racket/pkgs/racket-doc/syntax/scribblings/id-table.scrbl
Georges Dupéron 153dc01ccd Some small fixes to the documentation
* Wrong contract for syntax-local-value in the documentation.
* Clarified signature in documentation for expand-import, expand-export and pre-expand-export
* Corrected typo in documentation for "for".
* Fixed error message for function which seems to have been renamed in the docs
* Fixed typo in a comment in the tests
* Fixed a typo in the documentation for set-subtract.
* Use double ellipses for the free-id-table-set*, free-id-table-set*!, bound-id-table-set* and bound-id-table-set*! operations
2016-10-19 20:52:45 -05:00

392 lines
14 KiB
Racket

#lang scribble/doc
@(require "common.rkt"
scribble/eval
(for-label syntax/id-table racket/dict))
@(define id-table-eval (make-base-eval))
@(id-table-eval '(require (for-syntax racket/base syntax/id-table)))
@title[#:tag "idtable"]{Dictionaries with Identifier Keys}
@defmodule[syntax/id-table]
This module provides two implementations of @deftech{identifier tables}:
dictionaries with identifier keys that use identifier-specific
comparisons instead of @racket[eq?] or @racket[equal?]. Identifier
tables implement the @racketmodname[racket/dict] interface, and they
are available in both mutable and immutable variants.
@section{Dictionaries for @racket[free-identifier=?]}
A free-identifier table is a dictionary whose keys are compared using
@racket[free-identifier=?]. Free-identifier tables implement the
dictionary interface of @racketmodname[racket/dict], so all of the
appropriate generic functions (@racket[dict-ref], @racket[dict-map],
etc) can be used on free-identifier tables.
A caveat for using these tables is that a lookup can fail with
unexpected results if the binding of an identifier changes between
key-value insertion and the lookup.
For example, consider the following use:
@interaction[#:eval id-table-eval
(define-syntax-rule (m)
(begin
(begin-for-syntax
(define table (make-free-id-table))
(code:comment "set table entry to #t")
(free-id-table-set! table #'x #t)
(code:comment "sanity check, it's set to #t")
(displayln (free-id-table-ref table #'x #f)))
(define x 'defined-now)
(begin-for-syntax
(code:comment "might expect to get #t, but prints #f")
(displayln (free-id-table-ref table #'x #f)))))
(m)]
The macro @racket[m] expands to code that initializes an identifier table
at compile-time and inserts a key-value pair for @racket[#'x] and
@racket[#t]. The @racket[#'x] identifier has no binding, however, until
the definition @racket[(define x 'defined-now)] is evaluated.
As a result, the lookup at the end of @racket[m] will return @racket[#f]
instead of @racket[#t] because the binding symbol for @racket[#'x] changes
after the initial key-value pair is put into the table. If the definition
is evaluated @emph{before} the initial insertion, both expressions will
print @racket[#t].
@deftogether[[
@defproc[(make-free-id-table
[init-dict dict? null]
[#:phase phase (or/c exact-integer? #f) (syntax-local-phase-level)])
mutable-free-id-table?]
@defproc[(make-immutable-free-id-table
[init-dict dict? null]
[#:phase phase (or/c exact-integer? #f) (syntax-local-phase-level)])
immutable-free-id-table?]]]{
Produces a mutable free-identifier table or immutable free-identifier
table, respectively. The dictionary uses @racket[free-identifier=?]
to compare keys, but also uses a hash table based on symbol equality
to make the dictionary efficient in the common case.
The identifiers are compared at phase level @racket[phase]. The
default phase, @racket[(syntax-local-phase-level)], is generally
appropriate for identifier tables used by macros, but code that
analyzes fully-expanded programs may need to create separate
identifier tables for each phase of the module.
The optional @racket[init-dict] argument provides the initial
mappings. It must be a dictionary, and its keys must all be
identifiers. If the @racket[init-dict] dictionary has multiple
distinct entries whose keys are @racket[free-identifier=?], only one
of the entries appears in the new id-table, and it is not specified
which entry is picked.
}
@defproc[(free-id-table? [v any/c]) boolean?]{
Returns @racket[#t] if @racket[v] was produced by
@racket[make-free-id-table] or
@racket[make-immutable-free-id-table], @racket[#f] otherwise.
}
@defproc[(mutable-free-id-table? [v any/c]) boolean?]{
Returns @racket[#t] if @racket[v] was produced by
@racket[make-free-id-table], @racket[#f] otherwise.
