gmane.lisp.steel-bank.general

Re: Using /bin/more through RUN-PROGRAM

Teemu Likonen — 2013-06-16T15:17:33


Thanks, but... I did various changes and tests but still couldn't make
it work. For example, ":input nil", ":pty t" and using the value of
PROCESS-PTY slot for output stream still prints nothing. I'm using SBCL
version 1.1.8.17-123a8dd.

So I'm out of ideas. If someone knows how it would work I'd appreciate a
simple code example like this.


    (let ((out (sb-ext:process-pty
                (sb-ext:run-program "/bin/more" nil
                                    :wait nil
                                    :input nil
                                    :output t
                                    :error t
                                    :pty t))))
      (dotimes (i 100)
        (format out "~A~%" i))
      (force-output out))
------------------------------------------------------------------------------
This SF.net email is sponsored by Windows:

Build for Windows Store.

http://p.sf.net/sfu/windows-dev2dev_______________________________________________
Sbcl-help mailing list
Sbcl-help< at >lists.sourcefo

Re: Using /bin/more through RUN-PROGRAM

Philipp Marek — 2013-06-16T14:24:05

...
...
more / less expect to get a PTY, so you'll need the :PTY T arguments, too.





------------------------------------------------------------------------------
This SF.net email is sponsored by Windows:

Build for Windows Store.

http://p.sf.net/sfu/windows-dev2dev

Using /bin/more through RUN-PROGRAM

Teemu Likonen — 2013-06-16T05:04:00

I want to use pager /bin/more (or /usr/bin/less) internally in my
program so that certain output text is automatically written to the
pager program. Can't make it work, though. I thought the following code
should work but it doesn't:


    ;;; test.lisp

    (let ((out (sb-ext:process-input
                (sb-ext:run-program "/bin/more" nil
                                    :wait nil
                                    :input :stream
                                    :output t
                                    :error t))))
      (dotimes (i 100)
        (format out "~A~%" i))
      (force-output out))


Let's run it:

    $ sbcl --script test.lisp 
    $ 

Nothing is printed. If I switch "/bin/more" to "/bin/cat" it prints
normally the numbers from 0 to 99.

Is there a way to use a pager program internally in SBCL?
------------------------------------------------------------------------------
This SF.net email is sponsored by Windows:

Build for Windows Store.

http://p.sf.net/sfu/windows-dev2dev_____

Re: shadow special forms?

Alex Lier — 2013-06-14T22:53:54

Thank you very much.
This solution works perfectly for me.

Greetings
Alex



------------------------------------------------------------------------------
This SF.net email is sponsored by Windows:

Build for Windows Store.

http://p.sf.net/sfu/windows-dev2dev
_______________________________________________
Sbcl-help mailing list
Sbcl-help< at >lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sbcl-help

Question about introspecting setf expanders

Didier Verna — 2013-06-13T19:40:23

  Hello,

Given a symbol ACCESS-FN, I can tell whether a setf expander in short
form is installed for it by doing (sb-int:info :setf :inverse ACCESS-FN).

My question is: can I do the opposite? That is, given a symbol UPDATE-FN
(or a function, or even a macro), is there a way to figure out whether
there is a setf expander for it and what is the corresponding ACCESS-FN?
 
Thank you!

Re: shadow special forms?

Pascal J. Bourguignon — 2013-06-02T19:17:44


For an example, have a look at the cl-stepper in
https://gitorious.org/com-informatimago/com-informatimago/trees/master/common-lisp/lisp

Re: shadow special forms?

Stig Hemmer — 2013-06-02T19:08:07


I believe the short answer is that you can't.

According to the language standard, you are not allowed to fiddle with
symbols in the COMMON-LISP package.

SBCL allows you to bypass this with package locks, but when doing so you
are strictly on your own.  If it doesn't work, well, that's just too bad.
 Also, even if it does work, it might break in the next version of SBCL.

Are you sure that you need to fiddle with COMMON-LISP:FUNCTION (or
whatever).   Might you not get the results you need with
FAKE-COMMON-LISP:FUNCTION?

The idea to set up a package that USEs and reEXPORTs most of CL, only
substituting the symbols you want to play with.

You might need to give a definition like
(defmacro fcl:function (name)
   `(cl:function ,name))
to make it work outside the regions in question.

Stig Hemmer
------------------------------------------------------------------------------
Get 100% visibility into Java/.NET code with AppDynamics Lite
It's a free troubleshooting tool designed for production
Get down to code-l

shadow special forms?

