% Copyright 2005-2017 Cisco Systems, Inc.

%

% Licensed under the Apache License, Version 2.0 (the "License");

% you may not use this file except in compliance with the License.

% You may obtain a copy of the License at

%

% http://www.apache.org/licenses/LICENSE-2.0

%

% Unless required by applicable law or agreed to in writing, software

% distributed under the License is distributed on an "AS IS" BASIS,

% WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.

% See the License for the specific language governing permissions and

% limitations under the License.

\chapter{Binding Forms\label{CHPTBINDING}}

This chapter describes {\ChezScheme} extensions to the set of Revised$^6$

Report binding forms.

See Chapter~\ref{TSPL:CHPTBINDING} of {\TSPLFOUR} or the Revised$^6$ Report

for a description of standard binding forms.

\section{Definitions\label{SECTDEFINITIONS}}

A \index{definitions}definition in Revised$^6$ Report Scheme is a

\index{\scheme{define}}\index{variable definition}variable definition,

\index{\scheme{define-syntax}}\index{keyword definition}keyword

definition, or derived definition, i.e., a syntactic extension that

expands into a definition.

In addition, the forms within a

\index{\scheme{begin}}\scheme{begin}

expression appearing after a sequence

of definitions is spliced onto the end of the sequence of definitions

so that definitions at the front of the \scheme{begin} expression

are treated as if they were part of the outer sequence of definitions.

A \index{\scheme{let-syntax}}\scheme{let-syntax} or

\index{\scheme{letrec-syntax}}\scheme{letrec-syntax} form

is treated similarly, so that definitions at the front of the body

are treated as if they were part of the outer sequence of definitions,

albeit scoped where the bindings of the \scheme{let-syntax} or

\scheme{letrec-syntax} form are visible.

{\ChezScheme} extends the set of definitions to include

\index{modules}\index{\scheme{module}}\scheme{module} forms,

\index{\scheme{import}}\scheme{import} forms,

\index{\scheme{import-only}}\scheme{import-only} forms,

\index{\scheme{meta}}\scheme{meta} definitions, and

\index{\scheme{alias}}\scheme{alias} forms, although the

\scheme{module}, \scheme{import}, \scheme{import-only},

\scheme{meta}, and \scheme{alias} keywords are not available

in a library or RNRS top-level program unless the

\scheme{scheme} library is included in the library or

top-level programs imports.

These forms are described in Chatper~\ref{CHPTSYNTAX}.

In Revised$^6$ Report Scheme, definitions can appear at the front of

a \scheme{lambda} or similar body (e.g., a \scheme{let} or \scheme{letrec}

body), at the front of a library body, or intermixed with expressions

within an RNRS top-level program body.

In {\ChezScheme}, definitions may also be used in the

interactive top-level, i.e., they can be intermixed with expressions in

the REPL or in program text to be loaded from a file

via \index{\scheme{load}}\scheme{load} (Section~\ref{SECTMISCCOMPILEEVAL}).

The Revised$^6$ Report does not mandate the existence nor specify the

semantics of an interactive top-level, nor of a \scheme{load}

procedure.

The macro expander uses the same two-pass algorithm for expanding

top-level \scheme{begin} expressions as it uses for a \scheme{lambda},

\scheme{library}, or top-level program body.

(This algorithm is described in Section~\ref{SECTSYNTAXDEFINITIONS} of {\TSPLFOUR}.)

As a result,

\schemedisplay

(begin

(define-syntax a (identifier-syntax 3))

(define x a))

\endschemedisplay

and

\schemedisplay

(begin

(define x a)

(define-syntax a (identifier-syntax 3)))

\endschemedisplay

both result in the giving \scheme{x} the value 3,

even though an unbound variable reference to \scheme{a} would result if

the two forms within the latter \scheme{begin} expression where run

independently at top level.

Similarly, the \scheme{begin} form produced by a use of

\schemedisplay

(define-syntax define-constant

(syntax-rules ()

[(_ x e)

(begin

(define t e)

(define-syntax x (identifier-syntax t)))]))

\endschemedisplay

and the \scheme{begin} form produced by a use of

\schemedisplay

(define-syntax define-constant

(syntax-rules ()

[(_ x e)

(begin

(define-syntax x (identifier-syntax t))

(define t e))]))

\endschemedisplay

are equivalent.

The Revised$^6$ Report specifies that internal variable definitions be

treated like \scheme{letrec*}, while earlier reports required internal

variable definitions to be treated like \scheme{letrec}.

