Modules and macros
CHICKEN supports standard R5RS syntax-rules macros and a low-level
macro system based on explicit renaming.
Macro definitions
define-syntax
[syntax] (define-syntax IDENTIFIER TRANSFORMER)
Defines a macro named IDENTIFIER that will transform an expression
with IDENTIFIER in operator position according to TRANSFORMER.
The transformer expression must be a procedure with three arguments or
a syntax-rules form. If syntax-rules is used, the usual R5RS
semantics apply. If TRANSFORMER is a procedure, then it will
be called on expansion with the complete s-expression of the macro
invocation, a rename procedure that hygienically renames identifiers
and a comparison procedure that compares (possibly renamed) identifiers.
define-syntax may be used to define local macros that are visible
throughout the rest of the body in which the definition occurred, i.e.
(let ()
...
(define-syntax foo ...)
(define-syntax bar ...)
...)
is expanded into
(let ()
...
(letrec-syntax ((foo ...) (bar ...))
...) )
syntax-rules partially supports SRFI-46
in allowing the ellipsis identifier to be user-defined by passing it as the first
argument to the syntax-rules form.
The effect of destructively modifying the s-expression passed to a
transformer procedure is undefined.
define-compiled-syntax
[syntax] (define-compiled-syntax IDENTIFIER TRANSFORMER)
Equivalent to define-syntax, but when compiled, will also define the macro
at runtime.
syntax
[syntax] (syntax EXPRESSION)
Similar to quote but retains syntactical context information for
embedded identifiers.
strip-syntax
[procedure] (strip-syntax EXPRESSION)
Strips all syntactical information from EXPRESSION, returning a new expression
where symbols have all context-information removed.
Explicit renaming macros
The low-level macro facility that CHICKEN provides is called "explicit
renaming" and allows writing hygienic or non-hygienic macros procedurally.
When given a lambda-expression instead of a syntax-rules form,
define-syntax evaluates the procedure in a distinct expansion
environment (initially having access to the exported identifiers
of the scheme module). The procedure takes an expression and two
other arguments and returns a transformed expression.
For example, the transformation
procedure for a call macro such that
(call proc arg ...) expands
into (proc arg ...) can be written as
(lambda (exp rename compare)
(cdr exp))
Expressions are represented as lists in the traditional manner,
except that identifiers are represented as special uninterned symbols.
The second argument to a transformation procedure is a renaming procedure that
takes the representation of an identifier as its argument and returns the
representation of a fresh identifier that occurs nowhere else in the
program. For example, the transformation procedure for a simplified
version of the let macro might be written as
(lambda (exp rename compare)
(let ((vars (map car (cadr exp)))
(inits (map cadr (cadr exp)))
(body (cddr exp)))
`((lambda ,vars ,@body)
,@inits)))
This would not be hygienic, however. A
hygienic let macro must rename the identifier lambda to protect it
from being captured by a local binding. The renaming effectively
creates a fresh alias for lambda, one that cannot be captured by
any subsequent binding:
(lambda (exp rename compare)
(let ((vars (map car (cadr exp)))
(inits (map cadr (cadr exp)))
(body (cddr exp)))
`((,(rename 'lambda) ,vars ,@body)
,@inits)))
The expression returned by the transformation procedure will be
expanded in the syntactic environment obtained from the syntactic
environment of the macro application by binding any fresh identifiers
generated by the renaming procedure to the denotations of the original
identifiers in the syntactic environment in which the macro was
defined. This means that a renamed identifier will denote the same
thing as the original identifier unless the transformation procedure
that renamed the identifier placed an occurrence of it in a binding
position.
Identifiers obtained from any two calls to the renaming procedure with
the same argument will necessarily be the same, but will denote the
same syntactical binding. It is an error if the renaming procedure is
called after the transformation procedure has returned.
The third argument to a transformation procedure is a comparison
predicate that takes the representations of two identifiers as its
arguments and returns true if and only if they denote the same thing
in the syntactic environment that will be used to expand the
transformed macro application. For example, the transformation
procedure for a simplified version of the cond macro can be written
as
(lambda (exp rename compare)
(let ((clauses (cdr exp)))
(if (null? clauses)
`(,(rename 'quote) unspecified)
(let* ((first (car clauses))
(rest (cdr clauses))
(test (car first)))
(cond ((and (symbol? test)
(compare test (rename 'else)))
`(,(rename 'begin) ,@(cdr first)))
(else `(,(rename 'if)
,test
(,(rename 'begin) ,@(cdr first))
(cond ,@rest))))))))
In this example the identifier else is renamed before being passed
to the comparison predicate, so the comparison will be true if and
only if the test expression is an identifier that denotes the same
thing in the syntactic environment of the expression being transformed
as else denotes in the syntactic environment in which the cond
macro was defined. If else were not renamed before being passed to
the comparison predicate, then it would match a local variable that
happened to be named else, and the macro would not be hygienic.
