T * malloc( void ) { return (T *)malloc( sizeof(T) ); }
423// select type and size from left-hand side
424int * ip = malloc();  double * dp = malloc();  struct S {...} * sp = malloc();
425\end{cfa}
426
427Call-site inferencing and nested functions provide a localized form of inheritance.
428For example, the \CFA @qsort@ only sorts in ascending order using @<@.
429However, it is trivial to locally change this behaviour:
430\begin{cfa}
431forall( otype T | { int ?<?( T, T ); } ) void qsort( const T * arr, size_t size ) { /* use C qsort */ }
432int main() {
433        int ?<?( double x, double y ) { return x `>` y; } $\C{// locally override behaviour}$
434        qsort( vals, 10 );                                                      $\C{// descending sort}$
435}
436\end{cfa}
437The local version of @?<?@ performs @?>?@ overriding the built-in @?<?@ so it is passed to @qsort@.
438Hence, programmers can easily form local environments, adding and modifying appropriate functions, to maximize reuse of other existing functions and types.
439
440To reducing duplication, it is possible to distribute a group of @forall@ (and storage-class qualifiers) over functions/types, so each block declaration is prefixed by the group (see example in Appendix~\ref{s:CforallStack}).
441\begin{cfa}
442forall( otype `T` ) {                                                   $\C{// distribution block, add forall qualifier to declarations}$
443        struct stack { stack_node(`T`) * head; };       $\C{// generic type}$
444        inline {                                                                        $\C{// nested distribution block, add forall/inline to declarations}$
445                void push( stack(`T`) & s, `T` value ) ...      $\C{// generic operations}$
446        }
447}
448\end{cfa}
449
450
451\vspace*{-2pt}
452\subsection{Traits}
453
454\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:
455
456\begin{cquote}
457\lstDeleteShortInline@%
458\begin{tabular}{@{}l@{\hspace{\parindentlnth}}|@{\hspace{\parindentlnth}}l@{}}
459\begin{cfa}
460trait `sumable`( otype T ) {
461        void `?{}`( T &, zero_t ); // 0 literal constructor
462        T ?+?( T, T );                   // assortment of additions
463        T `?+=?`( T &, T );
464        T ++?( T & );
465        T ?++( T & );
466};
467\end{cfa}
468&
469\begin{cfa}
470forall( otype T `| sumable( T )` ) // use trait
471T sum( T a[$\,$], size_t size ) {
472        `T` total = { `0` };  // initialize by 0 constructor
473        for ( size_t i = 0; i < size; i += 1 )
474                total `+=` a[i]; // select appropriate +
475        return total;
476}
477\end{cfa}
478\end{tabular}
479\lstMakeShortInline@%
480\end{cquote}
481
482In 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:
483\begin{cfa}
484trait otype( dtype T | sized(T) ) {  // sized is a pseudo-trait for types with known size and alignment
485        void ?{}( T & );                                                $\C{// default constructor}$
486        void ?{}( T &, T );                                             $\C{// copy constructor}$
487        void ?=?( T &, T );                                             $\C{// assignment operator}$
488        void ^?{}( T & );                                               $\C{// destructor}$
489};
490\end{cfa}
491Given 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.
492
493In 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.
494Hence, trait names play no part in type equivalence;
495the names are simply macros for a list of polymorphic assertions, which are expanded at usage sites.
496Nevertheless, trait names form a logical subtype-hierarchy with @dtype@ at the top, where traits often contain overlapping assertions, \eg operator @+@.
497Traits are used like interfaces in Java or abstract base-classes in \CC, but without the nominal inheritance-relationships.
498Instead, 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.
499Hence, 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.
500(Nominal inheritance can be approximated with traits using marker variables or functions, as is done in Go.)
501
502% Nominal inheritance can be simulated with traits using marker variables or functions:
503% \begin{cfa}
504% trait nominal(otype T) {
505%     T is_nominal;
506% };
507% int is_nominal;                                                               $\C{// int now satisfies the nominal trait}$
508% \end{cfa}
509%
510% 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:
511% \begin{cfa}
512% trait pointer_like(otype Ptr, otype El) {
513%     lvalue El *?(Ptr);                                                $\C{// Ptr can be dereferenced into a modifiable value of type El}$
514% }
515% struct list {
516%     int value;
517%     list * next;                                                              $\C{// may omit "struct" on type names as in \CC}$
518% };
519% typedef list * list_iterator;
520%
521% lvalue int *?( list_iterator it ) { return it->value; }
522% \end{cfa}
523% 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@).
524% 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.
525
526
527\vspace*{-2pt}
528\section{Generic Types}
529
530A significant shortcoming of standard C is the lack of reusable type-safe abstractions for generic data structures and algorithms.
531Broadly speaking, there are three approaches to implement abstract data-structures in C.
532One approach is to write bespoke data-structures for each context in which they are needed.
533While 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.
534A 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.
535However, 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.
536A 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.
537Furthermore, writing and using preprocessor macros is unnatural and inflexible.
538
539\CC, Java, and other languages use \newterm{generic types} to produce type-safe abstract data-types.
540\CFA generic types integrate efficiently and naturally with the existing polymorphic functions, while retaining backwards compatibility with C and providing separate compilation.
541However, for known concrete parameters, the generic-type definition can be inlined, like \CC templates.
542
543A 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:
544\begin{cquote}
545\lstDeleteShortInline@%
546\begin{tabular}{@{}l|@{\hspace{\parindentlnth}}l@{}}
547\begin{cfa}
548`forall( otype R, otype S )` struct pair {
549        R first;        S second;
550};
551`forall( otype T )` // dynamic
552T value( pair(const char *, T) p ) { return p.second; }
553`forall( dtype F, otype T )` // dtype-static (concrete)
554T value( pair(F *, T * ) p) { return *p.second; }
555\end{cfa}
556&
557\begin{cfa}
558pair(const char *, int) p = {"magic", 42}; // concrete
559int i = value( p );
560pair(void *, int *) q = { 0, &p.second }; // concrete
561i = value( q );
562double d = 1.0;
563pair(double *, double *) r = { &d, &d }; // concrete
564d = value( r );
565\end{cfa}
566\end{tabular}
567\lstMakeShortInline@%
568\end{cquote}
569
570\CFA classifies generic types as either \newterm{concrete} or \newterm{dynamic}.
571Concrete types have a fixed memory layout regardless of type parameters, while dynamic types vary in memory layout depending on their type parameters.
572A \newterm{dtype-static} type has polymorphic parameters but is still concrete.
573Polymorphic 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.
574
575\CFA generic types also allow checked argument-constraints.
576For example, the following declaration of a sorted set-type ensures the set key supports equality and relational comparison:
577\begin{cfa}
578forall( otype Key | { _Bool ?==?(Key, Key); _Bool ?<?(Key, Key); } ) struct sorted_set;
579\end{cfa}
580
581
582\subsection{Concrete Generic-Types}
583
584The \CFA translator template-expands concrete generic-types into new structure types, affording maximal inlining.
585To 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.
586A 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.
587For example, the concrete instantiation for @pair( const char *, int )@ is:
588\begin{cfa}
589struct _pair_conc0 {
590        const char * first;  int second;
591};
592\end{cfa}
593
594A concrete generic-type with dtype-static parameters is also expanded to a structure type, but this type is used for all matching instantiations.
595In 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:
596\begin{cfa}
597struct _pair_conc1 {
598        void * first, * second;
599};
600\end{cfa}
601
602
603\subsection{Dynamic Generic-Types}
604
605Though \CFA implements concrete generic-types efficiently, it also has a fully general system for dynamic generic types.
606As 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.
607Dynamic generic-types also have an \newterm{offset array} containing structure-member offsets.
608A dynamic generic-@union@ needs no such offset array, as all members are at offset 0, but size and alignment are still necessary.
609Access 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.
610
611The offset arrays are statically generated where possible.
612If 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;
613if 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.
614As an example, the body of the second @value@ function is implemented as:
615\begin{cfa}
616_assign_T( _retval, p + _offsetof_pair[1] ); $\C{// return *p.second}$
617\end{cfa}
618@_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.
619@_offsetof_pair@ is the offset array passed into @value@; this array is generated at the call site as:
620\begin{cfa}
621size_t _offsetof_pair[] = { offsetof( _pair_conc0, first ), offsetof( _pair_conc0, second ) }
622\end{cfa}
623
624In some cases the offset arrays cannot be statically generated.
625For 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.
626\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.
627The \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.
628These 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).
629Results of these layout functions are cached so that they are only computed once per type per function. %, as in the example below for @pair@.
630Layout functions also allow generic types to be used in a function definition without reflecting them in the function signature.
631For instance, a function that strips duplicate values from an unsorted @vector(T)@ likely has a pointer to the vector as its only explicit parameter, but uses some sort of @set(T)@ internally to test for duplicate values.
632This function could acquire the layout for @set(T)@ by calling its layout function with the layout of @T@ implicitly passed into the function.
633
634Whether a type is concrete, dtype-static, or dynamic is decided solely on the @forall@'s type parameters.
635This 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.
636If 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.
637
638
639\subsection{Applications}
640\label{sec:generic-apps}
641
642The reuse of dtype-static structure instantiations enables useful programming patterns at zero runtime cost.
643The 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@:
644\begin{cfa}
645forall( dtype T ) int lexcmp( pair( T *, T * ) * a, pair( T *, T * ) * b, int (* cmp)( T *, T * ) ) {
646        return cmp( a->first, b->first ) ? : cmp( a->second, b->second );
647}
648\end{cfa}
649Since @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.
650
651Another useful pattern enabled by reused dtype-static type instantiations is zero-cost \newterm{tag-structures}.
652Sometimes information is only used for type-checking and can be omitted at runtime, \eg:
653\begin{cquote}
654\lstDeleteShortInline@%
655\begin{tabular}{@{}l|@{\hspace{\parindentlnth}}l@{}}
656\begin{cfa}
657forall( dtype Unit ) struct scalar { unsigned long value; };
658struct metres {};
659struct litres {};
660forall( dtype U ) scalar(U) ?+?( scalar(U) a, scalar(U) b ) {
661        return (scalar(U)){ a.value + b.value };
662}
663\end{cfa}
664&
665\begin{cfa}
666scalar(metres) half_marathon = { 21_098 };
667scalar(litres) pool = { 2_500_000 };
668scalar(metres) marathon = half_marathon +
669                                                        half_marathon;
670scalar(litres) two_pools = pool + pool;
671`marathon + pool;`      // ERROR, mismatched types
672\end{cfa}
673\end{tabular}
674\lstMakeShortInline@%
675\end{cquote}
676@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 @?+?@.
677These implementations may even be separately compiled, unlike \CC template functions.
678However, the \CFA type-checker ensures matching types are used by all calls to @?+?@, preventing nonsensical computations like adding a length to a volume.
679
680
681\section{Tuples}
682\label{sec:tuples}
683
684In many languages, functions can return at most one value;
685however, many operations have multiple outcomes, some exceptional.
686Consider C's @div@ and @remquo@ functions, which return the quotient and remainder for a division of integer and float values, respectively.
687\begin{cfa}
688typedef struct { int quo, rem; } div_t;         $\C{// from include stdlib.h}$
689div_t div( int num, int den );
690double remquo( double num, double den, int * quo );
691div_t qr = div( 13, 5 );                                        $\C{// return quotient/remainder aggregate}$
692int q;
693double r = remquo( 13.5, 5.2, &q );                     $\C{// return remainder, alias quotient}$
694\end{cfa}
695@div@ aggregates the quotient/remainder in a structure, while @remquo@ aliases a parameter to an argument.
696Both approaches are awkward.
697Alternatively, a programming language can directly support returning multiple values, \eg in \CFA:
698\begin{cfa}
699[ int, int ] div( int num, int den );           $\C{// return two integers}$
700[ double, double ] div( double num, double den ); $\C{// return two doubles}$
701int q, r;                                                                       $\C{// overloaded variable names}$
702double q, r;
703[ q, r ] = div( 13, 5 );                                        $\C{// select appropriate div and q, r}$
704[ q, r ] = div( 13.5, 5.2 );                            $\C{// assign into tuple}$
705\end{cfa}
706This approach is straightforward to understand and use;
707therefore, why do few programming languages support this obvious feature or provide it awkwardly?
708To answer, there are complex consequences that cascade through multiple aspects of the language, especially the type-system.
709This section show these consequences and how \CFA handles them.
710
711
712\subsection{Tuple Expressions}
713
714The addition of multiple-return-value functions (MRVF) are \emph{useless} without a syntax for accepting multiple values at the call-site.
715The simplest mechanism for capturing the return values is variable assignment, allowing the values to be retrieved directly.
716As 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}.
717
718However, 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:
719\begin{cfa}
720printf( "%d %d\n", div( 13, 5 ) );                      $\C{// return values seperated into arguments}$
721\end{cfa}
722Here, the values returned by @div@ are composed with the call to @printf@ by flattening the tuple into separate arguments.
723However, the \CFA type-system must support significantly more complex composition:
724\begin{cfa}
725[ int, int ] foo$\(_1\)$( int );                        $\C{// overloaded foo functions}$
726[ double ] foo$\(_2\)$( int );
727void bar( int, double, double );
728`bar`( foo( 3 ), foo( 3 ) );
729\end{cfa}
730The type-resolver only has the tuple return-types to resolve the call to @bar@ as the @foo@ parameters are identical, which involves unifying the possible @foo@ functions with @bar@'s parameter list.
731No combination of @foo@s are an exact match with @bar@'s parameters, so the resolver applies C conversions.
732The minimal cost is @bar( foo@$_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.
733
734
735\subsection{Tuple Variables}
736
737An important observation from function composition is that new variable names are not required to initialize parameters from an MRVF.
738\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:
739\begin{cfa}
740[ int, int ] qr = div( 13, 5 );                         $\C{// tuple-variable declaration and initialization}$
741[ double, double ] qr = div( 13.5, 5.2 );
742\end{cfa}
743where the tuple variable-name serves the same purpose as the parameter name(s).
744Tuple variables can be composed of any types, except for array types, since array sizes are generally unknown in C.
745
746One way to access the tuple-variable components is with assignment or composition:
747\begin{cfa}
748[ q, r ] = qr;                                                          $\C{// access tuple-variable components}$
749printf( "%d %d\n", qr );
750\end{cfa}
751\CFA also supports \newterm{tuple indexing} to access single components of a tuple expression:
752\begin{cfa}
753[int, int] * p = &qr;                                           $\C{// tuple pointer}$
754int rem = qr`.1`;                                                       $\C{// access remainder}$
755int quo = div( 13, 5 )`.0`;                                     $\C{// access quotient}$
756p`->0` = 5;                                                                     $\C{// change quotient}$
757bar( qr`.1`, qr );                                                      $\C{// pass remainder and quotient/remainder}$
758rem = [div( 13, 5 ), 42]`.0.1`;                         $\C{// access 2nd component of 1st component}$
759\end{cfa}
760
761
762\subsection{Flattening and Restructuring}
763
764In function call contexts, tuples support implicit flattening and restructuring conversions.
765Tuple flattening recursively expands a tuple into the list of its basic components.
766Tuple structuring packages a list of expressions into a value of tuple type, \eg:
767\begin{cfa}
768int f( int, int );
769[int] g( [int, int] );
770[int] h( int, [int, int] );
771[int, int] x;
772int y;
773f( x );                                                                         $\C{// flatten}$
774g( y, 10 );                                                                     $\C{// structure}$
775h( x, y );                                                                      $\C{// flatten and structure}$
776\end{cfa}
777In the call to @f@, @x@ is implicitly flattened so the components of @x@ are passed as the two arguments.