}
@defproc[(immutable-free-id-table? [v any/c]) boolean?]{
Returns @racket[#t] if @racket[v] was produced by
@racket[make-immutable-free-id-table], @racket[#f] otherwise.
}
@defproc[(free-id-table-ref [table free-id-table?]
[id identifier?]
[failure any/c
(lambda () (raise (make-exn:fail .....)))])
any]{
Like @racket[hash-ref]. In particular, if @racket[id] is not found,
the @racket[failure] argument is applied if it is a procedure, or
simply returned otherwise.
}
@defproc[(free-id-table-ref! [table mutable-free-id-table?]
[id identifier?]
[failure any/c])
any]{
Like @racket[hash-ref!].
@history[#:added "6.3.0.6"]
}
@defproc[(free-id-table-set! [table mutable-free-id-table?]
[id identifier?]
[v any/c])
void?]{
Like @racket[hash-set!].
}
@defproc[(free-id-table-set [table immutable-free-id-table?]
[id identifier?]
[v any/c])
immutable-free-id-table?]{
Like @racket[hash-set].
}
@defproc[(free-id-table-set*! [table mutable-free-id-table?]
[id identifier?]
[v any/c] ... ...)
void?]{
Like @racket[hash-set*!].
@history[#:added "6.3.0.6"]
}
@defproc[(free-id-table-set* [table immutable-free-id-table?]
[id identifier?]
[v any/c] ... ...)
immutable-free-id-table?]{
Like @racket[hash-set*].
@history[#:added "6.3.0.6"]
}
@defproc[(free-id-table-remove! [table mutable-free-id-table?]
[id identifier?])
void?]{
Like @racket[hash-remove!].
}
@defproc[(free-id-table-remove [table immutable-free-id-table?]
[id identifier?])
immutable-free-id-table?]{
Like @racket[hash-remove].
}
@defproc[(free-id-table-update! [table mutable-free-id-table?]
[id identifier?]
[updater (any/c . -> . any/c)]
[failure any/c
(lambda () (raise (make-exn:fail .....)))])
void?]{
Like @racket[hash-update!].
@history[#:added "6.3.0.6"]
}
@defproc[(free-id-table-update [table immutable-free-id-table?]
[id identifier?]
[updater (any/c . -> . any/c)]
[failure any/c
(lambda () (raise (make-exn:fail .....)))])
immutable-free-id-table?]{
Like @racket[hash-update].
@history[#:added "6.3.0.6"]
}
@defproc[(free-id-table-map [table free-id-table?]
[proc (-> identifier? any/c any)])
list?]{
Like @racket[hash-map].
}
@defproc[(free-id-table-keys [table free-id-table?])
(listof identifier?)]{
Like @racket[hash-keys].
@history[#:added "6.3.0.3"]
}
@defproc[(free-id-table-values [table free-id-table?])
(listof any/c)]{
Like @racket[hash-values].
@history[#:added "6.3.0.3"]
}
@defproc[(in-free-id-table [table free-id-table?])
sequence?]{
Like @racket[in-hash].
@history[#:added "6.3.0.3"]
}
@defproc[(free-id-table-for-each [table free-id-table?]
[proc (-> identifier? any/c any)])
void?]{
Like @racket[hash-for-each].
}
@defproc[(free-id-table-count [table free-id-table?])
exact-nonnegative-integer?]{
Like @racket[hash-count].
}
@deftogether[[
@defproc[(free-id-table-iterate-first [table free-id-table?])
id-table-iter?]
@defproc[(free-id-table-iterate-next [table free-id-table?]
[position id-table-iter?])
id-table-iter?]
@defproc[(free-id-table-iterate-key [table free-id-table?]
[position id-table-iter?])
identifier?]
@defproc[(free-id-table-iterate-value [table bound-it-table?]
[position id-table-iter?])
identifier?]]]{
Like @racket[hash-iterate-first], @racket[hash-iterate-next],
@racket[hash-iterate-key], and @racket[hash-iterate-value],
respectively.
}
@defproc[(id-table-iter? [v any/c]) boolean?]{
Returns @racket[#t] if @racket[v] represents a position in an
identifier table (free or bound, mutable or immutable), @racket[#f]
otherwise.
}
@defproc[(free-id-table/c [key-ctc flat-contract?]
[val-ctc chaperone-contract?]
[#:immutable immutable? (or/c #t #f 'dont-care) 'dont-care])
contract?]{
Like @racket[hash/c], but for free-identifier tables. If
@racket[immutable?] is @racket[#t], the contract accepts only
immutable identifier tables; if @racket[immutable?] is @racket[#f],
the contract accepts only mutable identifier tables.
