it still has quite a few. @xref{Contributed Modules}.
@menu
-* Garbage Collection::
-* Metaobject Protocol::
-* Support For Unix::
-* Customization Hooks for Users::
-* Tools To Help Developers::
-* Resolution of Name Conflicts::
-* Stale Extensions::
-* Efficiency Hacks::
+* Reader Extensions::
+* Package-Local Nicknames::
+* Package Variance::
+* Garbage Collection::
+* Metaobject Protocol::
+* Extensible Sequences::
+* Support For Unix::
+* Customization Hooks for Users::
+* Tools To Help Developers::
+* Resolution of Name Conflicts::
+* Hash Table Extensions::
+* Random Number Generation::
+* Miscellaneous Extensions::
+* Stale Extensions::
+* Efficiency Hacks::
@end menu
+@node Reader Extensions
+@comment node-name, next, previous, up
+@section Reader Extensions
+@cindex Reader Extensions
+
+SBCL supports extended package prefix syntax, which allows specifying
+an alternate package instead of @code{*package*} for the reader to use
+as the default package for interning symbols:
+
+@lisp
+<package-name>::<form-with-interning-into-package>
+@end lisp
+
+Example:
+
+@lisp
+ 'foo::(bar quux zot) == '(foo::bar foo::quux foo::zot)
+@end lisp
+
+Doesn't alter @code{*package*}: if @code{foo::bar} would cause a
+read-time package lock violation, so does @code{foo::(bar)}.
+
+@node Package-Local Nicknames
+@comment node-name, next, previous, up
+@section Package-Local Nicknames
+@cindex Package-Local Nicknames
+
+SBCL allows giving packages local nicknames: they allow short and
+easy-to-use names to be used without fear of name conflict associated
+with normal nicknames.
+
+A local nickname is valid only when inside the package for which it
+has been specified. Different packages can use same local nickname for
+different global names, or different local nickname for same global
+name.
+
+Symbol @code{:package-local-nicknames} in @code{*features*} denotes the
+support for this feature.
+
+@findex @cl{defpackage}
+@defmac @cl{defpackage} name [[option]]* @result{} package
+
+Options are extended to include
+
+@itemize
+@item
+@code{:local-nicknames (@var{local-nickname} @var{actual-package-name})*}
+
+The package has the specified local nicknames for the corresponding
+actual packages.
+@end itemize
+
+Example:
+
+@lisp
+(defpackage :bar (:intern "X"))
+(defpackage :foo (:intern "X"))
+(defpackage :quux (:use :cl) (:local-nicknames (:bar :foo) (:foo :bar)))
+(find-symbol "X" :foo) ; => FOO::X
+(find-symbol "X" :bar) ; => BAR::X
+(let ((*package* (find-package :quux)))
+ (find-symbol "X" :foo)) ; => BAR::X
+(let ((*package* (find-package :quux)))
+ (find-symbol "X" :bar)) ; => FOO::X
+@end lisp
+@end defmac
+
+@include fun-sb-ext-package-local-nicknames.texinfo
+@include fun-sb-ext-package-locally-nicknamed-by-list.texinfo
+@include fun-sb-ext-add-package-local-nickname.texinfo
+@include fun-sb-ext-remove-package-local-nickname.texinfo
+
+@node Package Variance
+@comment node-name, next, previous, up
+@section Package Variance
+
+Common Lisp standard specifies that ``If the new definition is at
+variance with the current state of that package, the consequences are
+undefined;'' SBCL by default signals a full warning and retains as
+much of the package state as possible.
+
+This can be adjusted using @code{sb-ext:*on-package-variance*}:
+
+@include var-sb-ext-star-on-package-variance-star.texinfo
+
@node Garbage Collection
@comment node-name, next, previous, up
@section Garbage Collection
+@cindex Garbage collection
SBCL provides additional garbage collection functionality not
-specified by ANSI. Weak pointers allow references to objects to be
-maintained without keeping them from being garbage collected, and
-``finalization'' hooks are available to cause code to be executed when
-an object has been garbage collected. Additionally users can specify
-their own cleanup actions to be executed with garbage collection.
+specified by ANSI.