778In 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@.
779Finally, 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]@.
780The 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.
781
782
783\subsection{Tuple Assignment}
784
785An assignment where the left side is a tuple type is called \newterm{tuple assignment}.
786There 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.
787\begin{cfa}
788int x = 10;
789double y = 3.5;
790[int, double] z;
791z = [x, y];                                                                     $\C{// multiple assignment}$
792[x, y] = z;                                                                     $\C{// multiple assignment}$
793z = 10;                                                                         $\C{// mass assignment}$
794[y, x] = 3.14;                                                          $\C{// mass assignment}$
795\end{cfa}
796Both kinds of tuple assignment have parallel semantics, so that each value on the left and right side is evaluated before any assignments occur.
797As 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]@.
798This 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.
799For example, @[y, x] = 3.14@ performs the assignments @y = 3.14@ and @x = 3.14@, yielding @y == 3.14@ and @x == 3@;
800whereas, C cascading assignment @y = x = 3.14@ performs the assignments @x = 3.14@ and @y = x@, yielding @3@ in @y@ and @x@.
801Finally, 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.
802This example shows mass, multiple, and cascading assignment used in one expression:
803\begin{cfa}
804[void] f( [int, int] );
805f( [x, y] = z = 1.5 );                                          $\C{// assignments in parameter list}$
806\end{cfa}
807
808
809\subsection{Member Access}
810
811It is also possible to access multiple fields from a single expression using a \newterm{member-access}.
812The result is a single tuple-valued expression whose type is the tuple of the types of the members, \eg:
813\begin{cfa}
814struct S { int x; double y; char * z; } s;
815s.[x, y, z] = 0;
816\end{cfa}
817Here, the mass assignment sets all members of @s@ to zero.
818Since 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).
819\begin{cfa}
820[int, int, long, double] x;
821void f( double, long );
822x.[0, 1] = x.[1, 0];                                            $\C{// rearrange: [x.0, x.1] = [x.1, x.0]}$
823f( x.[0, 3] );                                                          $\C{// drop: f(x.0, x.3)}$
824[int, int, int] y = x.[2, 0, 2];                        $\C{// duplicate: [y.0, y.1, y.2] = [x.2, x.0.x.2]}$
825\end{cfa}
826It is also possible for a member access to contain other member accesses, \eg:
827\begin{cfa}
828struct A { double i; int j; };
829struct B { int * k; short l; };
830struct C { int x; A y; B z; } v;
831v.[x, y.[i, j], z.k];                                           $\C{// [v.x, [v.y.i, v.y.j], v.z.k]}$
832\end{cfa}
833
834
835\begin{comment}
836\subsection{Casting}
837
838In C, the cast operator is used to explicitly convert between types.
839In \CFA, the cast operator has a secondary use as type ascription.
840That 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:
841\begin{cfa}
842int f();     // (1)
843double f()// (2)
844
845f();       // ambiguous - (1),(2) both equally viable
846(int)f()// choose (2)
847\end{cfa}
848
849Since casting is a fundamental operation in \CFA, casts should be given a meaningful interpretation in the context of tuples.
850Taking a look at standard C provides some guidance with respect to the way casts should work with tuples:
851\begin{cfa}
852int f();
853void g();
854
855(void)f()// (1)
856(int)g()// (2)
857\end{cfa}
858In C, (1) is a valid cast, which calls @f@ and discards its result.
859On the other hand, (2) is invalid, because @g@ does not produce a result, so requesting an @int@ to materialize from nothing is nonsensical.
860Generalizing 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.
861
862Formally, 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 $T_i$.
863Excess 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.
864This approach follows naturally from the way that a cast to @void@ works in C.
865
866For example, in
867\begin{cfa}
868[int, int, int] f();
869[int, [int, int], int] g();
870
871([int, double])f();           $\C{// (1)}$
872([int, int, int])g();         $\C{// (2)}$
873([void, [int, int]])g();      $\C{// (3)}$
874([int, int, int, int])g();    $\C{// (4)}$
875([int, [int, int, int]])g()$\C{// (5)}$
876\end{cfa}
877
878(1) discards the last element of the return value and converts the second element to @double@.
879Since @int@ is effectively a 1-element tuple, (2) discards the second component of the second element of the return value of @g@.
880If @g@ is free of side effects, this expression is equivalent to @[(int)(g().0), (int)(g().1.0), (int)(g().2)]@.
881Since @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)]@).
882
883Note 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 requires more precise matching of types than allowed for function arguments and parameters.}.
884As such, (4) is invalid because the cast target type contains 4 components, while the source type contains only 3.
885Similarly, (5) is invalid because the cast @([int, int, int])(g().1)@ is invalid.
886That is, it is invalid to cast @[int, int]@ to @[int, int, int]@.
887\end{comment}
888
889
890\subsection{Polymorphism}
891
892Tuples also integrate with \CFA polymorphism as a kind of generic type.
893Due 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:
894\begin{cfa}
895forall( otype T, dtype U ) void f( T x, U * y );
896f( [5, "hello"] );
897\end{cfa}
898where @[5, "hello"]@ is flattened, giving argument list @5, "hello"@, and @T@ binds to @int@ and @U@ binds to @const char@.
899Tuples, however, may contain polymorphic components.
900For example, a plus operator can be written to sum two triples.
901\begin{cfa}
902forall( otype T | { T ?+?( T, T ); } ) [T, T, T] ?+?( [T, T, T] x, [T, T, T] y ) {
903        return [x.0 + y.0, x.1 + y.1, x.2 + y.2];
904}
905[int, int, int] x;
906int i1, i2, i3;
907[i1, i2, i3] = x + ([10, 20, 30]);
908\end{cfa}
909
910Flattening and restructuring conversions are also applied to tuple types in polymorphic type assertions.
911\begin{cfa}
912[int] f( [int, double], double );
913forall( otype T, otype U | { T f( T, U, U ); } ) void g( T, U );
914g( 5, 10.21 );
915\end{cfa}
916Hence, function parameter and return lists are flattened for the purposes of type unification allowing the example to pass expression resolution.
917This relaxation is possible by extending the thunk scheme described by Bilson~\cite{Bilson03}.
918% Whenever 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:
919% \begin{cfa}
920% int _thunk( int _p0, double _p1, double _p2 ) { return f( [_p0, _p1], _p2 ); }
921% \end{cfa}
922% so the thunk provides flattening and structuring conversions to inferred functions, improving the compatibility of tuples and polymorphism.
923% These thunks are generated locally using gcc nested-functions, rather hoisting them to the external scope, so they can easily access local state.
924
925
926\subsection{Variadic Tuples}
927\label{sec:variadic-tuples}
928
929To define variadic functions, \CFA adds a new kind of type parameter, @ttype@ (tuple type).
930Matching against a @ttype@ parameter consumes all remaining argument components and packages them into a tuple, binding to the resulting tuple of types.
931In 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.
932As such, @ttype@ variables are also called \newterm{argument packs}.
933
934Like variadic templates, @ttype@ polymorphic functions are primarily manipulated via recursion.
935Since nothing is known about a parameter pack by default, assertion parameters are key to doing anything meaningful.
936Unlike variadic templates, @ttype@ polymorphic functions can be separately compiled.
937For example, a generalized @sum@ function:
938\begin{cfa}
939int sum$\(_0\)$() { return 0; }
940forall( ttype Params | { int sum( Params ); } ) int sum$\(_1\)$( int x, Params rest ) {
941        return x + sum( rest );
942}
943sum( 10, 20, 30 );
944\end{cfa}
945Since @sum@\(_0\) does not accept any arguments, it is not a valid candidate function for the call @sum(10, 20, 30)@.
946In 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]@.
947The process continues until @Params@ is bound to @[]@, requiring an assertion @int sum()@, which matches @sum@\(_0\) and terminates the recursion.
948Effectively, 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))@.
949
950It is reasonable to take the @sum@ function a step further to enforce a minimum number of arguments:
951\begin{cfa}
952int sum( int x, int y ) { return x + y; }
953forall( ttype Params | { int sum( int, Params ); } ) int sum( int x, int y, Params rest ) {
954        return sum( x + y, rest );
955}
956\end{cfa}
957One more step permits the summation of any summable type with all arguments of the same type:
958\begin{cfa}
959trait summable( otype T ) {
960        T ?+?( T, T );
961};
962forall( otype R | summable( R ) ) R sum( R x, R y ) {
963        return x + y;
964}
965forall( otype R, ttype Params | summable(R) | { R sum(R, Params); } ) R sum(R x, R y, Params rest) {
966        return sum( x + y, rest );
967}
968\end{cfa}
969Unlike C variadic functions, it is unnecessary to hard code the number and expected types.
970Furthermore, this code is extendable for any user-defined type with a @?+?@ operator.
971Summing arbitrary heterogeneous lists is possible with similar code by adding the appropriate type variables and addition operators.
972
973It is also possible to write a type-safe variadic print function to replace @printf@:
974\begin{cfa}
975struct S { int x, y; };
976forall( otype T, ttype Params | { void print(T); void print(Params); } ) void print(T arg, Params rest) {
977        print(arg);  print(rest);
978}
979void print( const char * x ) { printf( "%s", x ); }
980void print( int x ) { printf( "%d", x ); }
981void print( S s ) { print( "{ ", s.x, ",", s.y, " }" ); }
982print( "s = ", (S){ 1, 2 }, "\n" );
983\end{cfa}
984This example showcases a variadic-template-like decomposition of the provided argument list.
985The individual @print@ functions allow printing a single element of a type.
986The polymorphic @print@ allows printing any list of types, where as each individual type has a @print@ function.
987The individual print functions can be used to build up more complicated @print@ functions, such as @S@, which cannot be done with @printf@ in C.
988This mechanism is used to seamlessly print tuples in the \CFA I/O library (see Section~\ref{s:IOLibrary}).
989
990Finally, it is possible to use @ttype@ polymorphism to provide arbitrary argument forwarding functions.
991For example, it is possible to write @new@ as a library function:
992\begin{cfa}
993forall( otype R, otype S ) void ?{}( pair(R, S) *, R, S );
994forall( dtype T, ttype Params | sized(T) | { void ?{}( T *, Params ); } ) T * new( Params p ) {
995        return ((T *)malloc()){ p };                    $\C{// construct into result of malloc}$
996}
997pair( int, char ) * x = new( 42, '!' );
998\end{cfa}
999The @new@ function provides the combination of type-safe @malloc@ with a \CFA constructor call, making it impossible to forget constructing dynamically allocated objects.
1000This function provides the type-safety of @new@ in \CC, without the need to specify the allocated type again, thanks to return-type inference.
1001
1002
1003\subsection{Implementation}
1004
1005Tuples are implemented in the \CFA translator via a transformation into \newterm{generic types}.
1006For each $N$, the first time an $N$-tuple is seen in a scope a generic type with $N$ type parameters is generated, \eg:
1007\begin{cfa}
1008[int, int] f() {
1009        [double, double] x;
1010        [int, double, int] y;
1011}
1012\end{cfa}
1013is transformed into:
1014\begin{cfa}
1015forall( dtype T0, dtype T1 | sized(T0) | sized(T1) ) struct _tuple2 {
1016        T0 field_0;  T1 field_1;                                        $\C{// generated before the first 2-tuple}$
1017};
1018_tuple2(int, int) f() {
1019        _tuple2(double, double) x;
1020        forall( dtype T0, dtype T1, dtype T2 | sized(T0) | sized(T1) | sized(T2) ) struct _tuple3 {
1021                T0 field_0;  T1 field_1;  T2 field_2;   $\C{// generated before the first 3-tuple}$
1022        };
1023        _tuple3(int, double, int) y;
1024}
1025\end{cfa}
1026Tuple expressions are then converted directly into compound literals, \eg @[5, 'x', 1.24]@ becomes @(_tuple3(int, char,@ @double)){ 5, 'x', 1.24 }@.
1027
1028\begin{comment}
1029Since tuples are essentially structures, tuple indexing expressions are just field accesses:
1030\begin{cfa}
1031void f(int, [double, char]);
1032[int, double] x;
1033
1034x.0+x.1;
1035printf("%d %g\n", x);
1036f(x, 'z');
1037\end{cfa}
1038Is transformed into:
1039\begin{cfa}
1040void f(int, _tuple2(double, char));
1041_tuple2(int, double) x;
1042
1043x.field_0+x.field_1;
1044printf("%d %g\n", x.field_0, x.field_1);
1045f(x.field_0, (_tuple2){ x.field_1, 'z' });
1046\end{cfa}
1047Note that due to flattening, @x@ used in the argument position is converted into the list of its fields.
1048In the call to @f@, the second and third argument components are structured into a tuple argument.
1049Similarly, tuple member expressions are recursively expanded into a list of member access expressions.
1050
1051Expressions that may contain side effects are made into \newterm{unique expressions} before being expanded by the flattening conversion.
1052Each unique expression is assigned an identifier and is guaranteed to be executed exactly once:
1053\begin{cfa}
1054void g(int, double);
1055[int, double] h();
1056g(h());
1057\end{cfa}
1058Internally, this expression is converted to two variables and an expression:
1059\begin{cfa}
1060void g(int, double);
1061[int, double] h();
1062
1063_Bool _unq0_finished_ = 0;
1064[int, double] _unq0;
1065g(
1066        (_unq0_finished_ ? _unq0 : (_unq0 = f(), _unq0_finished_ = 1, _unq0)).0,
1067        (_unq0_finished_ ? _unq0 : (_unq0 = f(), _unq0_finished_ = 1, _unq0)).1,
1068);
1069\end{cfa}
1070Since argument evaluation order is not specified by the C programming language, this scheme is built to work regardless of evaluation order.
1071The first time a unique expression is executed, the actual expression is evaluated and the accompanying boolean is set to true.
1072Every subsequent evaluation of the unique expression then results in an access to the stored result of the actual expression.
1073Tuple member expressions also take advantage of unique expressions in the case of possible impurity.
1074
1075Currently, the \CFA translator has a very broad, imprecise definition of impurity, where any function call is assumed to be impure.
1076This 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.
1077
1078The various kinds of tuple assignment, constructors, and destructors generate GNU C statement expressions.
1079A 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.
1080The 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.
1081However, 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.
1082\end{comment}
1083
1084
1085\section{Control Structures}
1086
1087\CFA identifies inconsistent, problematic, and missing control structures in C, and extends, modifies, and adds control structures to increase functionality and safety.
1088
1089
1090\subsection{\texorpdfstring{\protect\lstinline{if} Statement}{if Statement}}
1091
1092The @if@ expression allows declarations, similar to @for@ declaration expression:
1093\begin{cfa}
1094if ( int x = f() ) ...                                          $\C{// x != 0}$
1095if ( int x = f(), y = g() ) ...                         $\C{// x != 0 \&\& y != 0}$
1096if ( int x = f(), y = g(); `x < y` ) ...        $\C{// relational expression}$
1097\end{cfa}
1098Unless 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.}
1099The scope of the declaration(s) is local to the @if@ statement but exist within both the ``then'' and ``else'' clauses.
1100
1101
1102\subsection{\texorpdfstring{\protect\lstinline{switch} Statement}{switch Statement}}
1103
1104There 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.
1105
1106C 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.}.
