Matthew Flatt
190925ee32
avoid a compiler warning
2016-01-02 07:46:42 -07:00
Robby Findler
2529e63b74
make stronger recognize any/c on the right as stronger than any flat contracts
2016-01-01 19:49:38 -06:00
Robby Findler
b24882fd18
implement the (-> any/c ... any) special case for the new -> contract combinator
...
(new is being used in a relative sense here; it is the newer of the
two -> combinators; the old one is used currently only for ->m)
2016-01-01 19:49:38 -06:00
Matthew Flatt
fd7b8b29ea
avoid compiler warning
2016-01-01 14:33:16 -07:00
Matthew Flatt
c0915b02b0
pathlist-closure, tar, and tar-gzip: add #:path-filter
...
Also, add `#:skip-filtered-directory?` to `find-files`.
Less significantly, adjust `pathlist-closure` to be consistent in the
way that it includes a separator at the end of a directory path.
2016-01-01 13:52:37 -07:00
Matthew Flatt
c9a1dc781e
raco setup: use lazy ".zo" parsing for dpeendency checking
...
Lazy ".zo" parsing make the dependency check take about half as
long.
2016-01-01 13:52:37 -07:00
Matthew Flatt
d3bb9d0412
fix bytecode-reading time-complexity bug
...
When a tree marshaled to bytecode form has many shared pieces,
the unmarshaling process can lose track of the sharing in one
place and traverse too much of the structure as pieces are
loaded.
2016-01-01 13:52:37 -07:00
Matthew Flatt
8b7f5bc046
faster xexpr->string
2016-01-01 13:52:37 -07:00
Vincent St-Amour
923b5864a5
Small guide fixes.
...
Closes PR 15109.
2016-01-01 12:52:14 -06:00
Vincent St-Amour
8c8a76979b
Fix keyword argument name.
...
Closes PR 15089.
2016-01-01 12:52:14 -06:00
Vincent St-Amour
4c04d4afce
Fix discussion of filter.
...
Closes PR 15083.
2016-01-01 12:52:14 -06:00
Robby Findler
110f1bd9b0
different, more likely fix
2015-12-31 15:36:33 -06:00
Robby Findler
357b692d33
fix typo
2015-12-31 15:33:59 -06:00
Matthew Flatt
567679bf0a
{impersonate,chaperone}-hash: add equal-key-proc wrapper
...
The optional `equal-key-proc` wrapper effectively interposes on calls
to `equal?` and `equal-hash-code` for hash-table keys.
2015-12-31 09:22:57 -07:00
Robby Findler
e92b8610f2
port id-set/c to late-neg
2015-12-30 19:16:36 -06:00
Robby Findler
85c781452d
wording tweaks
2015-12-30 17:57:36 -06:00
Benjamin Greenman
0f7a946dba
put contract-property accessors in an itemlist
2015-12-30 17:52:33 -06:00
Benjamin Greenman
61e21dba22
typo: is-flat-contract? -> is-list-contract?
2015-12-30 17:52:26 -06:00
Robby Findler
df382ca939
improve warnings and fix small problems with contract combinator constructors
2015-12-30 17:13:23 -06:00
Robby Findler
cb2af327e6
port async-channel/c to late-neg proj
...
and drop other projection implementations
(mostly to reduce the testing burden)
2015-12-30 17:13:22 -06:00
Robby Findler
9ee264a0ea
improve performance of define-generic-created contracts and port to late-neg
...
This program runs about 10x faster than it did before this commit, but
seems to still be about 100x slower than the version where you change
an-s to just be (s).
#lang racket/base
(require racket/contract/base racket/generic)
(define-generics id [m id x])
(struct s () #:methods gen:id [(define (m g x) x)])
(define an-s
(contract (id/c [m (-> any/c integer? integer?)])
(s)
'pos 'neg))
(time
(for ([x (in-range 100000)])
(m an-s 2)))
2015-12-30 17:13:22 -06:00
Robby Findler
a97aa8389b
port id-table/c to flat-neg projection
2015-12-30 17:13:22 -06:00
Sam Tobin-Hochstadt
1c3422d420
Fix CI tests for match test move.
