factor/library/collections/hashtables.factor

288 lines
8.0 KiB
Factor

IN: hashtables-internals
USING: arrays hashtables kernel math sequences
sequences-internals ;
! This hashtable implementation uses only one auxilliary array
! in addition to the hashtable tuple itself. The array stores
! keys in even slots and values in odd slots. Values are looked
! up with a hashing strategy that uses linear probing to resolve
! collisions.
! There are two special objects: the ((tombstone)) marker and
! the ((empty)) marker. Neither of these markers can be used as
! hashtable keys.
! hash-count is the number of entries including deleted entries.
! hash-deleted is the number of deleted entries.
TUPLE: tombstone ;
: ((empty)) T{ tombstone f } ; inline
: ((tombstone)) T{ tombstone t } ; inline
: hash@ ( key keys -- n )
#! Return an even key index.
>r hashcode r> length 2 /i rem 2 * ;
: probe ( heys i -- hash i ) 2 + over length mod ;
: (key@) ( key keys i -- n )
3dup swap nth-unsafe {
{ [ dup ((tombstone)) eq? ] [ 2drop probe (key@) ] }
{ [ dup ((empty)) eq? ] [ 2drop 3drop -1 ] }
{ [ = ] [ 2nip ] }
{ [ t ] [ probe (key@) ] }
} cond ;
: key@ ( key hash -- n )
underlying 2dup hash@ (key@) ;
: if-key ( key hash true false -- | true: index key hash -- )
>r >r [ key@ ] 2keep pick -1 > r> r> if ; inline
: <hash-array> ( n -- array )
1+ 4 * ((empty)) <repeated> >array ;
: reset-hash ( n hash -- )
swap <hash-array> over set-underlying
0 over set-hash-count 0 swap set-hash-deleted ;
: (new-key@) ( key keys i -- n )
3dup swap nth-unsafe dup tombstone? [
2drop 2nip
] [
= [ 2nip ] [ probe (new-key@) ] if
] if ;
: new-key@ ( key hash -- n )
underlying 2dup hash@ (new-key@) ;
: nth-pair ( n seq -- key value )
[ nth-unsafe ] 2keep >r 1+ r> nth-unsafe ;
: set-nth-pair ( value key n seq -- )
[ set-nth-unsafe ] 2keep >r 1+ r> set-nth-unsafe ;
: hash-count+ dup hash-count 1+ swap set-hash-count ;
: hash-deleted+ dup hash-deleted 1+ swap set-hash-deleted ;
: hash-deleted- dup hash-deleted 1- swap set-hash-deleted ;
: change-size ( hash old -- )
dup ((tombstone)) eq? [
drop hash-deleted-
] [
((empty)) eq? [ hash-count+ ] [ drop ] if
] if ;
: (set-hash) ( value key hash -- )
#! Store a value without growing the hashtable. Internal.
2dup new-key@ swap
[ underlying 2dup nth-unsafe ] keep
( value key n underlying old hash )
swap change-size set-nth-pair ;
: (each-pair) ( quot array i -- | quot: k v -- )
over length over number= [
3drop
] [
[
swap nth-pair over tombstone?
[ 3drop ] [ rot call ] if
] 3keep 2 + (each-pair)
] if ; inline
: each-pair ( array quot -- | quot: k v -- )
swap 0 (each-pair) ; inline
: (all-pairs?) ( quot array i -- ? | quot: k v -- ? )
over length over number= [
3drop t
] [
3dup >r >r >r swap nth-pair over tombstone? [
3drop r> r> r> 2 + (all-pairs?)
] [
rot call
[ r> r> r> 2 + (all-pairs?) ] [ r> r> r> 3drop f ] if
] if
] if ; inline
: all-pairs? ( array quot -- ? | quot: k v -- ? )
swap 0 (all-pairs?) ; inline
: hash>seq ( i hash -- seq )
underlying dup length 2 /i
[ 2 * pick + over nth-unsafe ] map
[ tombstone? not ] subset 2nip ;
IN: hashtables
: <hashtable> ( capacity -- hashtable )
(hashtable) [ reset-hash ] keep ;
: hash* ( key hash -- value ? )
[
nip >r 1+ r> underlying nth-unsafe t
] [
3drop f f
] if-key ;
: hash-contains? ( key hash -- ? )
[ 3drop t ] [ 3drop f ] if-key ;
: ?hash* ( key hash -- value/f ? )
dup [ hash* ] [ 2drop f f ] if ;
: hash ( key hash -- value ) hash* drop ; inline
: ?hash ( key hash -- value )
dup [ hash ] [ 2drop f ] if ;
: clear-hash ( hash -- )
[ underlying length ] keep reset-hash ;
: remove-hash ( key hash -- )
[
nip
dup hash-deleted+
underlying >r >r ((tombstone)) dup r> r> set-nth-pair
] [
3drop
] if-key ;
: hash-size ( hash -- n ) dup hash-count swap hash-deleted - ;
: hash-empty? ( hash -- ? ) hash-size 0 = ;
: grow-hash ( hash -- )
[ dup underlying swap hash-size 1+ ] keep
[ reset-hash ] keep swap [ swap pick (set-hash) ] each-pair
drop ;
: ?grow-hash ( hash -- )
dup hash-count 3 * over underlying length >
[ dup grow-hash ] when drop ;
: set-hash ( value key hash -- )
[ (set-hash) ] keep ?grow-hash ;
: associate ( value key -- hashtable )
2 <hashtable> [ set-hash ] keep ;
: hash-keys ( hash -- keys ) 0 swap hash>seq ;
: hash-values ( hash -- keys ) 1 swap hash>seq ;
: hash>alist ( hash -- assoc )
dup hash-keys swap hash-values 2array flip ;
: alist>hash ( alist -- hash )
[ length <hashtable> ] keep
[ first2 swap pick (set-hash) ] each ;
: hash-each ( hash quot -- | quot: k v -- )
#! Apply a quotation to each key/value pair.