Alex Lier — 2013-06-02T12:02:47

------------------------------------------------------------------------------
Get 100% visibility into Java/.NET code with AppDynamics Lite
It's a free troubleshooting tool designed for production
Get down to code-level detail for bottlenecks, with <2% overhead.
Download for free and get started troubleshooting in minutes.
http://p.sf.net/sfu/appdyn_d2d_ap2_______________________________________________
Sbcl-help mailing list
Sbcl-help< at >lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sbcl-help

Re: there is no package with name "SB-EXT"

Peter Münster — 2013-04-26T13:54:30


Similar, I've found it:

Long time ago, I've added this line

echo ":CLISP:E::lisp::/usr/bin/clisp:" >/proc/sys/fs/binfmt_misc/register

to `/etc/init.d/boot.local'. (and forgotten in the meantime...)

Thus the shebang line was not used and clisp interpreted the file!

Sorry for the noise,

Re: there is no package with name "SB-EXT"

Stas Boukarev — 2013-04-26T12:50:54

That's exactly how you do it. Your SBCL installation must be broken.

there is no package with name "SB-EXT"

Peter Münster — 2013-04-26T12:35:04

Hi,

How could I please get the command line arguments in a script?

Example, that fails with `there is no package with name "SB-EXT"':

--8<---------------cut here---------------start------------->8---
#!/usr/bin/sbcl --script
(print sb-ext:*posix-argv*)
--8<---------------cut here---------------end--------------->8---

$ sbcl --version
SBCL 1.1.2-2.1.1-suse

TIA for any hints,

texinfo-5.1

2013-04-20T11:35:59

Hello *,

With the cutrent texinfo-5.1, when I go to the directory doc/manual and 
say
make html
or
make info
I get a long list of error messages and warnings (make pdf succeeds; in 
doc/internals, make html and make info succeed). There were some 
incompatible changes in the texinfo language. Are there plans to adjust 
sbcl.texinfo (and, maybe, other *.texinfo files) to texinfo-5.x? Should I 
file a bug?

Andrey

------------------------------------------------------------------------------
Precog is a next-generation analytics platform capable of advanced
analytics on semi-structured data. The platform includes APIs for building
apps and a phenomenal toolset for data science. Developers can use
our toolset for easy data analysis & visualization. Get a free account!
http://www2.precog.com/precogplatform/slashdotnewsletter

Re: asdf-2.33

Carlos Konstanski — 2013-04-13T16:31:17

For those who don't use gentoo, and are unfamiliar with the rinky-dinky
details of portage and its common lisp support:

Any gentoo user who is serious about their common lisp uses the lisp
overlay. The main portage tree's lisp packages are ancient. For some
reason they keep sbcl sorta kinda up to date, but all the asdf packages
are worthlessly old. I think it's because the lisp maintainers don't
have access to the main tree, so they are forced to do everything in an
overlay.

sbcl-1.1.6 is in the lisp overlay. Last I checked it was still masked
because it needed further testing. Masking it caused portage to try
installing the mainline sbcl, asdf-binary-locations and gentoo-init. It
is currently necessary to unmask dev-lisp/sbcl, dev-lisp/uiop and
dev-lisp/asdf in order to stay on the overlay's packages.

The lisp overlay maintainers have this bad habit of removing older
ebuilds as soon as a new one becomes available, making it impossible to
go back a version. That's why masking the latest sbcl in the overl

Re: asdf-2.33

Waldek Hebisch — 2013-04-13T15:53:00

BTW: Do you know that by default Gentoo still ships broken sbcl?
AFAICS default is 'dev-lisp/sbcl-1.0.55-r1' and in startup
script it pulls 'dev-lisp/gentoo-init-0.1' which uses
ASDF-BINARY-LOCATIONS.  Effect:

sbcl
(require :sb-posix)

gives compliants about ASDF 2 beeing incompatible with
ASDF-BINARY-LOCATIONS.  I saw the problem mentined in October 2012...

Re: asdf-2.33

2013-04-13T15:22:04

Sorry for the noise. I had 2 different versions of asdf installed; after I 
removed one, I can build maxima successfully (Gentoo linux, x86_64), and 
sbcl shows the best time running the maxima testsuite.

Andrey

------------------------------------------------------------------------------
Precog is a next-generation analytics platform capable of advanced
analytics on semi-structured data. The platform includes APIs for building
apps and a phenomenal toolset for data science. Developers can use
our toolset for easy data analysis & visualization. Get a free account!
http://www2.precog.com/precogplatform/slashdotnewsletter

Re: asdf-2.33

James M. Lawrence — 2013-04-13T15:03:23

You are headed for an unrelated problem -- there are reports that
maxima doesn't build on sbcl-1.1.6 due to the SVREF issue.