By default, {\ChezScheme} implements the Revised$^6$ Report semantics for

internal variable definitions, as for all other things, but this behavior

may be overridden via the \scheme{internal-defines-as-letrec*} parameter.

%----------------------------------------------------------------------------

\entryheader

\formdef{internal-defines-as-letrec*}{\categorythreadparameter}{internal-defines-as-letrec*}

\listlibraries

\endentryheader

\noindent

When this parameter is set to \scheme{#t} (the default), internal variable

definitions are evaluated using \scheme{letrec*} semantics.

It may be set to \scheme{#f} to revert to the \scheme{letrec} semantics

for internal variable definitions, for backward compatibility.

\section{Multiple-value Definitions}

%----------------------------------------------------------------------------

\noskipentryheader

\formdef{define-values}{\categorysyntax}{(define-values formals \var{expr})}

\listlibraries

\endnoskipentryheader

A \scheme{define-values} form is a definition and can appear anywhere

other definitions can appear.

It is like a \scheme{define} form but permits an arbitrary formals list

(like \scheme{lambda}) on the left-hand side.

It evaluates \var{expr} and binds the variables appearing in \var{formals}

to the resulting values, in the same manner as the formal parameters of a

procedure are bound to its arguments.

\schemedisplay

(let ()

(define-values (x y) (values 1 2))

(list x y)) ;=> (1 2)

(let ()

(define-values (x y . z) (values 1 2 3 4))

(list x y z)) ;=> (1 2 (3 4))

\endschemedisplay

A \scheme{define-values} form expands into a sequence of definitions, the

first for a hidden temporary bound to a data structure holding the values

returned by \var{expr} and the remainder binding each of the formals to

the corresponding value or list of values, extracted from the data

structure via a reference to the temporary.

Because the temporary must be defined before the other variables are

defined, this works for internal \scheme{define-values} forms only if

\scheme{internal-defines-as-letrec*} is set to the default value

\scheme{#t}.

\section{Recursive Bindings}

%----------------------------------------------------------------------------

\noskipentryheader

\formdef{rec}{\categorysyntax}{(rec \var{var} \var{expr})}

\returns value of \var{expr}

\listlibraries

\endnoskipentryheader

\noindent

The syntactic form \scheme{rec} creates a \index{recursive object}recursive object from \var{expr} by

establishing a binding of \var{var} within \var{expr} to the value of \var{expr}.

In essence, it is a special case of \scheme{letrec} for self-recursive objects.

This form is useful for creating recursive objects (especially procedures)

that do not depend on external variables for the recursion, which are

sometimes undesirable because the external bindings can change.

For example, a recursive procedure defined at top level depends on the value

of the top-level variable given as its name.

If the value of this variable should change, the meaning of the procedure

itself would change.

If the procedure is defined instead with \scheme{rec}, its meaning is independent

of the variable to which it is bound.

\schemedisplay

(map (rec sum

(lambda (x)

(if (= x 0)

0

(+ x (sum (- x 1))))))

'(0 1 2 3 4 5)) ;=> (0 1 3 6 10 15)

(define cycle

(rec self

(list (lambda () self))))

(eq? ((car cycle)) cycle) ;=> #t

\endschemedisplay

\noindent

The definition below expands \scheme{rec} in terms of \scheme{letrec}.

\schemedisplay

(define-syntax rec

(syntax-rules ()

[(_ x e) (letrec ((x e)) x)]))

\endschemedisplay

\section{Fluid Bindings}

%----------------------------------------------------------------------------

\noskipentryheader

\formdef{fluid-let}{\categorysyntax}{(fluid-let ((\var{var} \var{expr}) \dots) \var{body_1} \var{body_2} \dots)}

\returns the values of the body \scheme{\var{body_1} \var{body_2} \dots}

\listlibraries

\endentryheader

\noindent

\index{fluid binding}\index{assignments}The syntactic form \scheme{fluid-let}

provides a way to temporarily assign values to a set of variables.

The new values are in effect only during the evaluation of the

body of the \scheme{fluid-let} expression.

The scopes of the variables are not determined by \scheme{fluid-let}; as with

\scheme{set!}, the variables must be bound at top level or by an enclosing

\scheme{lambda} or other binding form.

It is possible, therefore, to control the scope of a variable with

\scheme{lambda} or \scheme{let} while establishing a temporary

value with \scheme{fluid-let}.