1107\CFA provides a shorthand for a non-contiguous list:
1108\begin{cquote}
1109\lstDeleteShortInline@%
1110\begin{tabular}{@{}l|@{\hspace{2\parindentlnth}}l@{}}
1111\multicolumn{1}{c@{\hspace{2\parindentlnth}}}{\textbf{\CFA}}    & \multicolumn{1}{c}{\textbf{C}}        \\
1112\begin{cfa}
1113case 2, 10, 34, 42:
1114\end{cfa}
1115&
1116\begin{cfa}
1117case 2: case 10: case 34: case 42:
1118\end{cfa}
1119\end{tabular}
1120\lstMakeShortInline@%
1121\end{cquote}
1122for a contiguous list:\footnote{gcc has the same mechanism but awkward syntax, \lstinline@2 ...42@, as a space is required after a number, otherwise the first period is a decimal point.}
1123\begin{cquote}
1124\lstDeleteShortInline@%
1125\begin{tabular}{@{}l@{\hspace{2\parindentlnth}}l@{}}
1126\multicolumn{1}{c@{\hspace{2\parindentlnth}}}{\textbf{\CFA}}    & \multicolumn{1}{c}{\textbf{C}}        \\
1127\begin{cfa}
1128case 2~42:
1129\end{cfa}
1130&
1131\begin{cfa}
1132case 2: case 3: ... case 41: case 42:
1133\end{cfa}
1134\end{tabular}
1135\lstMakeShortInline@%
1136\end{cquote}
1137and a combination:
1138\begin{cfa}
1139case -12~-4, -1~5, 14~21, 34~42:
1140\end{cfa}
1141
1142C allows placement of @case@ clauses \emph{within} statements nested in the @switch@ body (called Duff's device~\cite{Duff83});
1143\begin{cfa}
1144switch ( i ) {
1145  case 0:
1146        for ( int i = 0; i < 10; i += 1 ) {
1147                ...
1148  `case 1:`             // no initialization of loop index
1149                ...
1150        }
1151}
1152\end{cfa}
1153\CFA precludes this form of transfer \emph{into} a control structure because it causes undefined behaviour, especially with respect to missed initialization, and provides very limited functionality.
1154
1155C allows placement of declaration within the @switch@ body and unreachable code at the start, resulting in undefined behaviour:
1156\begin{cfa}
1157switch ( x ) {
1158        `int y = 1;`                                                    $\C{// unreachable initialization}$
1159        `x = 7;`                                                                $\C{// unreachable code without label/branch}$
1160  case 0:
1161        ...
1162        `int z = 0;`                                                    $\C{// unreachable initialization, cannot appear after case}$
1163        z = 2;
1164  case 1:
1165        `x = z;`                                                                $\C{// without fall through, z is undefined}$
1166}
1167\end{cfa}
1168\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.
1169\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.
1170
1171C @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};
1172@case@ clauses are made disjoint by the @break@ statement.
1173While fall through \emph{is} a useful form of control flow, it does not match well with programmer intuition, resulting in errors from missing @break@ statements.
1174For 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}), similar to Go.
1175
1176\begin{figure}
1177\centering
1178\lstDeleteShortInline@%
1179\begin{tabular}{@{}l|@{\hspace{\parindentlnth}}l@{}}
1180\multicolumn{1}{c|@{\hspace{\parindentlnth}}}{\textbf{\CFA}}    & \multicolumn{1}{c}{\textbf{C}}        \\
1181\begin{cfa}
1182`choose` ( day ) {
1183  case Mon~Thu:  // program
1184
1185  case Fri:    // program
1186        wallet += pay;
1187        `fallthrough;`
1188  case Sat:   // party
1189        wallet -= party;
1190
1191  case Sun:  // rest
1192
1193  default:    // print error
1194}
1195\end{cfa}
1196&
1197\begin{cfa}
1198switch ( day ) {
1199  case Mon: case Tue: case Wed: case Thu:  // program
1200        `break;`
1201  case Fri:    // program
1202        wallet += pay;
1203
1204  case Sat:   // party
1205        wallet -= party;
1206        `break;`
1207  case Sun:  // rest
1208        `break;`
1209  default:    // print error
1210}
1211\end{cfa}
1212\end{tabular}
1213\lstMakeShortInline@%
1214\caption{\lstinline|choose| versus \lstinline|switch| Statements}
1215\label{f:ChooseSwitchStatements}
1216\end{figure}
1217
1218Finally, Figure~\ref{f:FallthroughStatement} shows @fallthrough@ may appear in contexts other than terminating a @case@ clause, and have an explicit transfer label allowing separate cases but common final-code for a set of cases.
1219The target label must be below the @fallthrough@ and may not be nested in a control structure, \ie @fallthrough@ cannot form a loop, and the target label must be at the same or higher level as the containing @case@ clause and located at the same level as a @case@ clause;
1220the target label may be case @default@, but only associated with the current @switch@/@choose@ statement.
1221
1222\begin{figure}
1223\centering
1224\lstDeleteShortInline@%
1225\begin{tabular}{@{}l|@{\hspace{\parindentlnth}}l@{}}
1226\multicolumn{1}{c|@{\hspace{\parindentlnth}}}{\textbf{non-terminator}}  & \multicolumn{1}{c}{\textbf{target label}}     \\
1227\begin{cfa}
1228choose ( ... ) {
1229  case 3:
1230        if ( ... ) {
1231                ... `fallthrough;`  // goto case 4
1232        } else {
1233                ...
1234        }
1235        // implicit break
1236  case 4:
1237\end{cfa}
1238&
1239\begin{cfa}
1240choose ( ... ) {
1241  case 3:
1242        ... `fallthrough common;`
1243  case 4:
1244        ... `fallthrough common;`
1245  `common`: // below fallthrough at same level as case clauses
1246        ...      // common code for cases 3 and 4
1247        // implicit break
1248  case 4:
1249\end{cfa}
1250\end{tabular}
1251\lstMakeShortInline@%
1252\caption{\lstinline|fallthrough| Statement}
1253\label{f:FallthroughStatement}
1254\end{figure}
1255
1256
1257\subsection{\texorpdfstring{Labelled \protect\lstinline{continue} / \protect\lstinline{break}}{Labelled continue / break}}
1258
1259While C provides @continue@ and @break@ statements for altering control flow, both are restricted to one level of nesting for a particular control structure.
1260Unfortunately, this restriction forces programmers to use @goto@ to achieve the equivalent control-flow for more than one level of nesting.
1261To 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.
1262For both @continue@ and @break@, the target label must be directly associated with a @for@, @while@ or @do@ statement;
1263for @break@, the target label can also be associated with a @switch@, @if@ or compound (@{}@) statement.
1264Figure~\ref{f:MultiLevelExit} shows @continue@ and @break@ indicating the specific control structure, and the corresponding C program using only @goto@ and labels.
1265The innermost loop has 7 exit points, which cause continuation or termination of one or more of the 7 nested control-structures.
1266
1267\begin{figure}
1268\lstDeleteShortInline@%
1269\begin{tabular}{@{\hspace{\parindentlnth}}l|@{\hspace{\parindentlnth}}l@{\hspace{\parindentlnth}}l@{}}
1270\multicolumn{1}{@{\hspace{\parindentlnth}}c|@{\hspace{\parindentlnth}}}{\textbf{\CFA}}  & \multicolumn{1}{@{\hspace{\parindentlnth}}c}{\textbf{C}}      \\
1271\begin{cfa}
1272`LC:` {
1273        ... $declarations$ ...
1274        `LS:` switch ( ... ) {
1275          case 3:
1276                `LIF:` if ( ... ) {
1277                        `LF:` for ( ... ) {
1278                                ... break `LC`; ...
1279                                ... break `LS`; ...
1280                                ... break `LIF`; ...
1281                                ... continue `LF;` ...
1282                                ... break `LF`; ...
1283                        } // for
1284                } else {
1285                        ... break `LIF`; ...
1286                } // if
1287        } // switch
1288} // compound
1289\end{cfa}
1290&
1291\begin{cfa}
1292{
1293        ... $declarations$ ...
1294        switch ( ... ) {
1295          case 3:
1296                if ( ... ) {
1297                        for ( ... ) {
1298                                ... goto `LC`; ...
1299                                ... goto `LS`; ...
1300                                ... goto `LIF`; ...
1301                                ... goto `LFC`; ...
1302                                ... goto `LFB`; ...
1303                          `LFC:` ; } `LFB:` ;
1304                } else {
1305                        ... goto `LIF`; ...
1306                } `LIF:` ;
1307        } `LS:` ;
1308} `LC:` ;
1309\end{cfa}
1310&
1311\begin{cfa}
1312
1313
1314
1315
1316
1317
1318
1319// terminate compound
1320// terminate switch
1321// terminate if
1322// continue loop
1323// terminate loop
1324
1325
1326
1327// terminate if
1328
1329\end{cfa}
1330\end{tabular}
1331\lstMakeShortInline@%
1332\caption{Multi-level Exit}
1333\label{f:MultiLevelExit}
1334\end{figure}
1335
1336With respect to safety, both labelled @continue@ and @break@ are a @goto@ restricted in the following ways:
1337\begin{itemize}
1338\item
1339They cannot create a loop, which means only the looping constructs cause looping.
1340This restriction means all situations resulting in repeated execution are clearly delineated.
1341\item
1342They cannot branch into a control structure.
1343This restriction prevents missing declarations and/or initializations at the start of a control structure resulting in undefined behaviour.
1344\end{itemize}
1345The 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.
1346Furthermore, 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.
1347With @goto@, the label is at the end of the control structure, which fails to convey this important clue early enough to the reader.
1348Finally, using an explicit target for the transfer instead of an implicit target allows new constructs to be added or removed without affecting existing constructs.
1349Otherwise, the implicit targets of the current @continue@ and @break@, \ie the closest enclosing loop or @switch@, change as certain constructs are added or removed.
1350
1351
1352\subsection{Exception Handling}
1353
1354The following framework for \CFA exception-handling is in place, excluding some runtime type-information and virtual functions.
1355\CFA provides two forms of exception handling: \newterm{fix-up} and \newterm{recovery} (see Figure~\ref{f:CFAExceptionHandling})~\cite{Buhr92b,Buhr00a}.
1356Both 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.
1357\CFA restricts exception types to those defined by aggregate type @exception@.
1358The 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@).
1359If @resume@ or @throw@ have no exception type, it is a reresume/rethrow, meaning the currently exception continues propagation.
1360If there is no current exception, the reresume/rethrow results in a runtime error.
1361
1362\begin{figure}
1363\begin{cquote}
1364\lstDeleteShortInline@%
1365\begin{tabular}{@{}l|@{\hspace{\parindentlnth}}l@{}}
1366\multicolumn{1}{c|@{\hspace{\parindentlnth}}}{\textbf{Resumption}}      & \multicolumn{1}{c}{\textbf{Termination}}      \\
1367\begin{cfa}
1368`exception R { int fix; };`
1369void f() {
1370        R r;
1371        ... `resume( r );` ...
1372        ... r.fix // control returns here after handler
1373}
1374`try` {
1375        ... f(); ...
1376} `catchResume( R r )` {
1377        ... r.fix = ...; // return correction to raise
1378} // dynamic return to _Resume
1379\end{cfa}
1380&
1381\begin{cfa}
1382`exception T {};`
1383void f() {
1384
1385        ... `throw( T{} );` ...
1386        // control does NOT return here after handler
1387}
1388`try` {
1389        ... f(); ...
1390} `catch( T t )` {
1391        ... // recover and continue
1392} // static return to next statement
1393\end{cfa}
1394\end{tabular}
1395\lstMakeShortInline@%
1396\end{cquote}
1397\caption{\CFA Exception Handling}
1398\label{f:CFAExceptionHandling}
1399\end{figure}
1400
1401The set of exception types in a list of catch clause may include both a resumption and termination handler:
1402\begin{cfa}
1403try {
1404        ... resume( `R{}` ); ...
1405} catchResume( `R` r ) { ... throw( R{} ); ... } $\C{\color{red}// H1}$
1406   catch( `R` r ) { ... }                                       $\C{\color{red}// H2}$
1407
1408\end{cfa}
1409The resumption propagation raises @R@ and the stack is not unwound;
1410the exception is caught by the @catchResume@ clause and handler H1 is invoked.
1411The termination propagation in handler H1 raises @R@ and the stack is unwound;
1412the exception is caught by the @catch@ clause and handler H2 is invoked.
1413The termination handler is available because the resumption propagation did not unwind the stack.
1414
1415An additional feature is conditional matching in a catch clause:
1416\begin{cfa}
1417try {
1418        ... write( `datafile`, ... ); ...               $\C{// may throw IOError}$
1419        ... write( `logfile`, ... ); ...
1420} catch ( IOError err; `err.file == datafile` ) { ... } $\C{// handle datafile error}$
1421   catch ( IOError err; `err.file == logfile` ) { ... } $\C{// handle logfile error}$
1422   catch ( IOError err ) { ... }                        $\C{// handler error from other files}$
1423\end{cfa}
1424where the throw inserts the failing file-handle into the I/O exception.
1425Conditional catch cannot be trivially mimicked by other mechanisms because once an exception is caught, handler clauses in that @try@ statement are no longer eligible..
1426
1427The resumption raise can specify an alternate stack on which to raise an exception, called a \newterm{nonlocal raise}:
1428\begin{cfa}
1429resume( $\emph{exception-type}$, $\emph{alternate-stack}$ )
1430resume( $\emph{alternate-stack}$ )
1431\end{cfa}
1432These overloads of @resume@ raise the specified exception or the currently propagating exception (reresume) at another \CFA coroutine or task~\cite{Delisle18}.
1433Nonlocal 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 handler returns.
1434
1435To facilitate nonlocal raise, \CFA provides dynamic enabling and disabling of nonlocal exception-propagation.
1436The constructs for controlling propagation of nonlocal exceptions are the @enable@ and the @disable@ blocks:
1437\begin{cquote}
1438\lstDeleteShortInline@%
1439\begin{tabular}{@{}l@{\hspace{2\parindentlnth}}l@{}}
1440\begin{cfa}
1441enable $\emph{exception-type-list}$ {
1442        // allow non-local raise
1443}
1444\end{cfa}
1445&
1446\begin{cfa}
1447disable $\emph{exception-type-list}$ {
1448        // disallow non-local raise
1449}
1450\end{cfa}
1451\end{tabular}
1452\lstMakeShortInline@%
1453\end{cquote}
1454The arguments for @enable@/@disable@ specify the exception types allowed to be propagated or postponed, respectively.
1455Specifying no exception type is shorthand for specifying all exception types.
1456Both @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.
1457Coroutines and tasks start with non-local exceptions disabled, allowing handlers to be put in place, before non-local exceptions are explicitly enabled.
1458\begin{cfa}
1459void main( mytask & t ) {                                       $\C{// thread starts here}$
1460        // non-local exceptions disabled
1461        try {                                                                   $\C{// establish handles for non-local exceptions}$
1462                enable {                                                        $\C{// allow non-local exception delivery}$
1463                        // task body
1464                }
1465        // appropriate catchResume/catch handlers
1466        }
1467}
1468\end{cfa}
1469
1470Finally, \CFA provides a Java like  @finally@ clause after the catch clauses:
1471\begin{cfa}
1472try {
1473        ... f(); ...
1474// catchResume or catch clauses
1475} `finally` {
1476        // house keeping
1477}
1478\end{cfa}
1479The finally clause is always executed, i.e., if the try block ends normally or if an exception is raised.