------------------------------------------------------------------------------
Precog is a next-generation analytics platform capable of advanced
analytics on semi-structured data. The platform includes APIs for building
apps and a phenomenal toolset for data science. Developers can use
our toolset for easy data analysis & visualization. Get a free account!
http://www2.precog.com/precogplatform/slashdotnewsletter

Re: asdf-2.33

Christophe Rhodes — 2013-04-13T06:41:01


I'd say the one distributed with sbcl itself is the only one I can
suggest... otherwise, you might have more luck on the asdf mailing list.

Best wishes,

Christophe

------------------------------------------------------------------------------
Precog is a next-generation analytics platform capable of advanced
analytics on semi-structured data. The platform includes APIs for building
apps and a phenomenal toolset for data science. Developers can use
our toolset for easy data analysis & visualization. Get a free account!
http://www2.precog.com/precogplatform/slashdotnewsletter

asdf-2.33

2013-04-13T05:16:18

Hello *,

I maintain maxima in Gentoo. There are experimental ebuilds of sbcl-1.1.6 
and asdf-2.33 in Gentoo (in the lisp overlay). After installing them, I 
cannot compile maxima. The problem is in server.lisp which requires 
sb-posix. This is what I get:

grozin< at >dns ~ $ sbcl
This is SBCL 1.1.6, an implementation of ANSI Common Lisp.
More information about SBCL is available at <http://www.sbcl.org/>.

SBCL is free software, provided as is, with absolutely no warranty.
It is mostly in the public domain; some portions are provided under
BSD-style licenses.  See the CREDITS and COPYING files in the
distribution for more information.
* (require :sb-posix)
ASDF could not load sb-posix because Error while trying to load definition 
for system sb-posix from pathname /usr/lib64/sbcl/sb-posix/sb-posix.asd: 
invalid number of arguments: 3.

debugger invoked on a ASDF/FIND-SYSTEM:LOAD-SYSTEM-DEFINITION-ERROR: Error 
while trying to load definition for system sb-posix from pathname 
/usr/lib64/sbcl/sb-posix/sb-posix.as

Re: Reading a file line-by-line

Nikodemus Siivola — 2013-03-27T17:26:27

By the by:

        ^^^

                                                 ^^^

Yeah. Makes no sense to me either. Either I mismeasured either one, or
there were more lseek calls hiding inside other WITHOUT-INTERRUPTS
uses. Shows also how coherent I'm today.

Cheers,

 -- Nikodemus

------------------------------------------------------------------------------
Own the Future-Intel&reg; Level Up Game Demo Contest 2013
Rise to greatness in Intel's independent game demo contest.
Compete for recognition, cash, and the chance to get your game 
on Steam. $5K grand prize plus 10 genre and skill prizes. 
Submit your demo by 6/6/13. http://p.sf.net/sfu/intel_levelupd2d

Re: Reading a file line-by-line

Nikodemus Siivola — 2013-03-27T15:52:35


About 20% of time seems to be spend in lseek(2).

Attached patch gets rid of some of the redundant lseek calls,
improving performance on my test-case by about 40%. WARNING: virtually
untested.

The rest of the slowness is seems mostly due to our embarrassingly
slow external format code. (Assuming the file contains only 7-bit
characters, :ASCII, :LATIN-1, and :UTF-8 are all in the same ballpark,
though, which while not great isn't terrible either. Other formats --
and higher code points probably fare less well.)

Cheers,

 -- Nikodemus
------------------------------------------------------------------------------
Own the Future-Intel&reg; Level Up Game Demo Contest 2013
Rise to greatness in Intel's independent game demo contest.
Compete for recognition, cash, and the chance to get your game 
on Steam. $5K grand prize plus 10 genre and skill prizes. 
Submit your demo by 6/6/13. http://p.sf.net/sfu/intel_levelupd2d_______________________________________________
Sbcl-help mailing list
Sbcl-help< at >lists.sourcefor

Reading a file line-by-line

Rupert Swarbrick — 2013-03-25T22:28:23

Hi there,

I've been trying to write a bit of code for Maxima that reads in a
texinfo file and works out some byte offsets. The code basically boils
down to a loop of the form

(defun grok-info-file (pathname)
  (with-open-file (stream pathname :direction :input)
    (let ((line) (line-number 0) (pos) (line-offsets))
      (loop
         (setf pos (file-position stream))
         (when (eq (setf line (read-line stream nil :eof)) :eof) (return))

         (push pos line-offsets)
         (incf line-number)

         (some-complicated-stuff)))))

(if you care, the entire function can be found in src/parse-info.lisp at
http://sourceforge.net/p/maxima/code/?branch=ref/parse-info, although
SF's git viewing application is pretty nasty and seems to think the
source files are binary so won't show them)

Profiling this to see where all the time was going, I was surprised to
see that we were spending ages in FILE-POSITION. Looking at SBCL's
stream.lisp, it seems that we're spending the time calculating byte
lengths of