}
@;{----------}
@section{Dictionaries for @racket[bound-identifier=?]}
A bound-identifier table is a dictionary whose keys are compared using
@racket[bound-identifier=?]. Bound-identifier tables implement the
dictionary interface of @racketmodname[racket/dict], so all of the
appropriate generic functions (@racket[dict-ref], @racket[dict-map],
etc) can be used on bound-identifier tables.
@deftogether[[
@defproc[(make-bound-id-table
[init-dict dict? null]
[#:phase phase (or/c exact-integer? #f) (syntax-local-phase-level)])
mutable-bound-id-table?]
@defproc[(make-immutable-bound-id-table
[init-dict dict? null]
[#:phase phase (or/c exact-integer? #f) (syntax-local-phase-level)])
immutable-bound-id-table?]
@defproc[(bound-id-table? [v any/c]) boolean?]
@defproc[(mutable-bound-id-table? [v any/c]) boolean?]
@defproc[(immutable-bound-id-table? [v any/c]) boolean?]
@defproc[(bound-id-table-ref [table bound-id-table?]
[id identifier?]
[failure any/c
(lambda () (raise (make-exn:fail .....)))])
any]
@defproc[(bound-id-table-ref! [table mutable-bound-id-table?]
[id identifier?]
[failure any/c])
any]
@defproc[(bound-id-table-set! [table mutable-bound-id-table?]
[id identifier?]
[v any/c])
void?]
@defproc[(bound-id-table-set [table immutable-bound-id-table?]
[id identifier?]
[v any/c])
immutable-bound-id-table?]
@defproc[(bound-id-table-set*! [table mutable-bound-id-table?]
[id identifier?]
[v any/c] ... ...)
void?]
@defproc[(bound-id-table-set* [table immutable-bound-id-table?]
[id identifier?]
[v any/c] ... ...)
immutable-bound-id-table?]
@defproc[(bound-id-table-remove! [table mutable-bound-id-table?]
[id identifier?])
void?]
@defproc[(bound-id-table-remove [table immutable-bound-id-table?]
[id identifier?])
immutable-bound-id-table?]
@defproc[(bound-id-table-update! [table mutable-bound-id-table?]
[id identifier?]
[updater (any/c . -> . any/c)]
[failure any/c
(lambda () (raise (make-exn:fail .....)))])
void?]
@defproc[(bound-id-table-update [table immutable-bound-id-table?]
[id identifier?]
[updater (any/c . -> . any/c)]
[failure any/c
(lambda () (raise (make-exn:fail .....)))])
immutable-bound-id-table?]
@defproc[(bound-id-table-map [table bound-id-table?]
[proc (-> identifier? any/c any)])
list?]
@defproc[(bound-id-table-keys [table bound-id-table?])
(listof identifier?)]
@defproc[(bound-id-table-values [table bound-id-table?])
(listof any/c)]
@defproc[(in-bound-id-table [table bound-id-table?])
sequence?]
@defproc[(bound-id-table-for-each [table bound-id-table?]
[proc (-> identifier? any/c any)])
void?]
@defproc[(bound-id-table-count [table bound-id-table?])
exact-nonnegative-integer?]
@defproc[(bound-id-table-iterate-first [table bound-id-table?])
id-table-position?]
@defproc[(bound-id-table-iterate-next [table bound-id-table?]
[position id-table-position?])
id-table-position?]
@defproc[(bound-id-table-iterate-key [table bound-id-table?]
[position id-table-position?])
identifier?]
@defproc[(bound-id-table-iterate-value [table bound-id-table?]
[position id-table-position?])
identifier?]
@defproc[(bound-id-table/c [key-ctc flat-contract?]
[val-ctc chaperone-contract?]
[#:immutable immutable (or/c #t #f 'dont-care) 'dont-care])
contract?]
]]{
Like the procedures for free-identifier tables
(@racket[make-free-id-table], @racket[free-id-table-ref], etc), but
for bound-identifier tables, which use @racket[bound-identifier=?] to
compare keys.
@history[#:changed "6.3.0.3" "Added bound-id-table-keys, bound-id-table-values, in-bound-id-table."
#:changed "6.3.0.6"
@string-append{Added bound-id-table-ref!, bound-id-table-set*,
bound-id-table-set*!, bound-id-table-update!, and bound-id-table-update}]
}
@close-eval[id-table-eval]