1480If an exception is raised and caught, the handler is run before the finally clause.
1481Like a destructor (see Section~\ref{s:ConstructorsDestructors}), a finally clause can raise an exception but not if there is an exception being propagated.
1482Mimicking the @finally@ clause with mechanisms like RAII is non-trivial when there are multiple types and local accesses.
1483
1484
1485\subsection{\texorpdfstring{\protect\lstinline{with} Statement}{with Statement}}
1486\label{s:WithStatement}
1487
1488Heterogeneous data is often aggregated into a structure/union.
1489To reduce syntactic noise, \CFA provides a @with@ statement (see Pascal~\cite[\S~4.F]{Pascal}) to elide aggregate field-qualification by opening a scope containing the field identifiers.
1490\begin{cquote}
1491\vspace*{-\baselineskip}%???
1492\lstDeleteShortInline@%
1493\begin{cfa}
1494struct S { char c; int i; double d; };
1495struct T { double m, n; };
1496// multiple aggregate parameters
1497\end{cfa}
1498\begin{tabular}{@{}l@{\hspace{\parindentlnth}}|@{\hspace{\parindentlnth}}l@{}}
1499\begin{cfa}
1500void f( S & s, T & t ) {
1501        `s.`c; `s.`i; `s.`d;
1502        `t.`m; `t.`n;
1503}
1504\end{cfa}
1505&
1506\begin{cfa}
1507void f( S & s, T & t ) `with ( s, t )` {
1508        c; i; d;                // no qualification
1509        m; n;
1510}
1511\end{cfa}
1512\end{tabular}
1513\lstMakeShortInline@%
1514\end{cquote}
1515Object-oriented programming languages only provide implicit qualification for the receiver.
1516
1517In detail, the @with@ statement has the form:
1518\begin{cfa}
1519$\emph{with-statement}$:
1520        'with' '(' $\emph{expression-list}$ ')' $\emph{compound-statement}$
1521\end{cfa}
1522and may appear as the body of a function or nested within a function body.
1523Each expression in the expression-list provides a type and object.
1524The type must be an aggregate type.
1525(Enumerations are already opened.)
1526The object is the implicit qualifier for the open structure-fields.
1527
1528All expressions in the expression list are open in parallel within the compound statement, which is different from Pascal, which nests the openings from left to right.
1529The difference between parallel and nesting occurs for fields with the same name and type:
1530\begin{cfa}
1531struct S { int `i`; int j; double m; } s, w;
1532struct T { int `i`; int k; int m; } t, w;
1533with ( s, t ) {
1534        j + k;                                                                  $\C{// unambiguous, s.j + t.k}$
1535        m = 5.0;                                                                $\C{// unambiguous, s.m = 5.0}$
1536        m = 1;                                                                  $\C{// unambiguous, t.m = 1}$
1537        int a = m;                                                              $\C{// unambiguous, a = t.m }$
1538        double b = m;                                                   $\C{// unambiguous, b = s.m}$
1539        int c = s.i + t.i;                                              $\C{// unambiguous, qualification}$
1540        (double)m;                                                              $\C{// unambiguous, cast s.m}$
1541}
1542\end{cfa}
1543For parallel semantics, both @s.i@ and @t.i@ are visible, so @i@ is ambiguous without qualification;
1544for nested semantics, @t.i@ hides @s.i@, so @i@ implies @t.i@.
1545\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.
1546Qualification or a cast is used to disambiguate.
1547
1548There is an interesting problem between parameters and the function-body @with@, \eg:
1549\begin{cfa}
1550void ?{}( S & s, int i ) with ( s ) {           $\C{// constructor}$
1551        `s.i = i;`  j = 3;  m = 5.5;                    $\C{// initialize fields}$
1552}
1553\end{cfa}
1554Here, 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@.
1555To solve this problem, parameters are treated like an initialized aggregate:
1556\begin{cfa}
1557struct Params {
1558        S & s;
1559        int i;
1560} params;
1561\end{cfa}
1562and implicitly opened \emph{after} a function-body open, to give them higher priority:
1563\begin{cfa}
1564void ?{}( S & s, int `i` ) with ( s ) `{` `with( $\emph{\color{red}params}$ )` {
1565        s.i = `i`; j = 3; m = 5.5;
1566} `}`
1567\end{cfa}
1568Finally, a cast may be used to disambiguate among overload variables in a @with@ expression:
1569\begin{cfa}
1570with ( w ) { ... }                                                      $\C{// ambiguous, same name and no context}$
1571with ( (S)w ) { ... }                                           $\C{// unambiguous, cast}$
1572\end{cfa}
1573and @with@ expressions may be complex expressions with type reference (see Section~\ref{s:References}) to aggregate:
1574\begin{cfa}
1575struct S { int i, j; } sv;
1576with ( sv ) {                                                           $\C{// implicit reference}$
1577        S & sr = sv;
1578        with ( sr ) {                                                   $\C{// explicit reference}$
1579                S * sp = &sv;
1580                with ( *sp ) {                                          $\C{// computed reference}$
1581                        i = 3; j = 4;                                   $\C{\color{red}// sp--{\textgreater}i, sp--{\textgreater}j}$
1582                }
1583                i = 2; j = 3;                                           $\C{\color{red}// sr.i, sr.j}$
1584        }
1585        i = 1; j = 2;                                                   $\C{\color{red}// sv.i, sv.j}$
1586}
1587\end{cfa}
1588
1589Collectively, these control-structure enhancements reduce programmer burden and increase readability and safety.
1590
1591
1592\section{Declarations}
1593
1594Declarations in C have weaknesses and omissions.
1595\CFA attempts to correct and add to C declarations, while ensuring \CFA subjectively ``feels like'' C.
1596An 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.
1597Maintaining 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.
1598Nevertheless, some features from other approaches are undeniably convenient;
1599\CFA attempts to adapt these features to the C paradigm.
1600
1601
1602\subsection{Alternative Declaration Syntax}
1603
1604C declaration syntax is notoriously confusing and error prone.
1605For example, many C programmers are confused by a declaration as simple as:
1606\begin{cquote}
1607\lstDeleteShortInline@%
1608\begin{tabular}{@{}ll@{}}
1609\begin{cfa}
1610int * x[5]
1611\end{cfa}
1612&
1613\raisebox{-0.75\totalheight}{\input{Cdecl}}
1614\end{tabular}
1615\lstMakeShortInline@%
1616\end{cquote}
1617Is this an array of 5 pointers to integers or a pointer to an array of 5 integers?
1618If there is any doubt, it implies productivity and safety issues even for basic programs.
1619Another 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.
1620For example, a function returning a pointer to an array of integers is defined and used in the following way:
1621\begin{cfa}
1622int `(*`f`())[`5`]` {...};                                      $\C{// definition}$
1623 ... `(*`f`())[`3`]` += 1;                                      $\C{// usage}$
1624\end{cfa}
1625Essentially, the return type is wrapped around the function name in successive layers (like an onion).
1626While attempting to make the two contexts consistent is a laudable goal, it has not worked out in practice.
1627
1628\CFA provides its own type, variable and function declarations, using a different syntax~\cite[pp.~856--859]{Buhr94a}.
1629The new declarations place qualifiers to the left of the base type, while C declarations place qualifiers to the right.
1630The qualifiers have the same meaning but are ordered left to right to specify a variable's type.
1631\begin{cquote}
1632\lstDeleteShortInline@%
1633\begin{tabular}{@{}l@{\hspace{2\parindentlnth}}l@{\hspace{2\parindentlnth}}l@{}}
1634\multicolumn{1}{c@{\hspace{2\parindentlnth}}}{\textbf{\CFA}}    & \multicolumn{1}{c}{\textbf{C}}        \\
1635\begin{cfa}
1636`[5] *` int x1;
1637`* [5]` int x2;
1638`[* [5] int]` f( int p );
1639\end{cfa}
1640&
1641\begin{cfa}
1642int `*` x1 `[5]`;
1643int `(*`x2`)[5]`;
1644`int (*`f( int p )`)[5]`;
1645\end{cfa}
1646&
1647\begin{cfa}
1648// array of 5 pointers to int
1649// pointer to array of 5 int
1650// function returning pointer to array of 5 int and taking int
1651\end{cfa}
1652\end{tabular}
1653\lstMakeShortInline@%
1654\end{cquote}
1655The only exception is bit field specification, which always appear to the right of the base type.
1656% 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.
1657However, unlike C, \CFA type declaration tokens are distributed across all variables in the declaration list.
1658For instance, variables @x@ and @y@ of type pointer to integer are defined in \CFA as follows:
1659\begin{cquote}
1660\lstDeleteShortInline@%
1661\begin{tabular}{@{}l@{\hspace{2\parindentlnth}}l@{}}
1662\multicolumn{1}{c@{\hspace{2\parindentlnth}}}{\textbf{\CFA}}    & \multicolumn{1}{c}{\textbf{C}}        \\
1663\begin{cfa}
1664`*` int x, y;
1665int z;
1666\end{cfa}
1667&
1668\begin{cfa}
1669int `*`x, `*`y, z;
1670
1671\end{cfa}
1672\end{tabular}
1673\lstMakeShortInline@%
1674\end{cquote}
1675% The downside of the \CFA semantics is the need to separate regular and pointer declarations.
1676The separation of regular and pointer declarations by \CFA declarations enforces greater clarity with only slightly more syntax.
1677
1678\begin{comment}
1679Other examples are:
1680\begin{cquote}
1681\lstDeleteShortInline@%
1682\begin{tabular}{@{}l@{\hspace{2\parindentlnth}}l@{\hspace{2\parindentlnth}}l@{}}
1683\multicolumn{1}{c@{\hspace{2\parindentlnth}}}{\textbf{\CFA}}    & \multicolumn{1}{c@{\hspace{2\parindentlnth}}}{\textbf{C}}     \\
1684\begin{cfa}
1685[ 5 ] int z;
1686[ 5 ] * char w;
1687* [ 5 ] double v;
1688struct s {
1689        int f0:3;
1690        * int f1;
1691        [ 5 ] * int f2;
1692};
1693\end{cfa}
1694&
1695\begin{cfa}
1696int z[ 5 ];
1697char * w[ 5 ];
1698double (* v)[ 5 ];
1699struct s {
1700        int f0:3;
1701        int * f1;
1702        int * f2[ 5 ]
1703};
1704\end{cfa}
1705&
1706\begin{cfa}
1707// array of 5 integers
1708// array of 5 pointers to char
1709// pointer to array of 5 doubles
1710
1711// common bit field syntax
1712
1713
1714
1715\end{cfa}
1716\end{tabular}
1717\lstMakeShortInline@%
1718\end{cquote}
1719\end{comment}
1720
1721All 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:
1722\begin{cquote}
1723\lstDeleteShortInline@%
1724\begin{tabular}{@{}l@{\hspace{2\parindentlnth}}l@{\hspace{2\parindentlnth}}l@{}}
1725\multicolumn{1}{c@{\hspace{2\parindentlnth}}}{\textbf{\CFA}}    & \multicolumn{1}{c@{\hspace{2\parindentlnth}}}{\textbf{C}}     \\
1726\begin{cfa}
1727extern const * const int x;
1728static const * [5] const int y;
1729\end{cfa}
1730&
1731\begin{cfa}
1732int extern const * const x;
1733static const int (* const y)[5]
1734\end{cfa}
1735&
1736\begin{cfa}
1737// external const pointer to const int
1738// internal const pointer to array of 5 const int
1739\end{cfa}
1740\end{tabular}
1741\lstMakeShortInline@%
1742\end{cquote}
1743Specifiers must appear at the start of a \CFA function declaration\footnote{\label{StorageClassSpecifier}
1744The 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}}.
1745
1746The new declaration syntax can be used in other contexts where types are required, \eg casts and the pseudo-function @sizeof@:
1747\begin{cquote}
1748\lstDeleteShortInline@%
1749\begin{tabular}{@{}l@{\hspace{2\parindentlnth}}l@{}}
1750\multicolumn{1}{c@{\hspace{2\parindentlnth}}}{\textbf{\CFA}}    & \multicolumn{1}{c}{\textbf{C}}        \\
1751\begin{cfa}
1752y = (* int)x;
1753i = sizeof([ 5 ] * int);
1754\end{cfa}
1755&
1756\begin{cfa}
1757y = (int *)x;
1758i = sizeof(int * [ 5 ]);
1759\end{cfa}
1760\end{tabular}
1761\lstMakeShortInline@%
1762\end{cquote}
1763
1764The syntax of the new function-prototype declaration follows directly from the new function-definition syntax;
1765as well, parameter names are optional, \eg:
1766\begin{cfa}
1767[ int x ] f ( /* void */ );             $\C[2.5in]{// returning int with no parameters}$
1768[ int x ] f (...);                              $\C{// returning int with unknown parameters}$
1769[ * int ] g ( int y );                  $\C{// returning pointer to int with int parameter}$
1770[ void ] h ( int, char );               $\C{// returning no result with int and char parameters}$
1771[ * int, int ] j ( int );               $\C{// returning pointer to int and int with int parameter}$
1772\end{cfa}
1773This syntax allows a prototype declaration to be created by cutting and pasting source text from the function-definition header (or vice versa).
1774Like 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:
1775\begin{cquote}
1776\lstDeleteShortInline@%
1777\begin{tabular}{@{}l@{\hspace{2\parindentlnth}}l@{}}
1778\multicolumn{1}{c@{\hspace{2\parindentlnth}}}{\textbf{\CFA}}    & \multicolumn{1}{c}{\textbf{C}}        \\
1779\begin{cfa}
1780[double] foo(), foo( int ), foo( double ) {...}
1781\end{cfa}
1782&
1783\begin{cfa}
1784double foo1( void ), foo2( int ), foo3( double );
1785\end{cfa}
1786\end{tabular}
1787\lstMakeShortInline@%
1788\end{cquote}
1789where \CFA allows the last function in the list to define its body.
1790
1791The syntax for pointers to \CFA functions specifies the pointer name on the right, \eg:
1792\begin{cfa}
1793* [ int x ] () fp;                              $\C{// pointer to function returning int with no parameters}$
1794* [ * int ] ( int y ) gp;               $\C{// pointer to function returning pointer to int with int parameter}$
1795* [ ] ( int, char ) hp;                 $\C{// pointer to function returning no result with int and char parameters}$
1796* [ * int, int ] ( int ) jp;    $\C{// pointer to function returning pointer to int and int with int parameter}$
1797\end{cfa}
1798Note, the name of the function pointer is specified last, as for other variable declarations.
1799
1800Finally, new \CFA declarations may appear together with C declarations in the same program block, but cannot be mixed within a specific declaration.
1801Therefore, a programmer has the option of either continuing to use traditional C declarations or take advantage of the new style.
1802Clearly, both styles need to be supported for some time due to existing C-style header-files, particularly for UNIX-like systems.
1803
1804
1805\subsection{References}
1806\label{s:References}
1807
1808All variables in C have an \newterm{address}, a \newterm{value}, and a \newterm{type};
1809at 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.
1810The C type-system does not always track the relationship between a value and its address;
1811a 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'').
1812For 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.
1813Despite 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.
1814
1815Within a lexical scope, lvalue expressions have an \newterm{address interpretation} for writing a value or a \newterm{value interpretation} to read a value.
1816For example, in @x = y@, @x@ has an address interpretation, while @y@ has a value interpretation.
1817While 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.