+
+@include var-sb-ext-star-after-gc-hooks-star.texinfo
+@include fun-sb-ext-gc.texinfo
+
+@subsection Finalization
+@cindex Finalization
+
+Finalization allows code to be executed after an object has been
+garbage collected. This is useful for example for releasing foreign
+memory associated with a Lisp object.
@include fun-sb-ext-finalize.texinfo
@include fun-sb-ext-cancel-finalization.texinfo
+
+@subsection Weak Pointers
+@cindex Weak pointers
+
+Weak pointers allow references to objects to be maintained without
+keeping them from being garbage collected: useful for building caches
+among other things.
+
+Hash tables can also have weak keys and values: @pxref{Hash Table
+Extensions}.
+
@include fun-sb-ext-make-weak-pointer.texinfo
@include fun-sb-ext-weak-pointer-value.texinfo
-@include var-sb-ext-star-after-gc-hooks-star.texinfo
+
+@subsection Introspection and Tuning
+
+@include var-sb-ext-star-gc-run-time-star.texinfo
+@include fun-sb-ext-bytes-consed-between-gcs.texinfo
+@include fun-sb-ext-dynamic-space-size.texinfo
+@include fun-sb-ext-get-bytes-consed.texinfo
+@include fun-sb-ext-gc-logfile.texinfo
+@include fun-sb-ext-generation-average-age.texinfo
+@include fun-sb-ext-generation-bytes-allocated.texinfo
+@include fun-sb-ext-generation-bytes-consed-between-gcs.texinfo
+@include fun-sb-ext-generation-minimum-age-before-gc.texinfo
+@include fun-sb-ext-generation-number-of-gcs-before-promotion.texinfo
+@include fun-sb-ext-generation-number-of-gcs.texinfo
@node Metaobject Protocol
@comment node-name, next, previous, up
@section Metaobject Protocol
+@subsection AMOP Compatibility of Metaobject Protocol
+
SBCL supports a metaobject protocol which is intended to be compatible
with AMOP; present exceptions to this (as distinct from current bugs)
are:
@itemize
-
+
@item
-@findex compute-effective-method
-@findex sb-mop:compute-effective-method
+@findex @sbmop{compute-effective-method}
@code{compute-effective-method} only returns one value, not two.
There is no record of what the second return value was meant to
indicate, and apparently no clients for it.
-
+
+@item
+@tindex @cl{generic-function}
+@tindex @cl{standard-generic-function}
+@tindex @sbmop{funcallable-standard-object}
+@tindex @cl{standard-object}
+@tindex @cl{function}
+The direct superclasses of @code{sb-mop:funcallable-standard-object} are
+@code{(function standard-object)}, not @code{(standard-object function)}.
+
+This is to ensure that the @code{standard-object} class is the last of
+the standardized classes before @code{t} appearing in the class
+precedence list of @code{generic-function} and
+@code{standard-generic-function}, as required by section 1.4.4.5 of the
+ANSI specification.
+
@item
-@findex ensure-generic-function
-@findex generic-function-declarations
-@findex sb-mop:generic-function-declarations
+@findex @cl{ensure-generic-function}
+@findex @sbmop{generic-function-declarations}
the arguments @code{:declare} and @code{:declarations} to
@code{ensure-generic-function} are both accepted, with the leftmost
argument defining the declarations to be stored and returned by
@code{:declare}. Portable code should use @code{:declare}.
@item
-@findex validate-superclass
-@findex finalize-inheritance
-@findex sb-mop:validate-superclass
-@findex sb-mop:finalize-inheritance
-@tindex standard-class
-@tindex funcallable-standard-class
-@tindex sb-mop:funcallable-standard-class
-@tindex function
-@findex sb-mop:class-prototype
-@findex class-prototype
-although SBCL obeys the requirement in AMOP for
-@code{validate-superclass} for @code{standard-class} and
-@code{funcallable-standard-class} to be compatible metaclasses, we
+@findex @sbmop{validate-superclass}
+@findex @sbmop{finalize-inheritance}
+@tindex @cl{standard-class}
+@tindex @sbmop{funcallable-standard-class}
+@tindex @cl{function}
+@findex @sbmop{class-prototype}
+although SBCL obeys the requirement in AMOP that
+@code{validate-superclass} should treat @code{standard-class} and
+@code{funcallable-standard-class} as compatible metaclasses, we
impose an additional requirement at class finalization time: a class
of metaclass @code{funcallable-standard-class} must have
@code{function} in its superclasses, and a class of metaclass
@code{standard-class} must not.