Although it is similar in appearance to \scheme{let}, its operation is more

like that of \scheme{set!}.

Each \var{var} is assigned, as with \scheme{set!}, to the value of the

corresponding \var{expr} within the body \scheme{\var{body_1} \var{body_2} \dots}.

Should the body

exit normally or by invoking a continuation made outside of the body

(see \scheme{call/cc}), the values in effect before the bindings were changed

are restored.

Should control return back to the body by the invocation of a continuation

created within the body, the bindings are changed once again to the values

in effect when the body last exited.

Fluid bindings are most useful for

maintaining variables that must be shared by a group of procedures.

Upon entry to the group of procedures, the shared variables are fluidly

bound to a new set of initial values so that on exit the original values

are restored automatically.

In this way, the group of procedures itself can be reentrant; it may call

itself directly or indirectly without affecting the values of its shared

variables.

\index{special bindings (in Lisp)}Fluid bindings are similar to

\emph{special} bindings in Common Lisp~\cite{Steele:common}, except that

(1) there is a single namespace for both lexical and fluid bindings, and

(2) the scope of a fluidly bound variable is not necessarily global.

\schemedisplay

(let ([x 3])

(+ (fluid-let ([x 5])

x)

x)) ;=> 8

(let ([x 'a])

(letrec ([f (lambda (y) (cons x y))])

(fluid-let ([x 'b])

(f 'c)))) ;=> (b . c)

(let ([x 'a])

(call/cc

(lambda (k)

(fluid-let ([x 'b])

(letrec ([f (lambda (y) (k '*))])

(f '*)))))

x) ;=> a

\endschemedisplay

\noindent

\scheme{fluid-let} may be defined in terms of \scheme{dynamic-wind} as follows.

\schemedisplay\label{defn:fluid-let}

(define-syntax fluid-let

(lambda (x)

(syntax-case x ()

[(_ () b1 b2 ...) #'(let () b1 b2 ...)]

[(_ ((x e) ...) b1 b2 ...)

(andmap identifier? #'(x ...))

(with-syntax ([(y ...) (generate-temporaries #'(x ...))])

#'(let ([y e] ...)

(let ([swap (lambda ()

(let ([t x]) (set! x y) (set! y t))

...)])

(dynamic-wind swap (lambda () b1 b2 ...) swap))))])))

\endschemedisplay

\section{Top-Level Bindings\label{SECTBINDINGTOPLEVEL}}

The procedures described in this section allow the direct manipulation

of \index{top-level values}top-level bindings for variables

and keywords.

They are intended primarily to support the definition of interpreters

or compilers for Scheme in Scheme but may be used to access or alter

top-level bindings anywhere within a program whether at top level or not.

%----------------------------------------------------------------------------

\entryheader

\formdef{define-top-level-value}{\categoryprocedure}{(define-top-level-value \var{symbol} \var{obj})}

\formdef{define-top-level-value}{\categoryprocedure}{(define-top-level-value \var{symbol} \var{obj} \var{env})}

\returns unspecified

\listlibraries

\endentryheader

\noindent

\scheme{define-top-level-value} is used to establish a binding

for the variable named by \var{symbol} to the value \var{obj}

in the environment \var{env}.

If \var{env} is not provided, it defaults to the

value of \scheme{interaction-environment}, i.e., the

top-level evaluation environment

(Section~\ref{SECTMISCENVIRONMENTS}).

An exception is raised with condition type \scheme{&assertion} if

\var{env} is not mutable.

A call to \scheme{define-top-level-value} is similar to a top-level

\index{\scheme{define}}\scheme{define} form, except that a call to

\scheme{define-top-level-value} need not occur at top-level and

the variable for which the binding is to be established can be

determined at run time, as can the environment.

\schemedisplay

(begin

(define-top-level-value 'xyz "hi")

xyz) ;=> "hi"

(let ([var 'xyz])

(define-top-level-value var "mom")

(list var xyz)) ;=> (xyz "mom")

\endschemedisplay

%----------------------------------------------------------------------------

\entryheader

\formdef{set-top-level-value!}{\categoryprocedure}{(set-top-level-value! \var{symbol} \var{obj})}

\formdef{set-top-level-value!}{\categoryprocedure}{(set-top-level-value! \var{symbol} \var{obj} \var{env})}

\returns unspecified

\listlibraries

\endentryheader

\noindent

\index{assignments}\scheme{set-top-level-value!} assigns

the variable named by \var{symbol} to the value \var{obj}

in the environment \var{env}.