1818In C, for any type @T@ there is a pointer type @T *@, the value of which is the address of a value of type @T@.
1819A 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 @&?@.
1820
1821\begin{cfa}
1822int x = 1, y = 2, * p1, * p2, ** p3;
1823p1 = &x;                                                                        $\C{// p1 points to x}$
1824p2 = &y;                                                                        $\C{// p2 points to y}$
1825p3 = &p1;                                                                       $\C{// p3 points to p1}$
1826*p2 = ((*p1 + *p2) * (**p3 - *p1)) / (**p3 - 15);
1827\end{cfa}
1828
1829Unfortunately, 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.
1830For 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.
1831However, 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.
1832To solve these problems, \CFA introduces reference types @T &@;
1833a @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:
1834
1835\begin{cfa}
1836int x = 1, y = 2, & r1, & r2, && r3;
1837&r1 = &x;                                                                       $\C{// r1 points to x}$
1838&r2 = &y;                                                                       $\C{// r2 points to y}$
1839&&r3 = &&r1;                                                            $\C{// r3 points to r2}$
1840r2 = ((r1 + r2) * (r3 - r1)) / (r3 - 15);       $\C{// implicit dereferencing}$
1841\end{cfa}
1842
1843Except for auto-dereferencing by the compiler, this reference example is exactly the same as the previous pointer example.
1844Hence, 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.
1845One 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:
1846
1847\begin{cfa}
1848`*`r2 = ((`*`r1 + `*`r2) * (`**`r3 - `*`r1)) / (`**`r3 - 15);
1849\end{cfa}
1850
1851References in \CFA are similar to those in \CC, with important improvements, which can be seen in the example above.
1852Firstly, \CFA does not forbid references to references.
1853This provides a much more orthogonal design for library implementors, obviating the need for workarounds such as @std::reference_wrapper@.
1854Secondly, \CFA references are rebindable, whereas \CC references have a fixed address.
1855Rebinding allows \CFA references to be default-initialized (\eg to a null pointer\footnote{
1856While 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, \eg using a pointer after its storage is deleted.}) and point to different addresses throughout their lifetime, like pointers.
1857Rebinding is accomplished by extending the existing syntax and semantics of the address-of operator in C.
1858
1859In 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.
1860In \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.
1861The 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.
1862This rebinding occurs to an arbitrary depth of reference nesting;
1863loosely speaking, nested address-of operators produce a nested lvalue pointer up to the depth of the reference.
1864These 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@.
1865More precisely:
1866\begin{itemize}
1867\item
1868if @R@ is an rvalue of type {@T &@$_1 \cdots$@ &@$_r$} where $r \ge 1$ references (@&@ symbols) then @&R@ has type {@T `*`&@$_{\color{red}2} \cdots$@ &@$_{\color{red}r}$}, \\ \ie @T@ pointer with $r-1$ references (@&@ symbols).
1869       
1870\item
1871if @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).
1872\end{itemize}
1873Since 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.
1874
1875By analogy to pointers, \CFA references also allow cv-qualifiers such as @const@:
1876\begin{cfa}
1877const int cx = 5;                                                       $\C{// cannot change cx}$
1878const int & cr = cx;                                            $\C{// cannot change cr's referred value}$
1879&cr = &cx;                                                                      $\C{// rebinding cr allowed}$
1880cr = 7;                                                                         $\C{// ERROR, cannot change cr}$
1881int & const rc = x;                                                     $\C{// must be initialized, like in \CC}$
1882&rc = &x;                                                                       $\C{// ERROR, cannot rebind rc}$
1883rc = 7;                                                                         $\C{// x now equal to 7}$
1884\end{cfa}
1885Given that a reference is meant to represent a lvalue, \CFA provides some syntactic shortcuts when initializing references.
1886There are three initialization contexts in \CFA: declaration initialization, argument/parameter binding, and return/temporary binding.
1887In each of these contexts, the address-of operator on the target lvalue is elided.
1888The 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@.
1889
1890More 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;
1891this conversion is used in any context in \CFA where an implicit conversion is allowed.
1892Similarly, 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.
1893The final reference conversion included in \CFA is ``rvalue-to-reference'' conversion, implemented by means of an implicit temporary.
1894When 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.
1895\begin{cfa}
1896struct S { double x, y; };
1897int x, y;
1898void f( int & i, int & j, S & s, int v[] );
1899f( 3, x + y, (S){ 1.0, 7.0 }, (int [3]){ 1, 2, 3 } ); $\C{// pass rvalue to lvalue \(\Rightarrow\) implicit temporary}$
1900\end{cfa}
1901This 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.
1902\CC allows a similar binding, but only for @const@ references; the more general semantics of \CFA are an attempt to avoid the \newterm{const poisoning} problem~\cite{Taylor10}, in which addition of a @const@ qualifier to one reference requires a cascading chain of added qualifiers.
1903
1904
1905\subsection{Type Nesting}
1906
1907Nested 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@).
1908Java 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.
1909Since \CFA in not object-oriented, adopting dynamic scoping does not make sense;
1910instead \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}).
1911\begin{figure}
1912\centering
1913\lstDeleteShortInline@%
1914\begin{tabular}{@{}l@{\hspace{3em}}l|l@{}}
1915\multicolumn{1}{c@{\hspace{3em}}}{\textbf{C Type Nesting}}      & \multicolumn{1}{c|}{\textbf{C Implicit Hoisting}}     & \multicolumn{1}{c}{\textbf{\CFA}}     \\
1916\begin{cfa}
1917struct S {
1918        enum C { R, G, B };
1919        struct T {
1920                union U { int i, j; };
1921                enum C c;
1922                short int i, j;
1923        };
1924        struct T t;
1925} s;
1926
1927int rtn() {
1928        s.t.c = R;
1929        struct T t = { R, 1, 2 };
1930        enum C c;
1931        union U u;
1932}
1933\end{cfa}
1934&
1935\begin{cfa}
1936enum C { R, G, B };
1937union U { int i, j; };
1938struct T {
1939        enum C c;
1940        short int i, j;
1941};
1942struct S {
1943        struct T t;
1944} s;
1945       
1946
1947
1948
1949
1950
1951
1952\end{cfa}
1953&
1954\begin{cfa}
1955struct S {
1956        enum C { R, G, B };
1957        struct T {
1958                union U { int i, j; };
1959                enum C c;
1960                short int i, j;
1961        };
1962        struct T t;
1963} s;
1964
1965int rtn() {
1966        s.t.c = `S.`R;  // type qualification
1967        struct `S.`T t = { `S.`R, 1, 2 };
1968        enum `S.`C c;
1969        union `S.T.`U u;
1970}
1971\end{cfa}
1972\end{tabular}
1973\lstMakeShortInline@%
1974\caption{Type Nesting / Qualification}
1975\label{f:TypeNestingQualification}
1976\end{figure}
1977In the C left example, types @C@, @U@ and @T@ are implicitly hoisted outside of type @S@ into the containing block scope.
1978In the \CFA right example, the types are not hoisted and accessible.
1979
1980
1981\subsection{Constructors and Destructors}
1982\label{s:ConstructorsDestructors}
1983
1984One 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.
1985However, this manual approach is verbose, and it is useful to manage resources other than memory (\eg file handles) using the same mechanism as memory.
1986\CC addresses these issues using Resource Aquisition Is Initialization (RAII), implemented by means of \newterm{constructor} and \newterm{destructor} functions;
1987\CFA adopts constructors and destructors (and @finally@) to facilitate RAII.
1988While 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.
1989Specifically, \CFA constructors and destructors are denoted by name and first parameter-type versus name and nesting in an aggregate type.
1990Constructor 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.
1991
1992In \CFA, a constructor is named @?{}@ and a destructor is named @^?{}@\footnote{%
1993The symbol \lstinline+^+ is used for the destructor name because it was the last binary operator that could be used in a unary context.}.
1994The name @{}@ comes from the syntax for the initializer: @struct S { int i, j; } s = `{` 2, 3 `}`@.
1995Like other \CFA operators, these names represent the syntax used to call the constructor or destructor, \eg @?{}(x, ...)@ or @^{}(x, ...)@.
1996The constructor and destructor have return type @void@, and the first parameter is a reference to the object type to be constructed or destructed.
1997While the first parameter is informally called the @this@ parameter, as in object-oriented languages, any variable name may be used.
1998Both constructors and destructors allow additional parametes after the @this@ parameter for specifying values for initialization/de-initialization\footnote{
1999Destruction parameters are useful for specifying storage-management actions, such as de-initialize but not deallocate.}.
2000\begin{cfa}
2001struct VLA { int len, * data; };
2002void ?{}( VLA & vla ) with ( vla ) { len = 10;  data = alloc( len ); }  $\C{// default constructor}$
2003void ^?{}( VLA & vla ) with ( vla ) { free( data ); } $\C{// destructor}$
2004{
2005        VLA x;                                                                  $\C{// implicit:  ?\{\}( x );}$
2006}                                                                                       $\C{// implicit:  ?\^{}\{\}( x );}$
2007\end{cfa}
2008@VLA@ is a \newterm{managed type}\footnote{
2009A 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.
2010A 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.
2011For details of the code-generation placement of implicit constructor and destructor calls among complex executable statements see~\cite[\S~2.2]{Schluntz17}.
2012
2013\CFA also provides syntax for \newterm{initialization} and \newterm{copy}:
2014\begin{cfa}
2015void ?{}( VLA & vla, int size, char fill ) with ( vla ) {  $\C{// initialization}$
2016        len = size;  data = alloc( len, fill );
2017}
2018void ?{}( VLA & vla, VLA other ) {                      $\C{// copy, shallow}$
2019        vla.len = other.len;  vla.data = other.data;
2020}
2021\end{cfa}
2022(Note, the example is purposely simplified using shallow-copy semantics.)
2023An 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).
2024\begin{cfa}
2025VLA va = `{` 20, 0 `}`,  * arr = alloc()`{` 5, 0 `}`;
2026\end{cfa}
2027Note, the use of a \newterm{constructor expression} to initialize the storage from the dynamic storage-allocation.
2028Like \CC, the copy constructor has two parameters, the second of which is a value parameter with the same type as the first parameter;
2029appropriate care is taken to not recursively call the copy constructor when initializing the second parameter.
2030
2031\CFA constructors may be explicitly call, like Java, and destructors may be explicitly called, like \CC.
2032Explicit 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.
2033While existing call syntax works for explicit calls to constructors and destructors, \CFA also provides a more concise \newterm{operator syntax} for both:
2034\begin{cfa}
2035{
2036        VLA  x,            y = { 20, 0x01 },     z = y; $\C{// z points to y}$
2037        //      ?{}( x );   ?{}( y, 20, 0x01 );   ?{}( z, y );
2038        ^x{};                                                                   $\C{// deallocate x}$
2039        x{};                                                                    $\C{// reallocate x}$
2040        z{ 5, 0xff };                                                   $\C{// reallocate z, not pointing to y}$
2041        ^y{};                                                                   $\C{// deallocate y}$
2042        y{ x };                                                                 $\C{// reallocate y, points to x}$
2043        x{};                                                                    $\C{// reallocate x, not pointing to y}$
2044        // ^?{}(z);  ^?{}(y);  ^?{}(x);
2045}
2046\end{cfa}
2047
2048To 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.
2049These default functions can be overridden by user-generated versions.
2050For 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;
2051if default zero-initialization is desired, the default constructors can be overridden.
2052For user-generated types, the four functions are also automatically generated.
2053@enum@ types are handled the same as their underlying integral type, and unions are also bitwise copied and no-op initialized and destructed.
2054For compatibility with C, a copy constructor from the first union member type is also defined.
2055For @struct@ types, each of the four functions are implicitly defined to call their corresponding functions on each member of the struct.
2056To better simulate the behaviour of C initializers, a set of \newterm{field constructors} is also generated for structures.
2057A 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.
2058To 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;
2059similarly, the generated default constructor is hidden upon declaration of any constructor.
2060These semantics closely mirror the rule for implicit declaration of constructors in \CC\cite[p.~186]{ANSI98:C++}.
2061
2062In some circumstance programmers may not wish to have implicit constructor and destructor generation and calls.
2063In these cases, \CFA provides the initialization syntax \lstinline|S x `@=` {}|, and the object becomes unmanaged, so implicit constructor and destructor calls are not generated.
2064Any C initializer can be the right-hand side of an \lstinline|@=| initializer, \eg \lstinline|VLA a @= { 0, 0x0 }|, with the usual C initialization semantics.
2065The same syntax can be used in a compound literal, \eg \lstinline|a = VLA`@`{ 0, 0x0 }|, to create a C-style literal.
2066The 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.
2067
2068
2069% \subsection{Default Parameters}
2070
2071
2072\section{Literals}
2073
2074C 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.
2075In 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.
2076
2077A 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.
2078\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.
2079
2080
2081\begin{comment}
2082\subsection{Integral Suffixes}
2083
2084New integral suffixes @hh@ (half of half of @int@) for @char@, @h@ (half of @int@) for @short@, and @z@ for @size_t@, and length suffixes for 8, 16, 32, 64, and 128 bit integers.
2085%Additional integral suffixes are added to cover all the integral types and lengths.
2086\begin{cquote}
2087\lstDeleteShortInline@%
2088\begin{tabular}{@{}l@{\hspace{2\parindentlnth}}l@{\hspace{2\parindentlnth}}l@{}}
2089\begin{cfa}
209020_`hh`     // signed char
209121_`hh`u   // unsigned char
209222_`h`       // signed short int
209323_u`h`     // unsigned short int
209424`z`         // size_t
2095\end{cfa}
2096&
2097\begin{cfa}
209820_`L8`      // int8_t
209921_u`l8`     // uint8_t
210022_`l16`     // int16_t
210123_u`l16`   // uint16_t
210224_`l32`     // int32_t
2103\end{cfa}
2104&
2105\begin{cfa}
210625_u`l32`      // uint32_t
210726_`l64`        // int64_t
210827_`l64`u      // uint64_t
210926_`L128`     // int128
211027_`L128`u   // unsigned int128
2111\end{cfa}
2112\end{tabular}
2113\lstMakeShortInline@%
2114\end{cquote}
2115\end{comment}
2116
2117
2118\subsection{0/1}
2119
2120In C, @0@ has the special property that it is the only ``false'' value;
2121from the standard, any value that compares equal to @0@ is false, while any value that compares unequal to @0@ is true.
2122As 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.
2123Operator overloading in \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.
2124To provide this precision, \CFA introduces a new type @zero_t@ as the type of literal @0@ (somewhat analagous to @nullptr_t@ and @nullptr@ in \CCeleven);
2125@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.
2126With this addition, \CFA rewrites @if (x)@ and similar expressions to @if ( (x) != 0 )@ or the appropriate analogue, and any type @T@ is ``truthy'' by defining an operator overload @int ?!=?( T, zero_t )@.
2127\CC makes types truthy by adding a conversion to @bool@;
2128prior to the addition of explicit cast operators in \CCeleven, this approach had the pitfall of making truthy types transitively convertable to any numeric type;
2129\CFA avoids this issue.
2130
2131Similarly, \CFA also has a special type for @1@, @one_t@;
2132like @zero_t@, @one_t@ has built-in implicit conversions to the various integral types so that @1@ maintains its expected semantics in legacy code for operations @++@ and @--@.
2133The addition of @one_t@ allows generic algorithms to handle the unit value uniformly for types where it is meaningful.