2015-12-30 10:39:45 -05:00
Robby Findler
f2f38cdf4a
remove overly aggressive use of procedure-specialize
2015-12-30 09:28:58 -06:00
Robby Findler
a5b3d6b3d0
uncopy some code
2015-12-29 21:46:06 -06:00
Robby Findler
46ace3172f
clean up interaction between strict set/c contracts and mutable sets
2015-12-29 20:28:22 -06:00
Robby Findler
757adac568
make #f always convert into the same (eq?) contract
2015-12-29 16:13:01 -06:00
Robby Findler
4bdde405f6
improve the way that 'struct-out' cooperates with the
...
rest of the contract system, creating and using a slightly
more legitmate blame record and calling into the late-neg
projections instead of using `contract`
2015-12-29 15:48:17 -06:00
Sam Tobin-Hochstadt
893bb56762
Rename to make it clear what to run.
2015-12-29 16:01:22 -05:00
Mira Leung
e2bd1b51bc
Using in-list in for/list
2015-12-29 15:59:40 -05:00
Mira Leung
404c5b2699
Better code style for f3d888
2015-12-29 15:59:40 -05:00
Mira Leung
56ea9f8b9a
Specify names of missing fields for match on structs
2015-12-29 15:59:40 -05:00
Robby Findler
22adc0253b
change sets generics so they default to supplying the optional argument
...
if the given procedure accepts it
2015-12-29 11:26:17 -06:00
Robby Findler
2c5aa96031
change some potentially error-prone provide all-from-outs to explicit provides
...
and make set load its implementation via relative requires
2015-12-29 11:25:45 -06:00
Robby Findler
7056cd5f2a
improve ->i in the case that the dependent contract
...
turns out to be a predicate.
In that case, just call it instead of creating all of the extra junk
that would normally be created by coercing the predicate to a contract
and invoking it
2015-12-29 11:20:59 -06:00
Matthew Flatt
a516304f6b
fix specialization over a nested lambda
2015-12-29 07:04:17 -06:00
Matthew Flatt
26560240f1
adjust "racket-playsound.exe" return code
2015-12-29 07:04:17 -06:00
Robby Findler
d4ca825640
fix set/c name method
2015-12-29 06:56:20 -06:00
Robby Findler
a44ce40b56
fix a bug in 50405a2ca
2015-12-28 22:17:39 -06:00
Robby Findler
f7465f81f1
convert stream/c to late-neg-projection
...
Also, tune the projection to get a few modest performance gains.
This program gets about 20% faster:
#lang racket
(define s (contract (stream/c exact-nonnegative-integer?)
(in-naturals) 'pos 'neg))
(time
(for ([x (in-range 1000)])
(for/fold ([s s])
([x (in-range 100)])
(stream-rest s))))
and this program gets about 15% faster:
#lang racket
(define f (contract (-> (stream/c exact-nonnegative-integer?)
any)
(λ (x) 1)
'pos 'neg))
(define l (make-list 10000 0))
(time (for ([x (in-range 1000)])
(f l) (f l) (f l) (f l) (f l)))
2015-12-28 21:57:03 -06:00
Robby Findler
ff31b01505
add some error checking for custom set type constructors
2015-12-28 16:59:43 -06:00
Robby Findler
50405a2ca9
fix chaperone-hash-set and impersonate-hash-set for custom-set-types
2015-12-28 16:59:42 -06:00
Matthew Flatt
e08188aeda
add source for "racket-playsound.exe"
2015-12-28 07:37:29 -06:00
Matthew Flatt
cefcbdf802
Windows: ensure 'kill subprocesses end with Racket
...
Use a job object to ensure that subprocesses that are meant
to be killed by the current custodian are reliably terminated
if Racket exits for any reason.
2015-12-28 07:37:29 -06:00
Matthew Flatt
afa01fa763
allow optimizations around procedure-specialize
...
Although `procedure-specialize` should be useful in places where
inlining does not apply, allowing inlining and related optimizations
through it, anyway.
2015-12-28 07:37:29 -06:00
Robby Findler
b3d05de304
improvements to set/c
...
- use chaperone-hash-set for set/c when the contract allows only hash-sets
- add a #:lazy flag to allow explicit choice of when to use laziness
(but have a backwards-compatible default that, roughly, eschews laziness
only when the resulting contract would be flat)
2015-12-27 22:56:04 -06:00
Robby Findler
bc12019af4
allow the interposition procedures to all be #f in chaperone-hash-set and impersonate-hash-set
2015-12-27 21:02:20 -06:00
Robby Findler
8f2874e4b5
fix impersonator properties for chaperone-hash-set and impersonate-hash-set
2015-12-26 22:37:00 -06:00
Robby Findler
daf19869de
chaperone-hash-set fixes
2015-12-26 22:09:14 -06:00
Robby Findler
1b6705f3d9
typo
2015-12-26 17:01:22 -06:00
