Fcalva/chibi-scheme
1
0
Fork 0
mirror of https://github.com/ashinn/chibi-scheme.git synced 2025-05-19 05:39:18 +02:00
Code Issues Projects Releases Packages Wiki Activity Actions
4019 commits 8 branches 19 tags 26 MiB
Commit graph

1164 commits

Author SHA1 Message Date
Marc Nieper-Wißkirchen
1f9b4796d6 Parse syntax-rules more strictly 
See issue #767.
 2021-08-03 17:05:37 +02:00
Alex Shinn
18d0adf13b signal error on improper lists passed to map/for-each 2021-07-28 17:00:40 +09:00
Alex Shinn
9710962cd2 don't assume map works on improper lists in cpp-define 2021-07-28 10:35:05 +09:00
Jürgen Geßwein
b23db00aed Fix indentation of test runner output 
Fix standard test runner so that its output is properly indented and
lines are properly terminated.

Refactor standard test runner so that it is possible to plug in another
runner with different output.
 2021-07-25 13:21:34 +02:00
Alex Shinn
5b8e196e0f parse-seq should check for ignored value in fast paths as well (issue #757) 2021-07-19 21:49:39 +09:00
Roger Crew
3337049811 shorter factor (issue #751 cont.) 2021-06-30 00:29:54 -07:00
Roger Crew
2759aaa306 add factor-alist and export it (chibi math prime) 
a more useful version of factor
 2021-06-29 23:48:12 -07:00
Roger Crew
680aede9ae totient and aliquot rewrite + corrected tests for n=1 (issue #751 cont.) 2021-06-29 23:48:12 -07:00
Roger Crew
b89bd9f889 faster factor, (factor 1) = () (issue #751 cont.) 
no need to go up to sqrt(n), Instead track i^2 and quit when that gets
larger than the (remaining) n (i.e., not the original n)
 2021-06-29 23:48:12 -07:00
Roger Crew
41aa1a918e miller-rabin-composite? rewrite (issue #751) 
modular-root-of-one? is replaced with the correct witness tester
 2021-06-29 23:48:09 -07:00
Roger Crew
7d39108e72 factor-twos cps version using first-bit-set 
first-bit-set is way faster than looping
 2021-06-29 23:43:53 -07:00
Alex Shinn
73da0a88d4 scan for appropriate 2nd element to take the mean with when calling vector-find-median on an even length vector (issue #754) 2021-06-29 21:09:41 +09:00
Alex Shinn
f3bccf1f7b removing unintended export 2021-06-28 17:31:14 +09:00
Alex Shinn
378b56a0c3 adding opt-lambda* and define-opt* 2021-06-28 17:27:22 +09:00
Alex Shinn
0fbd89dd00 ignore keywords in non-keyword positions (issue #753) 2021-06-28 16:19:43 +09:00
Alex Shinn
05c546e38d match fix for (a ...) patterns where a was already bound - thanks to Andy Wingo 2021-06-21 16:44:02 +09:00
Alex Shinn
5207bdfde2 Defining list->u8vector (issue #749). 2021-06-18 13:04:20 +09:00
Alexei Lozovsky
6f35aa75f4
Fix usage of signbit() in SRFI 144 
C standard defines signbit() as a macro returning "non-zero value" for
negative arguments (see 7.12.3.6 of C11 standard). SRFI 144's flsign-bit
is defined to return exactly 1.

Make sure to convert the result of signbit() call into "boolean int"
which is either 0 or 1.

This is not a theoretical issue. This causes SRFI 144 test suite to fail
on many architectures that are not x86_64.

GCC on x86_64 compiles signbit() as

        movmskpd %xmm0, %eax
        andl     $1, %eax

which indeed returns either 0 or 1. movmskpd extracts 2-bit sign mask
from the FP value in src register and stores that in low-order bits of
the dst register. Then the unneded extra bit is masked out, leaving only
the lowest bit set or unset.

However, other architectures don't have such conveniences and go with
more direct approach. For example, GCC on ARMv7 produces this:

        sub     sp, sp, #8
        vstr.64 d0, [sp]
        ldr     r0, [sp, #4]
        and     r0, r0, #0x80000000
        add     sp, sp, #8
        bx      lr

which effectively returns either 0 or -1. Generated code masks out
everything but the sign bit and returns the result as is. The value
0x80000000 is the representation of -1.