2134\TODO{Make this sentence true}
2135In 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)@;
2136analogous overloads for the decrement operators are present as well.
2137
2138
2139\subsection{User Literals}
2140
2141For readability, it is useful to associate units to scale literals, \eg weight (stone, pound, kilogram) or time (seconds, minutes, hours).
2142The left of Figure~\ref{f:UserLiteral} shows the \CFA alternative call-syntax (postfix: literal argument before function name), using the backquote, to convert basic literals into user literals.
2143The backquote is a small character, making the unit (function name) predominate.
2144For examples, the multi-precision integer-type in Section~\ref{s:MultiPrecisionIntegers} has user literals:
2145{\lstset{language=CFA,moredelim=**[is][\color{red}]{|}{|},deletedelim=**[is][]{`}{`}}
2146\begin{cfa}
2147y = 9223372036854775807L|`mp| * 18446744073709551615UL|`mp|;
2148y = "12345678901234567890123456789"|`mp| + "12345678901234567890123456789"|`mp|;
2149\end{cfa}
2150Because \CFA uses a standard function, all types and literals are applicable, as well as overloading and conversions, where @?`@ denotes a postfix-function name and @`@ denotes a postfix-function call.
2151}%
2152\begin{cquote}
2153\lstset{language=CFA,moredelim=**[is][\color{red}]{|}{|},deletedelim=**[is][]{`}{`}}
2154\lstDeleteShortInline@%
2155\begin{tabular}{@{}l@{\hspace{2\parindentlnth}}l@{\hspace{2\parindentlnth}}l@{\hspace{2\parindentlnth}}l@{}}
2156\multicolumn{1}{c@{\hspace{2\parindentlnth}}}{\textbf{postfix function}}        & \multicolumn{1}{c@{\hspace{2\parindentlnth}}}{\textbf{constant}}      & \multicolumn{1}{c@{\hspace{2\parindentlnth}}}{\textbf{variable/expression}}   & \multicolumn{1}{c}{\textbf{postfix pointer}}  \\
2157\begin{cfa}
2158int |?`h|( int s );
2159int |?`h|( double s );
2160int |?`m|( char c );
2161int |?`m|( const char * s );
2162int |?`t|( int a, int b, int c );
2163\end{cfa}
2164&
2165\begin{cfa}
21660 |`h|;
21673.5|`h|;
2168'1'|`m|;
2169"123" "456"|`m|;
2170[1,2,3]|`t|;
2171\end{cfa}
2172&
2173\begin{cfa}
2174int i = 7;
2175i|`h|;
2176(i + 3)|`h|;
2177(i + 3.5)|`h|;
2178
2179\end{cfa}
2180&
2181\begin{cfa}
2182int (* |?`p|)( int i );
2183|?`p| = |?`h|;
21843|`p|;
2185i|`p|;
2186(i + 3)|`p|;
2187\end{cfa}
2188\end{tabular}
2189\lstMakeShortInline@%
2190\end{cquote}
2191
2192The right of Figure~\ref{f:UserLiteral} shows the equivalent \CC version using the underscore for the call-syntax.
2193However, \CC restricts the types, \eg @unsigned long long int@ and @long double@ to represent integral and floating literals.
2194After which, user literals must match (no conversions);
2195hence, it is necessary to overload the unit with all appropriate types.
2196
2197\begin{figure}
2198\centering
2199\lstset{language=CFA,moredelim=**[is][\color{red}]{|}{|},deletedelim=**[is][]{`}{`}}
2200\lstDeleteShortInline@%
2201\begin{tabular}{@{}l@{\hspace{2\parindentlnth}}l@{}}
2202\multicolumn{1}{c@{\hspace{2\parindentlnth}}}{\textbf{\CFA}}    & \multicolumn{1}{c}{\textbf{\CC}}      \\
2203\begin{cfa}
2204struct W {
2205        double stones;
2206};
2207void ?{}( W & w ) { w.stones = 0; }
2208void ?{}( W & w, double w ) { w.stones = w; }
2209W ?+?( W l, W r ) {
2210        return (W){ l.stones + r.stones };
2211}
2212W |?`st|(double w) { return (W){ w }; }
2213W |?`lb|(double w) { return (W){ w/14.0 }; }
2214W |?`kg|(double w) { return (W){ w*0.16 }; }
2215
2216
2217
2218int main() {
2219        W w, heavy = { 20 };
2220        w = 155|`lb|;
2221        w = 0b_1111|`st|;
2222        w = 0_233|`lb|;
2223        w = 0x_9b_u|`kg|;
2224        w = 5.5|`st| + 8|`kg| + 25.01|`lb| + heavy;
2225}
2226\end{cfa}
2227&
2228\begin{cfa}
2229struct W {
2230        double stones;
2231        W() { stones = 0.0; }
2232        W( double w ) { stones = w; }
2233};
2234W operator+( W l, W r ) {
2235        return W( l.stones + r.stones );
2236}
2237W |operator""_st|(unsigned long long int w) {return W(w); }
2238W |operator""_lb|(unsigned long long int w) {return W(w/14.0); }
2239W |operator""_kg|(unsigned long long int w) {return W(w*0.16); }
2240W |operator""_st|(long double w ) { return W( w ); }
2241W |operator""_lb|(long double w ) { return W( w / 14.0 ); }
2242W |operator""_kg|(long double w ) { return W( w * 0.16 ); }
2243int main() {
2244        W w, heavy = { 20 };
2245        w = 155|_lb|;
2246        // binary unsupported
2247        w = 0${\color{red}\LstBasicStyle{'}}$233|_lb|;          // quote separator
2248        w = 0x9b|_kg|;
2249        w = 5.5d|_st| + 8|_kg| + 25.01|_lb| + heavy;
2250}
2251\end{cfa}
2252\end{tabular}
2253\lstMakeShortInline@%
2254\caption{User Literal}
2255\label{f:UserLiteral}
2256\end{figure}
2257
2258
2259\section{Libraries}
2260\label{sec:libraries}
2261
2262As 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.
2263\CFA has replacement libraries condensing hundreds of existing C names into tens of \CFA overloaded names, all without rewriting the actual computations.
2264In many cases, the interface is an inline wrapper providing overloading during compilation but zero cost at runtime.
2265The following sections give a glimpse of the interface reduction to many C libraries.
2266In 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.
2267
2268
2269\subsection{Limits}
2270
2271C library @limits.h@ provides lower and upper bound constants for the basic types.
2272\CFA name overloading is used to condense these typed constants, \eg:
2273\begin{cquote}
2274\lstDeleteShortInline@%
2275\begin{tabular}{@{}l@{\hspace{2\parindentlnth}}l@{}}
2276\multicolumn{1}{c@{\hspace{2\parindentlnth}}}{\textbf{Definition}}      & \multicolumn{1}{c}{\textbf{Usage}}    \\
2277\begin{cfa}
2278const short int `MIN` = -32768;
2279const int `MIN` = -2147483648;
2280const long int `MIN` = -9223372036854775808L;
2281\end{cfa}
2282&
2283\begin{cfa}
2284short int si = `MIN`;
2285int i = `MIN`;
2286long int li = `MIN`;
2287\end{cfa}
2288\end{tabular}
2289\lstMakeShortInline@%
2290\end{cquote}
2291The result is a significant reduction in names to access typed constants, \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}
2297MIN
2298MAX
2299PI
2300E
2301\end{cfa}
2302&
2303\begin{cfa}
2304CHAR_MIN, SHRT_MIN, INT_MIN, LONG_MIN, LLONG_MIN, FLT_MIN, DBL_MIN, LDBL_MIN
2305UCHAR_MAX, SHRT_MAX, INT_MAX, LONG_MAX, LLONG_MAX, FLT_MAX, DBL_MAX, LDBL_MAX
2306M_PI, M_PIl
2307M_E, M_El
2308\end{cfa}
2309\end{tabular}
2310\lstMakeShortInline@%
2311\end{cquote}
2312
2313
2314\subsection{Math}
2315
2316C library @math.h@ provides many mathematical functions.
2317\CFA function overloading is used to condense these mathematical functions, \eg:
2318\begin{cquote}
2319\lstDeleteShortInline@%
2320\begin{tabular}{@{}l@{\hspace{2\parindentlnth}}l@{}}
2321\multicolumn{1}{c@{\hspace{2\parindentlnth}}}{\textbf{Definition}}      & \multicolumn{1}{c}{\textbf{Usage}}    \\
2322\begin{cfa}
2323float `log`( float x );
2324double `log`( double );
2325double _Complex `log`( double _Complex x );
2326\end{cfa}
2327&
2328\begin{cfa}
2329float f = `log`( 3.5 );
2330double d = `log`( 3.5 );
2331double _Complex dc = `log`( 3.5+0.5I );
2332\end{cfa}
2333\end{tabular}
2334\lstMakeShortInline@%
2335\end{cquote}
2336The result is a significant reduction in names to access math functions, \eg:
2337\begin{cquote}
2338\lstDeleteShortInline@%
2339\begin{tabular}{@{}l@{\hspace{2\parindentlnth}}l@{}}
2340\multicolumn{1}{c@{\hspace{2\parindentlnth}}}{\textbf{\CFA}}    & \multicolumn{1}{c}{\textbf{C}}        \\
2341\begin{cfa}
2342log
2343sqrt
2344sin
2345\end{cfa}
2346&
2347\begin{cfa}
2348logf, log, logl, clogf, clog, clogl
2349sqrtf, sqrt, sqrtl, csqrtf, csqrt, csqrtl
2350sinf, sin, sinl, csinf, csin, csinl
2351\end{cfa}
2352\end{tabular}
2353\lstMakeShortInline@%
2354\end{cquote}
2355While \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).
2356For example, it is impossible to overload @atan@ for both one and two arguments;
2357instead the names @atan@ and @atan2@ are required (see Section~\ref{s:NameOverloading}).
2358The 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.
2359
2360
2361\subsection{Standard}
2362
2363C library @stdlib.h@ provides many general functions.
2364\CFA function overloading is used to condense these utility functions, \eg:
2365\begin{cquote}
2366\lstDeleteShortInline@%
2367\begin{tabular}{@{}l@{\hspace{2\parindentlnth}}l@{}}
2368\multicolumn{1}{c@{\hspace{2\parindentlnth}}}{\textbf{Definition}}      & \multicolumn{1}{c}{\textbf{Usage}}    \\
2369\begin{cfa}
2370unsigned int `abs`( int );
2371double `abs`( double );
2372double abs( double _Complex );
2373\end{cfa}
2374&
2375\begin{cfa}
2376unsigned int i = `abs`( -1 );
2377double d = `abs`( -1.5 );
2378double d = `abs`( -1.5+0.5I );
2379\end{cfa}
2380\end{tabular}
2381\lstMakeShortInline@%
2382\end{cquote}
2383The result is a significant reduction in names to access utility functions, \eg:
2384\begin{cquote}
2385\lstDeleteShortInline@%
2386\begin{tabular}{@{}l@{\hspace{2\parindentlnth}}l@{}}
2387\multicolumn{1}{c@{\hspace{2\parindentlnth}}}{\textbf{\CFA}}    & \multicolumn{1}{c}{\textbf{C}}        \\
2388\begin{cfa}
2389abs
2390strto
2391random
2392\end{cfa}
2393&
2394\begin{cfa}
2395abs, labs, llabs, fabsf, fabs, fabsl, cabsf, cabs, cabsl
2396strtol, strtoul, strtoll, strtoull, strtof, strtod, strtold
2397srand48, mrand48, lrand48, drand48
2398\end{cfa}
2399\end{tabular}
2400\lstMakeShortInline@%
2401\end{cquote}
2402In additon, there are polymorphic functions, like @min@ and @max@, that work on any type with operators @?<?@ or @?>?@.
2403
2404The following shows one example where \CFA \emph{extends} an existing standard C interface to reduce complexity and provide safety.
2405C/\Celeven provide a number of complex and overlapping storage-management operation to support the following capabilities:
2406\begin{description}%[topsep=3pt,itemsep=2pt,parsep=0pt]
2407\item[fill]
2408an allocation with a specified character.
2409\item[resize]
2410an existing allocation to decreased or increased its size.
2411In 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.
2412For an increase in storage size, new storage after the copied data may be filled.
2413\item[align]
2414an allocation on a specified memory boundary, \eg, an address multiple of 64 or 128 for cache-line purposes.
2415\item[array]
2416allocation with a specified number of elements.
2417An array may be filled, resized, or aligned.
2418\end{description}
2419Table~\ref{t:StorageManagementOperations} shows the capabilities provided by C/\Celeven allocation-functions and how all the capabilities can be combined into two \CFA functions.
2420\CFA storage-management functions extend the C equivalents by overloading, providing shallow type-safety, and removing the need to specify the base allocation-size.
2421Figure~\ref{f:StorageAllocation} contrasts \CFA and C storage-allocation performing the same operations with the same type safety.
2422
2423\begin{table}
2424\caption{Storage-Management Operations}
2425\label{t:StorageManagementOperations}
2426\centering
2427\lstDeleteShortInline@%
2428\lstMakeShortInline~%
2429\begin{tabular}{@{}r|r|l|l|l|l@{}}
2430\multicolumn{1}{c}{}&           & \multicolumn{1}{c|}{fill}     & resize        & align & array \\
2431\hline
2432C               & ~malloc~                      & no                    & no            & no            & no    \\
2433                & ~calloc~                      & yes (0 only)  & no            & no            & yes   \\
2434                & ~realloc~                     & no/copy               & yes           & no            & no    \\
2435                & ~memalign~            & no                    & no            & yes           & no    \\
2436                & ~posix_memalign~      & no                    & no            & yes           & no    \\
2437\hline
2438C11             & ~aligned_alloc~       & no                    & no            & yes           & no    \\
2439\hline
2440\CFA    & ~alloc~                       & yes/copy              & no/yes        & no            & yes   \\
2441                & ~align_alloc~         & yes                   & no            & yes           & yes   \\
2442\end{tabular}
2443\lstDeleteShortInline~%
2444\lstMakeShortInline@%
2445\end{table}
2446
2447\begin{figure}
2448\centering
2449\begin{cquote}
2450\begin{cfa}[aboveskip=0pt]
2451size_t  dim = 10;                                                       $\C{// array dimension}$
2452char fill = '\xff';                                                     $\C{// initialization fill value}$
2453int * ip;
2454\end{cfa}
2455\lstDeleteShortInline@%
2456\begin{tabular}{@{}l@{\hspace{2\parindentlnth}}l@{}}
2457\multicolumn{1}{c@{\hspace{2\parindentlnth}}}{\textbf{\CFA}}    & \multicolumn{1}{c}{\textbf{C}}        \\
2458\begin{cfa}
2459ip = alloc();
2460ip = alloc( fill );
2461ip = alloc( dim );
2462ip = alloc( dim, fill );
2463ip = alloc( ip, 2 * dim );
2464ip = alloc( ip, 4 * dim, fill );
2465
2466ip = align_alloc( 16 );
2467ip = align_alloc( 16, fill );
2468ip = align_alloc( 16, dim );
2469ip = align_alloc( 16, dim, fill );
2470\end{cfa}
2471&
2472\begin{cfa}
2473ip = (int *)malloc( sizeof( int ) );
2474ip = (int *)malloc( sizeof( int ) ); memset( ip, fill, sizeof( int ) );
2475ip = (int *)malloc( dim * sizeof( int ) );
2476ip = (int *)malloc( sizeof( int ) ); memset( ip, fill, dim * sizeof( int ) );
2477ip = (int *)realloc( ip, 2 * dim * sizeof( int ) );
2478ip = (int *)realloc( ip, 4 * dim * sizeof( int ) );
2479                        memset( ip, fill, 4 * dim * sizeof( int ) );
2480ip = memalign( 16, sizeof( int ) );
2481ip = memalign( 16, sizeof( int ) ); memset( ip, fill, sizeof( int ) );
2482ip = memalign( 16, dim * sizeof( int ) );
2483ip = memalign( 16, dim * sizeof( int ) ); memset( ip, fill, dim * sizeof( int ) );
2484\end{cfa}
2485\end{tabular}
2486\lstMakeShortInline@%
2487\end{cquote}
2488\caption{\CFA versus C Storage-Allocation}
2489\label{f:StorageAllocation}
2490\end{figure}
2491
2492Variadic @new@ (see Section~\ref{sec:variadic-tuples}) cannot support the same overloading because extra parameters are for initialization.