>r underlying r> each-pair ; inline
: hash-each-with ( obj hash quot -- | quot: obj k v -- )
swap [ 2swap [ >r -rot r> call ] 2keep ] hash-each 2drop ;
inline
: hash-all? ( hash quot -- | quot: k v -- ? )
#! Tests if every key/value pair satisfies the predicate.
>r underlying r> all-pairs? ; inline
: hash-all-with? ( obj hash quot -- | quot: obj k v -- ? )
swap
[ 2swap [ >r -rot r> call ] 2keep rot ] hash-all? 2nip ;
inline
: subhash? ( h1 h2 -- ? )
#! Test if h2 contains all the key/value pairs of h1.
swap [
>r swap hash* [ r> = ] [ r> 2drop f ] if
] hash-all-with? ; flushable
: hash-subset ( hash quot -- hash | quot: k v -- ? )
#! Make a new hash that only includes the key/value pairs
#! which satisfy the predicate.
over hash-size <hashtable> rot [
2swap [
>r pick pick >r >r call [
r> r> swap r> set-hash
] [
r> r> r> 3drop
] if
] 2keep
] hash-each nip ; inline
: hash-subset-with ( obj hash quot -- hash | quot: obj { k v } -- ? )
swap
[ 2swap [ >r -rot r> call ] 2keep rot ] hash-subset 2nip ;
inline
M: hashtable clone ( hash -- hash ) clone-growable ;
: hashtable= ( hash hash -- ? )
2dup subhash? >r swap subhash? r> and ;
M: hashtable = ( obj hash -- ? )
{
{ [ 2dup eq? ] [ 2drop t ] }
{ [ over hashtable? not ] [ 2drop f ] }
{ [ 2dup [ hash-size ] 2apply number= not ] [ 2drop f ] }
{ [ t ] [ hashtable= ] }
} cond ;
: ?hash ( key hash/f -- value/f )
dup [ hash ] [ 2drop f ] if ; flushable
: ?hash* ( key hash/f -- value/f )
dup [ hash* ] [ 2drop f f ] if ; flushable
: ?set-hash ( value key hash/f -- hash )
[ 2 <hashtable> ] unless* [ set-hash ] keep ;
: hash-stack ( key seq -- value )
#! Searches for a key in a sequence of hashtables,
#! where the most recently pushed hashtable is searched
#! first.
[ dupd hash-contains? ] find-last nip ?hash ; flushable
: hash-intersect ( hash1 hash2 -- hash1/\hash2 )
#! Remove all keys from hash2 not in hash1.
[ drop swap hash ] hash-subset-with ;
: hash-diff ( hash1 hash2 -- hash2-hash1 )
#! Remove all keys from hash2 in hash1.
[ drop swap hash not ] hash-subset-with ;
: hash-update ( hash1 hash2 -- )
#! Add all key/value pairs from hash2 to hash1.
[ swap rot set-hash ] hash-each-with ;
: hash-union ( hash1 hash2 -- hash1\/hash2 )
#! Make a new hashtable with all key/value pairs from
#! hash1 and hash2. Values in hash2 take precedence.
>r clone dup r> hash-update ;
: remove-all ( hash seq -- seq )
#! Remove all elements from the sequence that are keys
#! in the hashtable.
[ swap hash-contains? not ] subset-with ; flushable
: cache ( key hash quot -- value | quot: key -- value )
pick pick hash [
>r 3drop r>
] [
pick rot >r >r call dup r> r> set-hash
] if* ; inline
: map>hash ( seq quot -- hash | quot: key -- value )
#! Construct a hashtable with keys from the sequence, and
#! values obtained by applying the quotation to each key.
swap [ length <hashtable> ] keep
[ -rot [ >r over >r call r> r> set-hash ] 2keep ] each nip ;
inline