If \var{env} is not provided, it defaults to the

value of \scheme{interaction-environment}, i.e., the

top-level evaluation environment

(Section~\ref{SECTMISCENVIRONMENTS}).

An exception is raised with condition type \scheme{&assertion} if the

identifier named by \var{symbol} is not defined as a variable in \var{env}

or if the variable or environment is not mutable.

\scheme{set-top-level-value!} is similar to

\index{\scheme{set!}}\scheme{set!} when \scheme{set!}

is used on top-level variables except that the variable to be assigned

can be determined at run time, as can the environment.

\schemedisplay

(let ([v (let ([cons list])

(set-top-level-value! 'cons +)

(cons 3 4))])

(list v (cons 3 4))) ;=> ((3 4) 7)

\endschemedisplay

%----------------------------------------------------------------------------

\entryheader

\formdef{top-level-value}{\categoryprocedure}{(top-level-value \var{symbol})}

\formdef{top-level-value}{\categoryprocedure}{(top-level-value \var{symbol} \var{env})}

\returns the top-level value of the variable named by \var{symbol} in \var{env}

\listlibraries

\endentryheader

\noindent

If \var{env} is not provided, it defaults to the

value of \scheme{interaction-environment}, i.e., the

top-level evaluation environment

(Section~\ref{SECTMISCENVIRONMENTS}).

An exception is raised with condition type \scheme{&assertion} if the

identifier named by \var{symbol} is not defined as a variable in \var{env}.

\scheme{top-level-value} is similar to a top-level variable reference

except that the variable to be referenced can be determined at run time,

as can the environment.

\schemedisplay

(let ([cons +])

(list (cons 3 4)

((top-level-value 'cons) 3 4))) ;=> (7 (3 . 4))

(define e (copy-environment (scheme-environment)))

(define-top-level-value 'pi 3.14 e)

(top-level-value 'pi e) ;=> 3.14

(set-top-level-value! 'pi 3.1416 e)

(top-level-value 'pi e) ;=> 3.1416

\endschemedisplay

%----------------------------------------------------------------------------

\entryheader

\formdef{top-level-bound?}{\categoryprocedure}{(top-level-bound? \var{symbol})}

\formdef{top-level-bound?}{\categoryprocedure}{(top-level-bound? \var{symbol} \var{env})}

\returns \scheme{#t} if \var{symbol} is defined as a variable in \var{env}, \scheme{#f} otherwise

\listlibraries

\endentryheader

\noindent

If \var{env} is not provided, it defaults to the

value of \scheme{interaction-environment}, i.e., the

top-level evaluation environment

(Section~\ref{SECTMISCENVIRONMENTS}).

This predicate is useful in an interpreter to check for the existence of

a top-level binding before requesting the value with

\scheme{top-level-value}.

\schemedisplay

(top-level-bound? 'xyz) ;=> #f

(begin

(define-top-level-value 'xyz 3)

(top-level-bound? 'xyz)) ;=> #t

(define e (copy-environment (interaction-environment)))

(define-top-level-value 'pi 3.14 e)

(top-level-bound? 'pi) ;=> #f

(top-level-bound? 'pi e) ;=> #t

\endschemedisplay

%----------------------------------------------------------------------------

\entryheader

\formdef{top-level-mutable?}{\categoryprocedure}{(top-level-mutable? \var{symbol})}

\formdef{top-level-mutable?}{\categoryprocedure}{(top-level-mutable? \var{symbol} \var{env})}

\returns \scheme{#t} if \var{symbol} is mutable in \var{env}, \scheme{#f} otherwise

\listlibraries

\endentryheader

\noindent

If \var{env} is not provided, it defaults to the

value of \scheme{interaction-environment}, i.e., the

top-level evaluation environment

(Section~\ref{SECTMISCENVIRONMENTS}).

This predicate is useful in an interpreter to check whether a variable

can be assigned before assigning it with

\scheme{set-top-level-value!}.