Some macros are non-hygienic by design. For example, the
following defines a loop macro that implicitly binds exit to an
escape procedure. The binding of exit is intended to capture free
references to exit in the body of the loop, so exit is not
renamed.
(define-syntax loop
(lambda (x r c)
(let ((body (cdr x)))
`(,(r 'call-with-current-continuation)
(,(r 'lambda) (exit)
(,(r 'let) ,(r 'f) () ,@body (,(r 'f))))))))
Suppose a while macro is implemented using loop, with the intent
that exit may be used to escape from the while loop. The while
macro cannot be written as
(define-syntax while
(syntax-rules ()
((while test body ...)
(loop (if (not test) (exit #f))
body ...))))
because the reference to exit that is inserted by the while macro
is intended to be captured by the binding of exit that will be
inserted by the loop macro. In other words, this while macro is
not hygienic. Like loop, it must be written using procedurally:
(define-syntax while
(lambda (x r c)
(let ((test (cadr x))
(body (cddr x)))
`(,(r 'loop)
(,(r 'if) (,(r 'not) ,test) (exit #f))
,@body))))
Modules
To allow some control over visible bindings and to organize code at
the global level, a simple module system is available. A module
defines a set of toplevel expressions that are initially evaluated in
an empty syntactical environment. By importing other modules,
exported value- and macro-bindings are made visible inside the
environment of the module that imports them.
Note that modules are purely syntactical - they do not change the
control flow or delay the execution of the contained toplevel
forms. The body of a module is executed at load-time, when code is
loaded or accessed via the uses declaration, just like normal
toplevel expressions. Exported macro-definitions are compiled as
well, and can be accessed in interpreted or compiled code by loading
and importing the compiled file that contains the module.
Imported toplevel bindings can be assigned (with set!), any modifications
to these will change the global value and will be visible to other
modules that export or import the same toplevel binding.
A module is initially empty (has no visible bindings). You must at least
import the scheme module to do anything useful. To access any
of the non-standard macros and procedures, import the chicken
module.
CHICKEN's module system has the following features and shortcomings:
- Indirect exports for syntax definitions must be listed
- Separation of compile/expansion-time and run-time code is provided, which allows cross compilation
- Module-generating code is only created, when needed
- Supports batch-compilation of separate compilation units
- No separate "identifier" type is used, all identifiers appearing in code and processed in expansions are symbols
- The module system is fully optional
module
[syntax] (module NAME (EXPORT ...) BODY ...)
[syntax] (module NAME * BODY ...)
Defines a module with the name NAME, a set of exported bindings
and a contained sequence of toplevel expressions that are evaluated in
an empty syntactical environment. EXPORT may be a symbol or a list
of the form (IDENTIFIER1 IDENTIFIER2 ...). In the former case the
identifier given is exported from the module and can be imported at
the toplevel or in other modules. The latter case exports
IDENTIFIER1 (which should name a macro) and also arranges for the
remaining identifiers in the list to be visible in the expansion of
the macro (this is a hint to the module expander to export bindings
referenced by syntax-definitions which make use of them, but which
would normally be internal to the module - that allows some
optimization and also records).
Nested modules, modules not at toplevel (i.e. local modules) or
mutually recursive modules are not supported.
When compiled, the module information, including exported macros
is stored in the generated binary and available when loading
it into interpreted or compiled code. Note that this is different
to normal macros (outside of module declarations), which are normally
not exported from compiled code.
As a special case, specifying * instead of an export-list
will export all definitions.
Note that the module system is only a device for controlling the
mapping of identifiers to value or syntax bindings. Modules do not
instantiate separate environments that contain their own bindings, as
do many other module systems. Redefinition of value or syntax bindings
will modify the original, imported definition.
export
[syntax] (export EXPORT ...)
Allows augmenting module-exports from inside the module-body.
EXPORT is if the same form as an export-specifier in a
module export list. An export must precede its first occurrence
(either use or definition).
import
[syntax] (import IMPORT ...)
Imports module bindings into the current syntactical environment.
The visibility of any imported bindings is limited to the current
module, if used inside a module-definition, or to the current
compilation unit, if compiled and used outside of a module.
Importing a module does not load or link it - this is a separate
operation from importing its bindings.
IMPORT may be a module name, or an import specifier.
An IMPORT defines a set of bindings that are to be made visible
in the current scope.
