suzanne.soy/racket
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
35,041 Commits 17 Branches 54 Tags 278 MiB
c4401313d4
Commit Graph

35041 Commits

This Branch
This Branch
All Branches
Author SHA1 Message Date
Matthew Flatt
fe68c9ab81 fix docs again on struct in racket/signature 
The repair of 7bfe2eadab wasn't right; the `struct` form is treated
correctly in `define-signature`, but not in `racket/signature`.
 2015-02-08 06:52:24 -07:00
Matthew Flatt
2ada651dd3 {chaperone,impersonate}-struct: allow structure type as a witness 
Also, do not allow `struct-type` as a wrapped operation in
`chaperone-stuct` without a witness.

Related to PR 14970
 2015-02-08 06:52:24 -07:00
Matthew Flatt
8e8c9842fa raco {pkg,planet} <subcommand>: improve error for ambiguous <subcommand> 
Report a user error instead of an internal error.

Closes PR 14969
 2015-02-08 06:07:09 -07:00
Matthew Flatt
be8f70fffb racket/unit: static checking of initialization dependencies 
When using `compound-unit/infer` and similar, check the `link` clause
against each unit's static information for initialization dependencies.
Also, propagate dependency information in `define-compount-unit`.
 2015-02-06 09:22:01 +01:00
Matthew Flatt
53fb33144e add unit-static-init-dependencies 2015-02-06 08:55:58 +01:00
Matthew Flatt
7bfe2eadab fix docs for struct in define-signature 
Unofrtunately, `struct` in a signature corresponds to `define-struct`.
 2015-02-06 08:55:58 +01:00
Leif Andersen
5e04ec431f Update .travis.yml for raco pkg new. 2015-02-05 17:30:32 -05:00
Ryan Culpepper
991340e0b0 db/sqlite3: fix custodian shutdown wrt locking 2015-02-04 20:19:13 -05:00
Matthew Flatt
acdb0b0e90 fix prefab-key? for inferred field count 
Instead of inferring a field count of 0, accept a key that
works with some number of fields.

Closes PR 14964
 2015-02-03 10:48:18 +01:00
Robby Findler
ffd77693ee fix a bug in random contract generation that could cause 
nested structure to have contract-random-generate-fail stuck
into it

and clean up some confusing structure in the random generator
 2015-02-02 22:15:05 -06:00
vraid
68074f7fd7 fix typo 2015-02-02 17:27:42 -05:00
Matthew Flatt
8265c968fc openssl: recognize version "1.0.1j" 
Closes PR 14954
 2015-01-30 13:29:36 -07:00
Matthew Flatt
8d49a91dce pretty-print: fix for a current inspector that sees through internals 
If the current inspector can inspect the internal `unquoted` structure
type, then `unquoted?` must come before `struct?`.
 2015-01-29 19:33:36 -07:00
Ryan Culpepper
3029867e9f fix reified-syntax-class-curry (missing role argument) 2015-01-29 20:37:50 -05:00
Matthew Flatt
b7e86ef6c9 syntax/modcollapse: fix to new collapse-module-path-index variant 2015-01-28 16:21:43 -07:00
Matthew Flatt
83974a42ee native-libs script: build MPFR for Windows as thread-safe 2015-01-27 20:07:49 -07:00
Matthew Flatt
60704c9198 Windows: fix reparsing with UNC targets 2015-01-27 18:07:46 -07:00
Matthew Flatt
f30b3a50fd Windows: fix problems with junctions and symlinks 
Racket wasn't reparsing correctly; the strategy worked ok
for links created by `mklink`, but not with other tools that
leave the "printed name" field blank.

A consequence of various fixes is that reparse points like
"My Documents" (in a typical configuration) correctly resolve
to actual paths like "Documents".

Finally, `directory-exists?` didn't handle root directories like
"C:/" correctly. The query would actually report properties of
the OS-level current working directory, and when junctions are
involved, the current directory can be a link instead of a directory.