\schemedisplay

(define xyz 3)

(top-level-mutable? 'xyz) ;=> #t

(set-top-level-value! 'xyz 4)

(top-level-value 'xyz) ;=> 4

(define e (copy-environment (interaction-environment) #f))

(top-level-mutable? 'xyz e) ;=> #f

(set-top-level-value! 'xyz e) ;=> \var{exception: xyz is immutable}

\endschemedisplay

%----------------------------------------------------------------------------

\entryheader

\formdef{define-top-level-syntax}{\categoryprocedure}{(define-top-level-syntax \var{symbol} \var{obj})}

\formdef{define-top-level-syntax}{\categoryprocedure}{(define-top-level-syntax \var{symbol} \var{obj} \var{env})}

\returns unspecified

\listlibraries

\endentryheader

\noindent

\scheme{define-top-level-syntax} is used to establish a top-level binding

for the identifier named by \var{symbol} to the value of \var{obj}

in the environment \var{env}.

The value must be a procedure, the result of a call to

\scheme{make-variable-transformer}, or the result of a call to

\scheme{top-level-syntax}.

If \var{env} is not provided, it defaults to the

value of \scheme{interaction-environment}, i.e., the

top-level evaluation environment

(Section~\ref{SECTMISCENVIRONMENTS}).

An exception is raised with condition type \scheme{&assertion} if

\var{env} is not mutable.

A call to \scheme{define-top-level-syntax} is similar to a top-level

\index{\scheme{define-syntax}}\scheme{define-syntax} form, except that a call to

\scheme{define-top-level-syntax} need not occur at top-level and

the identifier for which the binding is to be established can be

determined at run time, as can the environment.

\schemedisplay

(define-top-level-syntax 'let1

(syntax-rules ()

[(_ x e b1 b2 ...) (let ([x e]) b1 b2 ...)]))

(let1 a 3 (+ a 1)) ;=> 4

\endschemedisplay

\scheme{define-top-level-syntax} can also be used to attach

to an identifier arbitrary compile-time bindings obtained

via \scheme{top-level-syntax}.

%----------------------------------------------------------------------------

\entryheader

\formdef{top-level-syntax}{\categoryprocedure}{(top-level-syntax \var{symbol})}

\formdef{top-level-syntax}{\categoryprocedure}{(top-level-syntax \var{symbol} \var{env})}

\returns unspecified

\listlibraries

\endentryheader

\noindent

\scheme{top-level-syntax} is used to retrieve the transformer, compile-time

value, or other compile-time binding to which

the identifier named by \var{symbol} is bound in the environment \var{env}.

If \var{env} is not provided, it defaults to the

value of \scheme{interaction-environment}, i.e., the

top-level evaluation environment

(Section~\ref{SECTMISCENVIRONMENTS}).

All identifiers bound in an environment have compile-time bindings, including

variables.

An exception is raised with condition type \scheme{&assertion} if the

identifier named by \var{symbol} is not defined as a keyword in \var{env}.

\schemedisplay

(define-top-level-syntax 'also-let (top-level-syntax 'let))

(also-let ([x 3] [y 4]) (+ x y)) ;=> 7

(define foo 17)

(define-top-level-syntax 'also-foo (top-level-syntax 'foo))

also-foo ;=> 17

(set! also-foo 23)

also-foo ;=> 23

foo ;=> 23

\endschemedisplay

The effect of the last example can be had more clearly with \scheme{alias}:

\schemedisplay

(define foo 17)

(alias also-foo foo)

also-foo ;=> 17

(set! also-foo 23)

also-foo ;=> 23

foo ;=> 23

\endschemedisplay

%----------------------------------------------------------------------------

\entryheader

\formdef{top-level-syntax?}{\categoryprocedure}{(top-level-syntax? \var{symbol})}

\formdef{top-level-syntax?}{\categoryprocedure}{(top-level-syntax? \var{symbol} \var{env})}

\returns \scheme{#t} if \var{symbol} is bound as a keyword in \var{env}, \scheme{#f} otherwise

\listlibraries

\endentryheader

\noindent

If \var{env} is not provided, it defaults to the

value of \scheme{interaction-environment}, i.e., the

top-level evaluation environment

(Section~\ref{SECTMISCENVIRONMENTS}).

All identifiers bound in an environment have compile-time bindings, including

variables, so this predicate amounts to a bound check, but is more general

than \scheme{top-level-bound?}, which returns true only for bound variables.

\schemedisplay

(define xyz 'hello)

(top-level-syntax? 'cons) ;=> #t

(top-level-syntax? 'lambda) ;=> #t

(top-level-syntax? 'hello) ;=> #t

(top-level-syntax? 'cons (scheme-environment)) ;=> #t

(top-level-syntax? 'lambda (scheme-environment)) ;=> #t

(top-level-syntax? 'hello (scheme-environment)) ;=> #f

\endschemedisplay