-@findex typep
-@findex class-of
-@findex subtypep
-At class finalization, a class prototype which is accessible by a
-standard mop function @code{sb-mop:class-prototype}. The user can
-then ask whether this object is a @code{function} or not in several
-different ways: whether it is a function according to @code{typep};
-whether its @code{class-of} is @code{subtypep} @code{function}, or
-whether @code{function} appears in the superclasses of the class. The
-additional consistency requirement comes from the desire to make all
-of these answers the same.
+@findex @cl{typep}
+@findex @cl{class-of}
+@findex @cl{subtypep}
+After a class has been finalized, it is associated with a class
+prototype which is accessible by a standard mop function
+@code{sb-mop:class-prototype}. The user can then ask whether this
+object is a @code{function} or not in several different ways: whether it
+is a function according to @code{typep}; whether its @code{class-of} is
+@code{subtypep} @code{function}, or whether @code{function} appears in
+the superclasses of the class. The additional consistency requirement
+comes from the desire to make all of these answers the same.
The following class definitions are bad, and will lead to errors
either immediately or if an instance is created:
@end lisp
and leads to a class whose instances are funcallable and have one slot.
+@tindex @sbmop{funcallable-standard-object}
+Note that this requirement also applies to the class
+@code{sb-mop:funcallable-standard-object}, which has metaclass
+@code{sb-mop:funcallable-standard-class} rather than
+@code{standard-class} as AMOP specifies.
+
@item
the requirement that ``No portable class @math{C_p} may inherit, by
virtue of being a direct or indirect subclass of a specified class, any
of software from colliding with each other.
@item
-@findex slot-value-using-class
-@findex sb-mop:slot-value-using-class
-@findex (setf slot-value-using-class)
-@findex (setf sb-mop:slot-value-using-class)
-@findex slot-boundp-using-class
-@findex sb-mop:slot-boundp-using-class
+@findex @sbmop{slot-value-using-class}
+@findex @setf{@sbmop{slot-value-using-class}}
+@findex @sbmop{slot-boundp-using-class}
specializations of the @code{new-value} argument to @code{(setf
sb-mop:slot-value-using-class)} are not allowed: all user-defined
methods must have a specializer of the class @code{t}.
Note that application code may specialize on the @code{new-value}
argument of slot accessors.
+@item
+@findex @cl{defclass}
+@findex @sbmop{ensure-class}
+@findex @sbmop{ensure-class-using-class}
+@findex @cl{find-class}
+@findex @cl{class-name}
+the class named by the @code{name} argument to @code{ensure-class}, if
+any, is only redefined if it is the proper name of that class;
+otherwise, a new class is created.
+
+This is consistent with the description of @code{ensure-class} in AMOP
+as the functional version of @code{defclass}, which has this behaviour;
+however, it is not consistent with the weaker requirement in AMOP, which
+states that any class found by @code{find-class}, no matter what its
+@code{class-name}, is redefined.
+
+@end itemize
+
+@subsection Metaobject Protocol Extensions
+
+In addition, SBCL supports extensions to the Metaobject protocol from
+AMOP; at present, they are:
+
+@itemize
+
+@item
+@findex @cl{defmethod}
+@findex @cl{find-class}
+@findex @sbmop{intern-eql-specializer}
+@findex @sbpcl{make-method-specializers-form}
+@findex @sbmop{make-method-lambda}
+compile-time support for generating specializer metaobjects from
+specializer names in @code{defmethod} forms is provided by the
+@code{make-method-specializers-form} function, which returns a form
+which, when evaluated in the lexical environment of the
+@code{defmethod}, returns a list of specializer metaobjects. This
+operator suffers from similar restrictions to those affecting
+@code{make-method-lambda}, namely that the generic function must be
+defined when the @code{defmethod} form is expanded, so that the
+correct method of @code{make-method-specializers-form} is invoked.
+The system-provided method on @code{make-method-specializers-form}
+generates a call to @code{find-class} for each symbol specializer
+name, and a call to @code{intern-eql-specializer} for each @code{(eql
+@var{x})} specializer name.