Relevant to PR 14950 and PR 14912
 2015-01-27 17:48:52 -07:00
Matthew Flatt
d3383e3e35 dynamic-require: fix re-export shortcut 2015-01-27 13:46:33 -07:00
Matthew Flatt
a72ef3ec05 syntax-local-lift-require: fix problems for meta-compile-time use 
Various repairs correct problems with `local-require` in a
phase-1 context.
 2015-01-27 09:49:28 -07:00
Matthew Flatt
7bee7bbadc collapse-module-path-index: support relative module path flattening 
Unlike `collapse-module-path`, it makes sense for
`collapse-module-path-index` to convert a relative module path index
to a plain module path. In other words, `collapse-module-path-index`
can convert a module path index to a module path.
 2015-01-27 08:40:10 -07:00
Robby Findler
d74b0a6bf4 between/c (unlike integer-in) can actually just disappear 
when the arguments are = to each other
 2015-01-25 19:23:38 -06:00
Robby Findler
6551cc00d6 make (integer-in x x) produce (and/c x exact?) 
fix bug in 9f59d57ee
 2015-01-25 15:31:04 -06:00
Robby Findler
87a231b792 fix a bug in the way that an old style projection 
was created when a val-first-projection was needed
 2015-01-25 15:08:12 -06:00
Robby Findler
ffbf01ad4e more trickyness for = and numbers in random generation 2015-01-25 15:08:12 -06:00
Gustavo Massaccesi
e36382d500 Add SCHEME_PRIM_WANTS_FLONUM_SECOND flag to flexpt 2015-01-25 07:51:45 -07:00
Gustavo Massaccesi
6ab68eb97d Add SCHEME_PRIM_PRODUCES_FIXNUM flag to unsafe-fxvector-ref 2015-01-25 07:51:45 -07:00
Gustavo Massaccesi
6d5597c090 Change type of result of unsafe-??vector-ref 2015-01-25 07:51:45 -07:00
Gustavo Massaccesi
9ecf98a0a0 Add documentation of unsafe-fxvector-length and friends 2015-01-25 07:51:45 -07:00
Matthew Flatt
e2261096cb doc clarifications on eqv? 
Part of the clarification is duplicating information about numbers
and character in the documentation of `eqv?`. Since those two type
are the only special cases of `eqv?`, the duplication seems helpful
and managable.
 2015-01-24 10:12:35 -07:00
Matthew Flatt
68c5d3d1d6 fix error message for inexact->exact on +inf.f 2015-01-24 10:12:35 -07:00
Matthew Flatt
bac11bf8f5 db/sqlite3: fix recur for busy db 2015-01-24 10:12:35 -07:00
Robby Findler
eb7c6653f3 make the contract random generator sometimes 
generate inexact/exact variants of numbers
when the exact/inexact one was originally
present in the contract
 2015-01-23 22:40:54 -06:00
Robby Findler
cc642c3382 change +nan.0 and +nan.f, when viewed as contracts, to 
be equal?-based contracts instead of = based contracts.

Before this change, the contract (or/c 1 2 +nan.0) was the same
contract as (or/c 1 2), because +nan.0 was the same contract as
the predicate (lambda (x) (= x +nan.0)), which is the same as
(lambda (x) #f). Now, +nan.0 and +nan.f are the only numbers
that are treated as equal?-based contracts, but this means that
(or/c 1 2 +nan.0) actually accepts +nan.0.
 2015-01-23 21:57:51 -06:00
Matthew Flatt
778a95294c fix requested stack depth as needed by on-demand JITting 
Found by stack-overflow checking added in 3408209f66.
 2015-01-23 12:10:04 -07:00
Robby Findler
430a4b08c4 fix typo 2015-01-23 09:41:07 -06:00
Robby Findler
6bf6f4d392 add a generator for exact-integer 2015-01-22 20:04:36 -06:00
Matthew Flatt
2ffb546c95 fix vector-set-performance-stats! for CGC 
Also, fix the build for a no-futures, no-places configuration.
 2015-01-22 13:03:00 -07:00
Matthew Flatt
db40c2f4ce corrections to GC out-of-memory handling 2015-01-22 10:16:32 -07:00
Matthew Flatt
cffb63be56 correction to recent repair to places 
Corrects 5b20690876
 2015-01-22 10:16:32 -07:00
Matthew Flatt
7196dc0e74 add peak memory use to vector-set-performance-stats! 2015-01-22 10:16:32 -07:00
Robby Findler
9f59d57ee4 make (integer-in n n) produce just n 2015-01-21 21:53:31 -06:00
Robby Findler
cb68141ff6 do tests for f2bacdc6f properly 2015-01-21 10:05:09 -06:00
Robby Findler
7003ca6eb3 add test case for previous commit 2015-01-21 08:39:47 -06:00
Markus Bertheau
f2bacdc6f3 Correct 'german time format 
We separate hour and minute with a colon, not with a period.
 2015-01-21 08:39:46 -06:00
Matthew Flatt
0c13a4a1f1 places: avoid redundant atexit() registrations 
Register only in the original place.
 2015-01-21 06:11:05 -07:00
Matthew Flatt
8035ee354c document format of GC logged message text 2015-01-21 05:39:04 -07:00
Matthew Flatt
1893f73fac fix GC peak-memory logging 2015-01-21 05:10:51 -07:00
Matthew Flatt
3b866db378 raco setup: render each document with a separate namespace 
Rending a document can deserialize values, which can load modules
that would otherwise not be loaded by Scribble, so render each
document with a fresh namespace that is discarded after rendering.
 2015-01-21 04:53:14 -07:00
Matthew Flatt
5b20690876 places: no allocation while low-level locks are held 2015-01-20 19:26:42 -07:00