Note that the imported bindings are only visible in the next toplevel
expression (regardless of whether the import appears inside or outside
a module):
(begin
(import m1)
... ; imports not visible here
... ; imports visible here
only
[import specifier] (only IMPORT IDENTIFIER ...)
Only import the listed value- or syntax bindings from the set given
by IMPORT.
except
[import specifier] (except IMPORT IDENTIFIER ...)
Remove the listed identifiers from the import-set defined by IMPORT.
rename
[import specifier] (rename IMPORT (OLD1 NEW1) ...)
Renames identifiers imported from IMPORT.
prefix
[import specifier] (prefix IMPORT SYMBOL)
Prefixes all imported identifiers with SYMBOL.
import-for-syntax
[syntax] (import-for-syntax IMPORT ...)
Similar to import, but imports exported bindings of a module into
the environment in which macro transformers are evaluated.
Note: currently this isn't fully correct - value bindings are still
imported into the normal environment because a separate import
environment for syntax has not been implemented (syntactic bindings
are kept separate correctly).
import libraries
import libraries allow the syntactical (compile-time)
and run-time parts of a compiled module to be separated into a normal
compiled file and a shared library that only contains macro definitions
and module information. This reduces the size of executables and
simplifies compiling code that uses modules for a different architecture
than the machine the compiler is executing on (i.e. "cross" compilation).
By using the emit-import-library compiler-option or declaration,
a separate file is generated that only contains syntactical information
(including macros) for a module. import will automatically find and
load an import library for a currently unknown module, if the import-
library is either in the extension repository or the current include
path. Import libraries may also be explicitly loaded into the
compiler by using the -extend compiler option. Interpreted code
can simply load the import library to make the module-definition
available. Macro-support definitions defined with define-for-syntax
and expansion-time expressions of the form (begin-for-syntax ...)
will be added to import libraries to make them available for exported
macros. Note that these definitions will ruthlessly pollute the
toplevel namespace and so they should be used sparingly.
Predefined modules
Import libraries for the following modules are initially
available:
[module] scheme
Exports the standard R5RS bindings.
[module] chicken
Everything from the library, eval and expand library units.
[module] extras
[module] data-structures
[module] ports
[module] lolevel
[module] posix
[module] regex
[module] srfi-1
[module] srfi-4
[module] srfi-13
[module] srfi-14
[module] srfi-18
[module] srfi-69
[module] tcp
[module] utils
Modules exporting the bindings from the respective library units.
[module] foreign
Exports all macros and procedures that are used to access foreign
C/C++ code.
Examples of using modules
Here is a silly little test module to demonstrate how modules
are defined and used:
- ; hello.scm
(module test (hello greet)
(import scheme)
(define-syntax greet
(syntax-rules ()
((_ whom)
(begin
(display "Hello, ")
(display whom)
(display " !\n") ) ) ) )
(define (hello)
(greet "world") ) )
The module test exports one value (hello) and one syntax
binding (greet). To use it in csi, the interpreter,
simply load and import it:
#;1> ,l hello.scm
; loading hello.scm ...
; loading /usr/local/lib/chicken/4/scheme.import.so ...
#;1> (import test)
#;2> (hello)
Hello, world !
#;3> (greet "you")
Hello, you !
The module can easily be compiled
% csc -s hello.scm
and used in an identical manner:
#;1> ,l hello.so
; loading hello.so ...
#;1> (import test)
#;2> (hello)
Hello, world !
#;3> (greet "you")
Hello, you !
If you want to keep macro-definitions in a separate file, use import
libraries:
% csc -s hello.scm -j test
% csc -s test.import.scm
#;1> ,l hello.so
; loading hello.so ...
#;1> (import test)
; loading ./test.import.so ...
#;2> (hello)
Hello, world !
#;3> (greet "you")
Hello, you !
If an import library (compiled or in source-form) is located
somewhere in the extensions-repository or include path, it
is automatically loaded on import. Otherwise you have to
load it manually:
#;1> ,l hello.so
; loading hello.so ...
#;1> ,l test.import.so
; loading test.import.so ...
#;1> (import test)
#;2>
Caveats
The macro- and module system has been implemented relatively
recently and is likely to contain bugs. Please contact the
maintainers if you encounter behavior that you think is
not correct or that triggers an error where there shouldn't
be one.
- In evaluated code, loading a file containing module information and importing from a module contained in that file will not work when the loading and import is performed in the same toplevel expression (this does not apply, when import libraries are used)
- Currently value bindings imported by import and import-for-syntax share the same import-environment.
Previous: Non-standard macros and special forms
Next: Declarations