+
+@item
+@findex @cl{find-method}
+@findex @sbpcl{parse-specializer-using-class}
+@findex @sbpcl{unparse-specializer-using-class}
+run-time support for converting between specializer names and
+specializer metaobjects, mostly for the purposes of
+@code{find-method}, is provided by
+@code{parse-specializer-using-class} and
+@code{unparse-specializer-using-class}, which dispatch on their first
+argument, the generic function associated with a method with the given
+specializer. The system-provided methods on those methods convert
+between classes and proper names and between lists of the form
+@code{(eql @var{x})} and interned eql specializer objects.
+
+@item
+@vindex @sbpcl{+slot-unbound+}
+@findex @sbmop{standard-instance-access}
+@findex @sbmop{funcallable-standard-instance-access}
+distinguishing unbound instance allocated slots from bound ones when
+using @code{standard-instance-access} and
+@code{funcallable-standard-instance-access} is possible by comparison
+to the constant @code{+slot-unbound+}.
+
+@end itemize
+
+@node Extensible Sequences
+@comment node-name, next, previous, up
+@section Extensible Sequences
+
+@menu
+* Iterator Protocol::
+* Simple Iterator Protocol::
+@end menu
+
+ANSI Common Lisp has a class @code{sequence} with subclasses @code{list} and
+@code{vector} on which the ``sequence functions'' like @code{find},
+@code{subseq}, etc. operate. As an extension to the ANSI specification,
+SBCL allows additional subclasses of @code{sequence} to be defined
+@footnote{A motivation, rationale and additional examples for the design
+of this extension can be found in the paper @cite{Rhodes, Christophe
+(2007): User-extensible sequences in Common Lisp} available for download
+at
+@url{http://www.doc.gold.ac.uk/~mas01cr/papers/ilc2007/sequences-20070301.pdf}.}.
+@tindex @cl{sequence}
+@tindex @cl{vector}
+@findex @cl{find}
+@findex @cl{subseq}
+
+Users of this extension just make instances of @cl{sequence} subclasses
+and transparently operate on them using sequence functions:
+@lisp
+(coerce (subseq (make-instance 'my-sequence) 5 10) 'list)
+@end lisp
+From this perspective, no distinction between builtin and user-defined
+@code{sequence} subclasses should be necessary.
+@findex @cl{coerce}
+@findex @cl{subseq}
+@findex @cl{make-instance}
+@tindex @cl{list}
+
+Providers of the extension, that is of user-defined @code{sequence}
+subclasses, have to adhere to a ``sequence protocol'' which consists of
+a set of generic functions in the @code{sequence} package.
+@c
+A minimal @code{sequence} subclass has to specify @code{standard-object} and
+@code{sequence} as its superclasses and has to be the specializer of the
+@code{sequence} parameter of methods on at least the following generic
+functions:
+@tindex @cl{sequence}
+@tindex @cl{standard-object}
+
+@include fun-sb-sequence-length.texinfo
+@include fun-sb-sequence-elt.texinfo
+@include fun-sb-sequence-setf-elt.texinfo
+@include fun-sb-sequence-adjust-sequence.texinfo
+@include fun-sb-sequence-make-sequence-like.texinfo
+
+@code{make-sequence-like} is needed for functions returning
+freshly-allocated sequences such as @code{subseq} or
+@code{copy-seq}. @code{adjust-sequence} is needed for functions which
+destructively modify their arguments such as @code{delete}. In fact, all
+other sequence functions can be implemented in terms of the above
+functions and actually are, if no additional methods are
+defined. However, relying on these generic implementations, in
+particular not implementing the iterator protocol can incur a high
+performance penalty @xref{Iterator Protocol}.