2493Hence, there are @new@ and @anew@ functions for single and array variables, and the fill value is the arguments to the constructor, \eg:
2494\begin{cfa}
2495struct S { int i, j; };
2496void ?{}( S & s, int i, int j ) { s.i = i; s.j = j; }
2497S * s = new( 2, 3 );                                            $\C{// allocate storage and run constructor}$
2498S * as = anew( dim, 2, 3 );                                     $\C{// each array element initialized to 2, 3}$
2499\end{cfa}
2500Note, \CC can only initialization array elements via the default constructor.
2501
2502Finally, the \CFA memory-allocator has \newterm{sticky properties} for dynamic storage: fill and alignment are remembered with an object's storage in the heap.
2503When a @realloc@ is performed, the sticky properties are respected, so that new storage is correctly aligned and initialized with the fill character.
2504
2505
2506\subsection{I/O}
2507\label{s:IOLibrary}
2508
2509The goal of \CFA I/O is to simplify the common cases, while fully supporting polymorphism and user defined types in a consistent way.
2510The approach combines ideas from \CC and Python.
2511The \CFA header file for the I/O library is @fstream@.
2512
2513The common case is printing out a sequence of variables separated by whitespace.
2514\begin{cquote}
2515\lstDeleteShortInline@%
2516\begin{tabular}{@{}l@{\hspace{2\parindentlnth}}l@{}}
2517\multicolumn{1}{c@{\hspace{2\parindentlnth}}}{\textbf{\CFA}}    & \multicolumn{1}{c}{\textbf{\CC}}      \\
2518\begin{cfa}
2519int x = 1, y = 2, z = 3;
2520sout | x `|` y `|` z | endl;
2521\end{cfa}
2522&
2523\begin{cfa}
2524
2525cout << x `<< " "` << y `<< " "` << z << endl;
2526\end{cfa}
2527\\
2528\begin{cfa}[showspaces=true,aboveskip=0pt,belowskip=0pt]
25291` `2` `3
2530\end{cfa}
2531&
2532\begin{cfa}[showspaces=true,aboveskip=0pt,belowskip=0pt]
25331 2 3
2534\end{cfa}
2535\end{tabular}
2536\lstMakeShortInline@%
2537\end{cquote}
2538The \CFA form has half the characters of the \CC form, and is similar to Python I/O with respect to implicit separators.
2539Similar simplification occurs for tuple I/O, which prints all tuple values separated by ``\lstinline[showspaces=true]@, @''.
2540\begin{cfa}
2541[int, [ int, int ] ] t1 = [ 1, [ 2, 3 ] ], t2 = [ 4, [ 5, 6 ] ];
2542sout | t1 | t2 | endl;                                  $\C{// print tuples}$
2543\end{cfa}
2544\begin{cfa}[showspaces=true,aboveskip=0pt]
25451`, `2`, `3 4`, `5`, `6
2546\end{cfa}
2547Finally, \CFA uses the logical-or operator for I/O as it is the lowest-priority overloadable operator, other than assignment.
2548Therefore, fewer output expressions require parenthesis.
2549\begin{cquote}
2550\lstDeleteShortInline@%
2551\begin{tabular}{@{}ll@{}}
2552\textbf{\CFA:}
2553&
2554\begin{cfa}
2555sout | x * 3 | y + 1 | z << 2 | x == y | (x | y) | (x || y) | (x > z ? 1 : 2) | endl;
2556\end{cfa}
2557\\
2558\textbf{\CC:}
2559&
2560\begin{cfa}
2561cout << x * 3 << y + 1 << `(`z << 2`)` << `(`x == y`)` << (x | y) << (x || y) << (x > z ? 1 : 2) << endl;
2562\end{cfa}
2563\\
2564&
2565\begin{cfa}[showspaces=true,aboveskip=0pt]
25663 3 12 0 3 1 2
2567\end{cfa}
2568\end{tabular}
2569\lstMakeShortInline@%
2570\end{cquote}
2571There 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.
2572\begin{comment}
2573The implicit separator character (space/blank) is a separator not a terminator.
2574The rules for implicitly adding the separator are:
2575\begin{itemize}
2576\item
2577A separator does not appear at the start or end of a line.
2578\item
2579A separator does not appear before or after a character literal or variable.
2580\item
2581A separator does not appear before or after a null (empty) C string, which is a local mechanism to disable insertion of the separator character.
2582\item
2583A separator does not appear before a C string starting with the characters: \lstinline[mathescape=off,basicstyle=\tt]@([{=$@
2584\item
2585A separator does not appear after a C string ending with the characters: \lstinline[basicstyle=\tt]@,.;!?)]}%@
2586\item
2587{\lstset{language=CFA,deletedelim=**[is][]{`}{`}}
2588A 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@
2589}%
2590\end{itemize}
2591\end{comment}
2592There are functions to set and get the separator string, and manipulators to toggle separation on and off in the middle of output.
2593
2594
2595\subsection{Multi-precision Integers}
2596\label{s:MultiPrecisionIntegers}
2597
2598\CFA has an interface to the GMP multi-precision signed-integers~\cite{GMP}, similar to the \CC interface provided by GMP.
2599The \CFA interface wraps GMP functions into operator functions to make programming with multi-precision integers identical to using fixed-sized integers.
2600The \CFA type name for multi-precision signed-integers is @Int@ and the header file is @gmp@.
2601Figure~\ref{f:GMPInterface} shows a multi-precision factorial-program contrasting the GMP interface in \CFA and C.
2602
2603\begin{figure}
2604\centering
2605\lstDeleteShortInline@%
2606\begin{tabular}{@{}l@{\hspace{2\parindentlnth}}@{\hspace{2\parindentlnth}}l@{}}
2607\multicolumn{1}{c@{\hspace{2\parindentlnth}}}{\textbf{\CFA}}    & \multicolumn{1}{@{\hspace{2\parindentlnth}}c}{\textbf{C}}     \\
2608\begin{cfa}
2609#include <gmp>
2610int main( void ) {
2611        sout | "Factorial Numbers" | endl;
2612        Int fact = 1;
2613        sout | 0 | fact | endl;
2614        for ( unsigned int i = 1; i <= 40; i += 1 ) {
2615                fact *= i;
2616                sout | i | fact | endl;
2617        }
2618}
2619\end{cfa}
2620&
2621\begin{cfa}
2622#include <gmp.h>
2623int main( void ) {
2624        `gmp_printf`( "Factorial Numbers\n" );
2625        `mpz_t` fact;  `mpz_init_set_ui`( fact, 1 );
2626        `gmp_printf`( "%d %Zd\n", 0, fact );
2627        for ( unsigned int i = 1; i <= 40; i += 1 ) {
2628                `mpz_mul_ui`( fact, fact, i );
2629                `gmp_printf`( "%d %Zd\n", i, fact );
2630        }
2631}
2632\end{cfa}
2633\end{tabular}
2634\lstMakeShortInline@%
2635\caption{GMP Interface \CFA versus C}
2636\label{f:GMPInterface}
2637\end{figure}
2638
2639
2640\section{Polymorphic Evaluation}
2641\label{sec:eval}
2642
2643\CFA adds parametric polymorphism to C.
2644A runtime evaluation is performed to compare the cost of alternative styles of polymorphism.
2645The goal is to compare just the underlying mechanism for implementing different kinds of polymorphism.
2646% Though \CFA provides significant added functionality over C, these features have a low runtime penalty.
2647% In fact, it is shown that \CFA's generic programming can enable faster runtime execution than idiomatic @void *@-based C code.
2648The experiment is a set of generic-stack micro-benchmarks~\cite{CFAStackEvaluation} in C, \CFA, and \CC (see implementations in Appendix~\ref{sec:BenchmarkStackImplementations}).
2649Since all these languages share a subset essentially comprising standard C, maximal-performance benchmarks should show little runtime variance, differing only in length and clarity of source code.
2650A more illustrative comparison measures the costs of idiomatic usage of each language's features.
2651Figure~\ref{fig:BenchmarkTest} shows the \CFA benchmark tests for a generic stack based on a singly linked-list.
2652The benchmark test is similar for the other languages.
2653The experiment uses element types @int@ and @pair(short, char)@, and pushes $N=40M$ elements on a generic stack, copies the stack, clears one of the stacks, and finds the maximum value in the other stack.
2654
2655\begin{figure}
2656\begin{cfa}[xleftmargin=3\parindentlnth,aboveskip=0pt,belowskip=0pt]
2657int main() {
2658        int max = 0, val = 42;
2659        stack( int ) si, ti;
2660
2661        REPEAT_TIMED( "push_int", N, push( si, val ); )
2662        TIMED( "copy_int", ti{ si }; )
2663        TIMED( "clear_int", clear( si ); )
2664        REPEAT_TIMED( "pop_int", N, int x = pop( ti ); if ( x > max ) max = x; )
2665
2666        pair( short, char ) max = { 0h, '\0' }, val = { 42h, 'a' };
2667        stack( pair( short, char ) ) sp, tp;
2668
2669        REPEAT_TIMED( "push_pair", N, push( sp, val ); )
2670        TIMED( "copy_pair", tp{ sp }; )
2671        TIMED( "clear_pair", clear( sp ); )
2672        REPEAT_TIMED( "pop_pair", N, pair(short, char) x = pop( tp ); if ( x > max ) max = x; )
2673}
2674\end{cfa}
2675\caption{\protect\CFA Benchmark Test}
2676\label{fig:BenchmarkTest}
2677\end{figure}
2678
2679The structure of each benchmark implemented is: C with @void *@-based polymorphism, \CFA with parametric polymorphism, \CC with templates, and \CC using only class inheritance for polymorphism, called \CCV.
2680The \CCV variant illustrates an alternative object-oriented idiom where all objects inherit from a base @object@ class, mimicking a Java-like interface;
2681hence runtime checks are necessary to safely down-cast objects.
2682The 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.
2683Note, the C benchmark uses unchecked casts as C has no runtime mechanism to perform such checks, while \CFA and \CC provide type-safety statically.
2684
2685Figure~\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.
2686The graph plots the median of 5 consecutive runs of each program, with an initial warm-up run omitted.
2687All code is compiled at \texttt{-O2} by gcc or g++ 6.4.0, with all \CC code compiled as \CCfourteen.
2688The 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.
2689
2690\begin{figure}
2691\centering
2692\input{timing}
2693\caption{Benchmark Timing Results (smaller is better)}
2694\label{fig:eval}
2695\end{figure}
2696
2697\begin{table}
2698\caption{Properties of benchmark code}
2699\label{tab:eval}
2700\centering
2701\newcommand{\CT}[1]{\multicolumn{1}{c}{#1}}
2702\begin{tabular}{rrrrr}
2703                                                                        & \CT{C}        & \CT{\CFA}     & \CT{\CC}      & \CT{\CCV}             \\ \hline
2704maximum memory usage (MB)                       & 10,001        & 2,502         & 2,503         & 11,253                \\
2705source code size (lines)                        & 201           & 191           & 125           & 294                   \\
2706redundant type annotations (lines)      & 27            & 0                     & 2                     & 16                    \\
2707binary size (KB)                                        & 14            & 257           & 14            & 37                    \\
2708\end{tabular}
2709\end{table}
2710
2711The 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;
2712this inefficiency is exacerbated by the second level of generic types in the pair benchmarks.
2713By contrast, the \CFA and \CC variants run in roughly equivalent time for both the integer and pair because of equivalent storage layout, with the inlined libraries (\ie no separate compilation) and greater maturity of the \CC compiler contributing to its lead.
2714\CCV is slower than C largely due to the cost of runtime type-checking of down-casts (implemented with @dynamic_cast@);
2715The outlier for \CFA, pop @pair@, results from the complexity of the generated-C polymorphic code.
2716The gcc compiler is unable to optimize some dead code and condense nested calls;
2717a compiler designed for \CFA could easily perform these optimizations.
2718Finally, the binary size for \CFA is larger because of static linking with the \CFA libraries.
2719
2720\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 39 and 42 lines, respectively.
2721The difference between the \CFA and \CC line counts is primarily declaration duplication to implement separate compilation; a header-only \CFA library would be similar in length to the \CC version.
2722On 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.
2723\CCV does not use the \CC standard template library by construction, and in fact includes the definition of @object@ and wrapper classes for @char@, @short@, and @int@ in its line count, which inflates this count somewhat, as an actual object-oriented language would include these in the standard library;
2724with their omission, the \CCV line count is similar to C.
2725We 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.
2726
2727Line-count is a fairly rough measure of code complexity;
2728another important factor is how much type information the programmer must specify manually, especially where that information is not compiler-checked.
2729Such 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 pointer arguments and format codes, or \CCV, with its extensive use of untype-checked downcasts, \eg @object@ to @integer@ when popping a stack.
2730To quantify this manual typing, the ``redundant type annotations'' line in Table~\ref{tab:eval} counts the number of lines on which the type of a known variable is respecified, either as a format specifier, explicit downcast, type-specific function, or by name in a @sizeof@, struct literal, or @new@ expression.
2731The \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.
2732The \CFA benchmark is able to eliminate all redundant type annotations through use of the polymorphic @alloc@ function discussed in Section~\ref{sec:libraries}.
2733
2734We conjecture these results scale across most generic data-types as the underlying polymorphic implement is constant.
2735
2736
2737\section{Related Work}
2738
2739
2740\subsection{Polymorphism}
2741
2742ML~\cite{ML} was the first language to support parametric polymorphism.
2743Like \CFA, it supports universal type parameters, but not the use of assertions and traits to constrain type arguments.
2744Haskell~\cite{Haskell10} combines ML-style polymorphism, polymorphic data types, and type inference with the notion of type classes, collections of overloadable methods that correspond in intent to traits in \CFA.
2745Unlike \CFA, Haskell requires an explicit association between types and their classes that specifies the implementation of operations.
2746These associations determine the functions that are assertion arguments for particular combinations of class and type, in contrast to \CFA where the assertion arguments are selected at function call sites based upon the set of operations in scope at that point.
2747Haskell also severely restricts the use of overloading: an overloaded name can only be associated with a single class, and methods with overloaded names can only be defined as part of instance declarations.
2748
2749\CC provides three disjoint polymorphic extensions to C: overloading, inheritance, and templates.
2750The 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.
2751In 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.
2752The key mechanism to support separate compilation is \CFA's \emph{explicit} use of assumed type properties.
2753Until \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;
2754furthermore, \CC concepts are restricted to template polymorphism.