Even on i386 signbit() is compiled as

        fldl    4(%esp)
        fxam
        fnstsw  %ax
        fstp    %st(0)
        andl    $512, %eax
        ret

which effectively returns either 0 or 512: fxam sets C1 bit FPU status
word to the sign of FP value, then the status word is extracted, the
"sign bit" is masked out, and left as is.
 2021-06-06 13:49:44 +09:00
Alex Shinn
5610653c20 servlets should assume html if no Content-Type is specified 2021-05-28 12:07:23 +09:00
Alex Shinn
05ca40fa51 chibi doc fixes for irregex manual 2021-05-14 13:49:07 +09:00
Alex Shinn
4a06997978 guard against failed module-ref 2021-05-13 22:10:30 +09:00
Alex Shinn
037a7b24fc exporting array-coeffs 2021-05-10 22:56:36 +09:00
Alex Shinn
7b2fbb0bf2 fixing ffi doc extraction for cond-expanded procedures 2021-05-10 18:40:33 +09:00
Alex Shinn
3eab7bf226 package SRFI 179 2021-05-10 17:49:02 +09:00
Alex Shinn
6e5278b7a1 separating out SRFI 179 base library 2021-05-07 22:39:53 +09:00
Alex Shinn
ca47a41ccf array simplification and performance tweaks 2021-05-07 16:15:48 +09:00
Alex Shinn
12ad1d37d8 add docs and tests for assert, unify with SRFI 145 2021-05-06 20:34:03 +09:00
Alex Shinn
249f4f21ba allow tracing macros 2021-05-06 10:17:03 +09:00
Alex Shinn
1db8a573f1 more extensive list of constructs not to descend into 2021-05-05 08:08:12 +09:00
Alex Shinn
9a9f974d69 assert improvements: dedup vars, ignore lambdas, allow report: 2021-05-05 07:56:14 +09:00
Alex Shinn
e3083062fa Flattening indexing further. 
We can pre-subtract each coeff times its lower bound from the
base coefficient in SRFI 179 indexers.
 2021-04-30 14:25:39 +09:00
Alex Shinn
d11106b2f7 Fix upper bounds checks in u64vectors. 2021-04-30 14:02:29 +09:00
Alex Shinn
3c138dc808 Fix validation on specialized-array-reshape. 2021-04-30 13:38:53 +09:00
Alex Shinn
76284f79f0 flattening array indexers 2021-04-28 22:53:16 +09:00
Alex Shinn
ef9daf22c8 handle include-shared in analyze-module 2021-04-26 14:27:11 +09:00
Alex Shinn
cd5bf03537
Merge pull request #740 from lassik/219 
Add SRFI 219: Define higher-order lambda
 2021-04-18 18:58:39 +09:00
Lassi Kortela
7178d22928 Add SRFI 219: Define higher-order lambda 2021-04-18 12:21:32 +03:00
Alex Shinn
d0510bebe6 simplify array-reduce 2021-04-16 23:35:16 +09:00
Alex Shinn
c8f5f49890 add initial SRFI 179 implementation 2021-04-16 19:53:51 +09:00
Alex Shinn
487ea21d77 check value domains on uvector-set! ops 2021-04-16 10:13:37 +09:00
Alex Shinn
d64f159608 fix docs on test-group 2021-04-16 10:10:18 +09:00
Alex Shinn
fc6e5da915 make-u1vector takes an optional fill 2021-04-15 17:02:44 +09:00
Alex Shinn
e74614d4b3 removing redundant uvector definitions 2021-04-14 17:06:34 +09:00
Alex Shinn
cfbd64f085 fixing bug in prime-above 2021-04-14 16:47:29 +09:00
Alex Shinn
3fc9c22245 fix regression from shadowed bindings in process->output+error+status 2021-04-09 23:09:53 +09:00
Alex Shinn
4bd4f08b59 fix longstanding todo and get signature from analyzed procedure forms 2021-04-08 23:00:12 +09:00
Alex Shinn
8c45c3fb19 better text display 2021-04-08 22:59:34 +09:00
Alex Shinn
f7b546769c update logging tests for fixes in string-split 2021-04-08 10:08:19 +09:00
Alex Shinn
8b27ce9726 add proper grammar support to srfi 130 string-split 2021-04-02 13:51:02 +09:00
Alex Shinn
d80589144d close stdout/err in process->foo utilities 
Relying on gc can accumulate many open fd's,
which is bad for code outside of chibi.
 2021-03-31 09:23:21 +09:00