+@tindex @cl{sequence}
+@findex @sequence{make-sequence-like}
+@findex @cl{subseq}
+@findex @cl{copy-seq}
+@findex @sequence{adjust-sequence}
+
+In addition to the mandatory functions above, methods on the following
+sequence functions can be defined:
+
+@include fun-sb-sequence-emptyp.texinfo
+
+@itemize
+@item
+@code{sb-sequence:count}, @code{sb-sequence:count-if}, @code{sb-sequence:count-if-not}
+
+@item
+@code{sb-sequence:find}, @code{sb-sequence:find-if}, @code{sb-sequence:find-if-not}
+
+@item
+@code{sb-sequence:position}, @code{sb-sequence:position-if}, @code{sb-sequence:position-if-not}
+
+@item
+@code{sb-sequence:subseq}
+
+@item
+@code{sb-sequence:copy-seq}
+
+@item
+@code{sb-sequence:fill}
+
+@item
+@code{sb-sequence:nsubstitute}, @code{sb-sequence:nsubstitute-if},
+@code{sb-sequence:nsubstitute-if-not}, @code{sb-sequence:substitute},
+@code{sb-sequence:substitute-if}, @code{sb-sequence:substitute-if-not}
+
+@item
+@code{sb-sequence:replace}
+
+@item
+@code{sb-sequence:nreverse}, @code{sb-sequence:reverse}
+
+@item
+@code{sb-sequence:reduce}
+
+@item
+@code{sb-sequence:mismatch}
+
+@item
+@code{sb-sequence:search}
+
+@item
+@code{sb-sequence:delete}, @code{sb-sequence:delete-if}, @code{sb-sequence:delete-if-not},
+@code{sb-sequence:remove}, @code{sb-sequence:remove-if}, @code{sb-sequence:remove-if-not},
+
+@item
+@code{sb-sequence:delete-duplicates}, @code{sb-sequence:remove-duplicates}
+
+@item
+@code{sb-sequence:sort}, @code{sb-sequence:stable-sort}
@end itemize
+In the spirit of @code{dolist}, generic sequences can be traversed using
+the macro
+@findex @cl{dolist}
+
+@include macro-sb-sequence-dosequence.texinfo
+
+@node Iterator Protocol
+@comment node-name, next, previous, up
+@subsection Iterator Protocol
+
+The iterator protocol allows subsequently accessing some or all elements
+of a sequence in forward or reverse direction. Users first call
+@code{make-sequence-iterator} to create an iteration state and
+receive functions to query and mutate it. These functions allow, among
+other things, moving to, retrieving or modifying elements of the
+sequence. An iteration state consists of a state object, a limit object,
+a from-end indicator and the following six functions to query or mutate
+this state:
+@findex @sequence{make-sequence-iterator}
+@deffn {Function} @code{step function} sequence iterator from-end
+Moves the iterator one position forward or backward in the associated
+sequence depending on the iteration direction.
+@end deffn
+@deffn {Function} @code{endp function} sequence iterator limit from-end
+Returns non-@code{nil} when the iterator has reached the end of the
+associated sequence with respect to the iteration direction.
+@end deffn
+@deffn {Function} @code{element function} sequence iterator
+Returns the sequence element associated to the current position of the
+iteration.
+@end deffn
+@deffn {Function} @code{setf element function} new-value sequence iterator
+Destructively modifies the associates sequence by replacing the sequence
+element associated to the current iteration position with a new value.
+@end deffn
+@deffn {Function} @code{index function} sequence iterator
+Returns the position of the iteration in the associated sequence.
+@end deffn
+@deffn {Function} @code{copy function} sequence iterator
+Returns a copy of the iteration state which can be mutated independently
+of the copied iteration state.
+@end deffn
+
+An iterator is created by calling:
+
+@include fun-sb-sequence-make-sequence-iterator.texinfo
+
+Note that @code{make-sequence-iterator} calls
+@code{make-simple-sequence-iterator} when there is no specialized
+method for a particular @code{sequence} subclass. @xref{Simple Iterator
+Protocol}.
+@findex @sequence{make-sequence-iterator}
+@findex @sequence{make-simple-sequence-iterator}
+@tindex @cl{sequence}
+
+The following convenience macros simplify traversing sequences using
+iterators:
+
+@include macro-sb-sequence-with-sequence-iterator.texinfo
+@include macro-sb-sequence-with-sequence-iterator-functions.texinfo
+
+@node Simple Iterator Protocol
+@comment node-name, next, previous, up
+@subsection Simple Iterator Protocol
+
+For cases in which the full flexibility and performance of the general
+sequence iterator protocol is not required, there is a simplified
+sequence iterator protocol consisting of a few generic functions which
+can be specialized for iterator classes:
+
+@include fun-sb-sequence-iterator-step.texinfo
+@include fun-sb-sequence-iterator-endp.texinfo
+@include fun-sb-sequence-iterator-element.texinfo
+@include fun-sb-sequence-setf-iterator-element.texinfo
+@include fun-sb-sequence-iterator-index.texinfo
+@include fun-sb-sequence-iterator-copy.texinfo
+
+Iterator objects implementing the above simple iteration protocol are
+created by calling the following generic function:
+
+@include fun-sb-sequence-make-simple-sequence-iterator.texinfo
+
@node Support For Unix
@comment node-name, next, previous, up
@section Support For Unix
+@menu
+* Command-line arguments::
+* Querying the process environment::
+* Running external programs::
+@end menu
+
+@node Command-line arguments
+@subsection Command-line arguments
+@vindex @sbext{@earmuffs{posix-argv}}
+
The UNIX command line can be read from the variable
-@code{sb-ext:*posix-argv*}. The UNIX environment can be queried with
-the @code{sb-ext:posix-getenv} function.