2755
2756Cyclone~\cite{Grossman06} also provides capabilities for polymorphic functions and existential types, similar to \CFA's @forall@ functions and generic types.
2757Cyclone 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.
2758Furthermore, 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@.
2759In \CFA terms, all Cyclone polymorphism must be dtype-static.
2760While 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.
2761Smith 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.
2762
2763Objective-C~\cite{obj-c-book} is an industrially successful extension to C.
2764However, Objective-C is a radical departure from C, using an object-oriented model with message-passing.
2765Objective-C did not support type-checked generics until recently \cite{xcode7}, historically using less-efficient runtime checking of object types.
2766The GObject~\cite{GObject} framework also adds object-oriented programming with runtime type-checking and reference-counting garbage-collection to C;
2767these features are more intrusive additions than those provided by \CFA, in addition to the runtime overhead of reference-counting.
2768Vala~\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.
2769Java~\cite{Java8} included generic types in Java~5, which are type-checked at compilation and type-erased at runtime, similar to \CFA's.
2770However, in Java, each object carries its own table of method pointers, while \CFA passes the method pointers separately to maintain a C-compatible layout.
2771Java is also a garbage-collected, object-oriented language, with the associated resource usage and C-interoperability burdens.
2772
2773D~\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.
2774However, 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.
2775D and Go are garbage-collected languages, imposing the associated runtime overhead.
2776The necessity of accounting for data transfer between managed runtimes and the unmanaged C runtime complicates foreign-function interfaces to C.
2777Furthermore, 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.
2778D 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.
2779Rust also possesses much more powerful abstraction capabilities for writing generic code than Go.
2780On 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.
2781\CFA, with its more modest safety features, allows direct ports of C code while maintaining the idiomatic style of the original source.
2782
2783
2784\subsection{Tuples/Variadics}
2785
2786Many programming languages have some form of tuple construct and/or variadic functions, \eg SETL, C, KW-C, \CC, D, Go, Java, ML, and Scala.
2787SETL~\cite{SETL} is a high-level mathematical programming language, with tuples being one of the primary data types.
2788Tuples in SETL allow subscripting, dynamic expansion, and multiple assignment.
2789C 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.
2790KW-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.
2791The main contributions of that work were adding MRVF, tuple mass and multiple assignment, and record-field access.
2792\CCeleven introduced @std::tuple@ as a library variadic template structure.
2793Tuples are a generalization of @std::pair@, in that they allow for arbitrary length, fixed-size aggregation of heterogeneous values.
2794Operations include @std::get<N>@ to extract values, @std::tie@ to create a tuple of references used for assignment, and lexicographic comparisons.
2795\CCseventeen proposes \emph{structured bindings}~\cite{Sutter15} to eliminate pre-declaring variables and use of @std::tie@ for binding the results.
2796This 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.
2797Furthermore, structured bindings are not a full replacement for @std::tie@, as it always declares new variables.
2798Like \CC, D provides tuples through a library variadic-template structure.
2799Go does not have tuples but supports MRVF.
2800Java'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.
2801Tuples are a fundamental abstraction in most functional programming languages, such as Standard ML~\cite{sml}, Haskell, and Scala~\cite{Scala}, which decompose tuples using pattern matching.
2802
2803
2804\subsection{C Extensions}
2805
2806\CC is the best known C-based language, and is similar to \CFA in that both are extensions to C with source and runtime backwards compatibility.
2807Specific difference between \CFA and \CC have been identified in prior sections, with a final observation that \CFA has equal or fewer tokens to express the same notion in many cases.
2808The key difference in design philosophies is that \CFA is easier for C programmers to understand by maintaining a procedural paradigm and avoiding complex interactions among extensions.
2809\CC, on the other hand, has multiple overlapping features (such as the three forms of polymorphism), many of which have complex interactions with its object-oriented design.
2810As a result, \CC has a steep learning curve for even experienced C programmers, especially when attempting to maintain performance equivalent to C legacy-code.
2811
2812There are several other C extension-languages with less usage and even more dramatic changes than \CC.
2813Objective-C and Cyclone are two other extensions to C with different design goals than \CFA, as discussed above.
2814Other languages extend C with more focused features.
2815$\mu$\CC~\cite{uC++book}, CUDA~\cite{Nickolls08}, ispc~\cite{Pharr12}, and Sierra~\cite{Leissa14} add concurrent or data-parallel primitives to C or \CC;
2816data-parallel features have not yet been added to \CFA, but are easily incorporated within its design, while concurrency primitives similar to those in $\mu$\CC have already been added~\cite{Delisle18}.
2817Finally, CCured~\cite{Necula02} and Ironclad \CC~\cite{DeLozier13} attempt to provide a more memory-safe C by annotating pointer types with garbage collection information; type-checked polymorphism in \CFA covers several of C's memory-safety issues, but more aggressive approaches such as annotating all pointer types with their nullability or requiring runtime garbage collection are contradictory to \CFA's backwards compatibility goals.
2818
2819
2820\section{Conclusion and Future Work}
2821
2822The 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.
2823While 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.
2824The 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.
2825The 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.
2826The work is a challenging design, engineering, and implementation exercise.
2827On the surface, the project may appear as a rehash of similar mechanisms in \CC.
2828However, 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.
2829All of these new features are being used by the \CFA development-team to build the \CFA runtime-system.
2830Finally, we demonstrate that \CFA performance for some idiomatic cases is better than C and close to \CC, showing the design is practically applicable.
2831
2832While all examples in the paper compile and run, a public beta-release of \CFA will take 6--8 months to reduce compilation time, provide better debugging, and add a few more libraries.
2833There is also new work on a number of \CFA features, including arrays with size, runtime type-information, virtual functions, user-defined conversions, and modules.
2834While \CFA polymorphic functions use dynamic virtual-dispatch with low runtime overhead (see Section~\ref{sec:eval}), it is not as low as \CC template-inlining.
2835Hence it may be beneficial to provide a mechanism for performance-sensitive code.
2836Two 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).
2837These approaches are not mutually exclusive and allow performance optimizations to be applied only when necessary, without suffering global code-bloat.
2838In general, we believe separate compilation, producing smaller code, works well with loaded hardware-caches, which may offset the benefit of larger inlined-code.
2839
2840
2841\section{Acknowledgments}
2842
2843The authors would like to recognize the design assistance of Glen Ditchfield, Richard Bilson, Thierry Delisle, Andrew Beach and Brice Dobry on the features described in this paper, and thank Magnus Madsen for feedback on the writing.
2844Funding for this project has been provided by 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.
2845
2846{%
2847\fontsize{9bp}{12bp}\selectfont%
2848\bibliography{pl}
2849}%
2850
2851\appendix
2852
2853\section{Benchmark Stack Implementations}
2854\label{sec:BenchmarkStackImplementations}
2855
2856Throughout, @/***/@ designates a counted redundant type annotation; code reformatted slightly for brevity.
2857
2858
2859\subsection{C}
2860
2861\begin{flushleft}
2862\lstDeleteShortInline@%
2863\begin{tabular}{@{}l@{\hspace{1.8\parindentlnth}}|@{\hspace{\parindentlnth}}l@{}}
2864\begin{cfa}[xleftmargin=0pt,aboveskip=0pt,belowskip=0pt]
2865typedef struct node {
2866        void * value;
2867        struct node * next;
2868} node;
2869typedef struct stack {
2870        struct node * head;
2871} stack;
2872void copy_stack( stack * s, const stack * t,
2873                                void * (*copy)( const void * ) ) {
2874        node ** cr = &s->head;
2875        for (node * nx = t->head; nx; nx = nx->next) {
2876                *cr = malloc( sizeof(node) ); /***/
2877                (*cr)->value = copy( nx->value );
2878                cr = &(*cr)->next;
2879        }
2880        *cr = NULL;
2881}
2882void clear_stack( stack * s, void (* free_el)( void * ) ) {
2883        for ( node * nx = s->head; nx; ) {
2884                node * cr = nx;
2885                nx = cr->next;
2886                free_el( cr->value );
2887                free( cr );
2888        }
2889        s->head = NULL;
2890}
2891\end{cfa}
2892&
2893\begin{cfa}[xleftmargin=0pt,aboveskip=0pt,belowskip=0pt]
2894stack new_stack() {
2895        return (stack){ NULL }; /***/
2896}
2897stack * assign_stack( stack * s, const stack * t,
2898                                void * (*copy_el)( const void * ),
2899                                void (*free_el)( void * ) ) {
2900        if ( s->head == t->head ) return s;
2901        clear_stack( s, free_el ); /***/
2902        copy_stack( s, t, copy_el ); /***/
2903        return s;
2904}
2905_Bool stack_empty( const stack * s ) {
2906        return s->head == NULL;
2907}
2908void push_stack( stack * s, void * v ) {
2909        node * n = malloc( sizeof(node) ); /***/
2910        *n = (node){ v, s->head }; /***/
2911        s->head = n;
2912}
2913void * pop_stack( stack * s ) {
2914        node * n = s->head;
2915        s->head = n->next;
2916        void * v = n->value;
2917        free( n );
2918        return v;
2919}
2920\end{cfa}
2921\end{tabular}
2922\lstMakeShortInline@%
2923\end{flushleft}
2924
2925
2926\subsection{\CFA}
2927\label{s:CforallStack}
2928
2929\begin{flushleft}
2930\lstDeleteShortInline@%
2931\begin{tabular}{@{}l|@{\hspace{\parindentlnth}}l@{}}
2932\begin{cfa}[xleftmargin=0pt,aboveskip=0pt,belowskip=0pt]
2933forall( otype T ) {
2934        struct node {
2935                T value;
2936                node(T) * next;
2937        };
2938        struct stack { node(T) * head; };
2939        void ?{}( stack(T) & s, stack(T) t ) { // copy
2940                node(T) ** cr = &s.head;
2941                for ( node(T) * nx = t.head; nx; nx = nx->next ) {
2942                        *cr = alloc();
2943                        ((*cr)->value){ nx->value };
2944                        cr = &(*cr)->next;
2945                }
2946                *cr = 0;
2947        }
2948        void clear( stack(T) & s ) with( s ) {
2949                for ( node(T) * nx = head; nx; ) {
2950                        node(T) * cr = nx;
2951                        nx = cr->next;
2952                        ^(*cr){};
2953                        free( cr );
2954                }
2955                head = 0;
2956        }
2957
2958\end{cfa}
2959&
2960\begin{cfa}[xleftmargin=0pt,aboveskip=0pt,belowskip=0pt]
2961        void ?{}( stack(T) & s ) { (s.head){ 0 }; }
2962        void ^?{}( stack(T) & s) { clear( s ); }
2963        stack(T) ?=?( stack(T) & s, stack(T) t ) {
2964                if ( s.head == t.head ) return s;
2965                clear( s );
2966                s{ t };
2967                return s;
2968        }
2969        _Bool empty( const stack(T) & s ) {
2970                return s.head == 0;
2971        }
2972        void push( stack(T) & s, T value ) with( s ) {
2973                node(T) * n = alloc();
2974                (*n){ value, head };
2975                head = n;
2976        }
2977        T pop( stack(T) & s ) with( s ) {
2978                node(T) * n = head;
2979                head = n->next;
2980                T v = n->value;
2981                ^(*n){};
2982                free( n );
2983                return v;
2984        }
2985}
2986\end{cfa}
2987\end{tabular}
2988\lstMakeShortInline@%
2989\end{flushleft}
2990
2991
2992\subsection{\CC}
2993
2994\begin{flushleft}
2995\lstDeleteShortInline@%
2996\begin{tabular}{@{}l|@{\hspace{\parindentlnth}}l@{}}
2997\begin{cfa}[xleftmargin=0pt,aboveskip=0pt,belowskip=0pt]
2998template<typename T> struct stack {
2999        struct node {
3000                T value;
3001                node * next;
3002                node( const T & v, node * n = nullptr ) :
3003                        value( v ), next( n ) {}
3004        };
3005        node * head;
3006        void copy( const stack<T> & o ) {
3007                node ** cr = &head;
3008                for ( node * nx = o.head; nx; nx = nx->next ) {
3009                        *cr = new node{ nx->value }; /***/
3010                        cr = &(*cr)->next;
3011                }
3012                *cr = nullptr;
3013        }
3014        void clear() {
3015                for ( node * nx = head; nx; ) {
3016                        node * cr = nx;
3017                        nx = cr->next;
3018                        delete cr;
3019                }
3020                head = nullptr;
3021        }
3022\end{cfa}
3023&
3024\begin{cfa}[xleftmargin=0pt,aboveskip=0pt,belowskip=0pt]
3025        stack() : head( nullptr ) {}
3026        stack( const stack<T> & o ) { copy( o ); }
3027        ~stack() { clear(); }
3028        stack & operator=( const stack<T> & o ) {
3029                if ( this == &o ) return *this;
3030                clear();
3031                copy( o );
3032                return *this;
3033        }
3034        bool empty() const {
3035                return head == nullptr;
3036        }
3037        void push( const T & value ) {
3038                head = new node{ value, head };  /***/
3039        }
3040        T pop() {
3041                node * n = head;
3042                head = n->next;
3043                T v = std::move( n->value );
3044                delete n;
3045                return v;
3046        }
3047};
3048
3049\end{cfa}
3050\end{tabular}
3051\lstMakeShortInline@%
3052\end{flushleft}
3053
3054
3055\subsection{\CCV}
3056
3057\begin{flushleft}
3058\lstDeleteShortInline@%
3059\begin{tabular}{@{}l|@{\hspace{\parindentlnth}}l@{}}
3060\begin{cfa}[xleftmargin=0pt,aboveskip=0pt,belowskip=0pt]
3061struct stack {
3062        struct node {
3063                ptr<object> value;
3064                node * next;
3065                node( const object & v, node * n = nullptr ) :
3066                                value( v.new_copy() ), next( n ) {}
3067        };
3068        node * head;
3069        void copy( const stack & o ) {
3070                node ** cr = &head;
3071                for ( node * nx = o.head; nx; nx = nx->next ) {
3072                        *cr = new node{ *nx->value }; /***/
3073                        cr = &(*cr)->next;
3074                }
3075                *cr = nullptr;
3076        }
3077        void clear() {
3078                for ( node * nx = head; nx; ) {
3079                        node * cr = nx;
3080                        nx = cr->next;
3081                        delete cr;
3082                }
3083                head = nullptr;
3084        }
3085\end{cfa}
3086&
3087\begin{cfa}[xleftmargin=0pt,aboveskip=0pt,belowskip=0pt]
3088        stack() : head( nullptr ) {}
3089        stack( const stack & o ) { copy( o ); }
3090        ~stack() { clear(); }
3091        stack & operator=( const stack & o ) {
3092                if ( this == &o ) return *this;
3093                clear();
3094                copy( o );
3095                return *this;
3096        }
3097        bool empty() const {
3098                return head == nullptr;
3099        }
3100        void push( const object & value ) {
3101                head = new node{ value, head }; /***/
3102        }
3103        ptr<object> pop() {
3104                node * n = head;
3105                head = n->next;
3106                ptr<object> v = std::move( n->value );
3107                delete n;
3108                return v;
3109        }
3110};
3111
3112\end{cfa}
3113\end{tabular}
3114\lstMakeShortInline@%
3115\end{flushleft}
3116
3117
3118\end{document}
3119
3120% Local Variables: %
3121% tab-width: 4 %
3122% compile-command: "make" %
3123% End: %
Note: See TracBrowser for help on using the repository browser.