+@code{sb-ext:*posix-argv*}.
+
+@node Querying the process environment
+@subsection Querying the process environment
+
+The UNIX environment can be queried with the
+@code{sb-ext:posix-getenv} function.
@include fun-sb-ext-posix-getenv.texinfo
+@node Running external programs
+@subsection Running external programs
+
External programs can be run with @code{sb-ext:run-program}.
+@footnote{In SBCL versions prior to 1.0.13, @code{sb-ext:run-program}
+searched for executables in a manner somewhat incompatible with other
+languages. As of this version, SBCL uses the system library routine
+@code{execvp(3)}, and no longer contains the function,
+@code{find-executable-in-search-path}, which implemented the old
+search. Users who need this function may find it
+in @file{run-program.lisp} versions 1.67 and earlier in SBCL's CVS
+repository here
+@url{http://sbcl.cvs.sourceforge.net/sbcl/sbcl/src/code/run-program.lisp?view=log}. However,
+we caution such users that this search routine finds executables that
+system library routines do not.}
@include fun-sb-ext-run-program.texinfo
+When @code{sb-ext:run-program} is called with @code{wait} equal to
+NIL, an instance of class @var{sb-ext:process} is returned. The
+following functions are available for use with processes:
+
@include fun-sb-ext-process-p.texinfo
@include fun-sb-ext-process-input.texinfo
@include fun-common-lisp-ed.texinfo
@include var-sb-ext-star-ed-functions-star.texinfo
+Conditions of type @code{warning} and @code{style-warning} are
+sometimes signaled at runtime, especially during execution of Common
+Lisp defining forms such as @code{defun}, @code{defmethod}, etc. To
+muffle these warnings at runtime, SBCL provides a variable
+@code{sb-ext:*muffled-warnings*}:
+
+@include var-sb-ext-star-muffled-warnings-star.texinfo
+
@node Tools To Help Developers
@comment node-name, next, previous, up
@section Tools To Help Developers
+@findex @cl{trace}
+@findex @cl{inspect}
SBCL provides a profiler and other extensions to the ANSI @code{trace}
facility. For more information, see @ref{Macro common-lisp:trace}.
@node Resolution of Name Conflicts
@section Resolution of Name Conflicts
+@tindex @sbext{name-conflict}
+@findex @sbext{name-conflict-symbols}
The ANSI standard (section 11.1.1.2.5) requires that name conflicts in
packages be resolvable in favour of any of the conflicting symbols. In
argument, which should be a member of the list returned by the condition
accessor @code{sb-ext:name-conflict-symbols}.
+@node Hash Table Extensions
+@comment node-name, next, previous, up
+@section Hash Table Extensions
+@cindex Hash tables
+
+Hash table extensions supported by SBCL are all controlled by keyword
+arguments to @code{make-hash-table}.
+
+@include fun-common-lisp-make-hash-table.texinfo
+
+@include macro-sb-ext-define-hash-table-test.texinfo
+
+@include macro-sb-ext-with-locked-hash-table.texinfo
+
+@include fun-sb-ext-hash-table-synchronized-p.texinfo
+
+@include fun-sb-ext-hash-table-weakness.texinfo
+
+@node Random Number Generation
+@comment node-name, next, previous, up
+@section Random Number Generation
+@cindex Random Number Generation
+
+The initial value of @code{*random-state*} is the same each time SBCL
+is started. This makes it possible for user code to obtain repeatable
+pseudo random numbers using only standard-provided functionality. See
+@code{seed-random-state} below for an SBCL extension that allows to
+seed the random number generator from given data for an additional
+possibility to achieve this. Non-repeatable random numbers can always
+be obtained using @code{(make-random-state t)}.
+
+The sequence of numbers produced by repeated calls to @code{random}
+starting with the same random state and using the same sequence of
+@code{limit} arguments is guaranteed to be reproducible only in the
+same version of SBCL on the same platform, using the same code under
+the same evaluator mode and compiler optimization qualities. Just two
+examples of differences that may occur otherwise: calls to
+@code{random} can be compiled differently depending on how much is
+known about the @code{limit} argument at compile time, yielding
+different results even if called with the same argument at run time,
+and the results can differ depending on the machine's word size, for
+example for limits that are fixnums under 64-bit word size but bignums
+under 32-bit word size.
+
+@include fun-sb-ext-seed-random-state.texinfo
+
+Some notes on random floats: The standard doesn't prescribe a specific
+method of generating random floats. The following paragraph describes
+SBCL's current implementation and should be taken purely informational,
+that is, user code should not depend on any of its specific properties.
+The method used has been chosen because it is common, conceptually
+simple and fast.
+
+To generate random floats, SBCL evaluates code that has an equivalent
+effect as
+@lisp
+(* limit
+ (float (/ (random (expt 2 23)) (expt 2 23)) 1.0f0))
+@end lisp
+(for single-floats) and correspondingly (with @code{52} and
+@code{1.0d0} instead of @code{23} and @code{1.0f0}) for double-floats.
+Note especially that this means that zero is a possible return value
+occurring with probability @code{(expt 2 -23)} respectively
+@code{(expt 2 -52)}. Also note that there exist twice as many
+equidistant floats between 0 and 1 as are generated. For example, the
+largest number that @code{(random 1.0f0)} ever returns is
+@code{(float (/ (1- (expt 2 23)) (expt 2 23)) 1.0f0)} while
+@code{(float (/ (1- (expt 2 24)) (expt 2 24)) 1.0f0)} is the
+largest single-float less than 1. This is a side effect of the fact
+that the implementation uses the fastest possible conversion from bits
+to floats.
+
+SBCL currently uses the Mersenne Twister as its random number
+generator, specifically the 32-bit version under both 32- and 64-bit
+word size. The seeding algorithm has been improved several times by
+the authors of the Mersenne Twister; SBCL uses the third version
+(from 2002) which is still the most recent as of June 2012. The
+implementation has been tested to provide output identical to the
+recommended C implementation.
+
+While the Mersenne Twister generates random numbers of much better
+statistical quality than other widely used generators, it uses only
+linear operations modulo 2 and thus fails some statistical
+tests@footnote{See chapter 7 "Testing widely used RNGs" in
+@cite{TestU01: A C Library for Empirical Testing of Random Number
+Generators} by Pierre L'Ecuyer and Richard Simard, ACM Transactions on
+Mathematical Software, Vol. 33, article 22, 2007.}.
+For example, the distribution of ranks of (sufficiently large) random
+binary matrices is much distorted compared to the theoretically
+expected one when the matrices are generated by the Mersenne Twister.
+Thus, applications that are sensitive to this aspect should use a
+different type of generator.
+
+@node Miscellaneous Extensions
+@comment node-name, next, previous, up
+@section Miscellaneous Extensions
+
+@include fun-sb-ext-array-storage-vector.texinfo
+@include fun-sb-ext-delete-directory.texinfo
+@include fun-sb-ext-get-time-of-day.texinfo
+@include macro-sb-ext-wait-for.texinfo
+@include fun-sb-ext-version-assert.texinfo
+
@node Stale Extensions
@comment node-name, next, previous, up
@section Stale Extensions
The @code{sb-ext:freeze-type} declaration declares that a
type will never change, which can make type testing
(@code{typep}, etc.) more efficient for structure types.
-
-The @code{sb-ext:constant-function} declaration specifies
-that a function will always return the same value for the same
-arguments, which may allow the compiler to optimize calls
-to it. This is appropriate for functions like @code{sqrt}, but
-is @emph{not} appropriate for functions like @code{aref},
-which can change their return values when the underlying data are
-changed.
-@c <!-- FIXME: This declaration does not seem to be supported in the
-@c current compiler. -->
projects / sbcl.git / blobdiff