From: Matthew D Swank
Subject: pathname weirdness
Date: 
Message-ID: <pan.2006.02.10.13.17.25.807936@c.net>
I am having trouble running examples from the Verrazano test suite in
sbcl 0.9.9, and I have isolated the problem to the following error:

(with-open-file (out (merge-pathnames "vzntemp.cpp" 
                                      (make-pathname :device "/"
                                                     :directory "/tmp"))
                     :direction :output
                     :if-exists :supersede)
  (format out "~a~%" "this is a test"))

-->
The path #P"//tmp/vzntemp.cpp" does not exist.
   [Condition of type SB-INT:SIMPLE-FILE-ERROR]

Restarts:
  0: [ABORT-REQUEST] Abort handling SLIME request.
  1: [TERMINATE-THREAD] Terminate this thread 

However, 

(with-open-file (out #P"//tmp/vzntemp.cpp"
                     :direction :output
                     :if-exists :supersede)
  (format out "~a~%" "this is a test"))

works just fine.

What is going on?

Matt

-- 
"You do not really understand something unless you can
 explain it to your grandmother." — Albert Einstein.
From: Peter Seibel
Subject: Re: pathname weirdness
Date: 
Message-ID: <m2bqxfqo8e.fsf@gigamonkeys.com>
Matthew D Swank <·······································@c.net> writes:

> I am having trouble running examples from the Verrazano test suite in
> sbcl 0.9.9, and I have isolated the problem to the following error:
>
> (with-open-file (out (merge-pathnames "vzntemp.cpp" 
>                                       (make-pathname :device "/"
>                                                      :directory "/tmp"))
>                      :direction :output
>                      :if-exists :supersede)
>   (format out "~a~%" "this is a test"))
>
> -->
> The path #P"//tmp/vzntemp.cpp" does not exist.
>    [Condition of type SB-INT:SIMPLE-FILE-ERROR]
>
> Restarts:
>   0: [ABORT-REQUEST] Abort handling SLIME request.
>   1: [TERMINATE-THREAD] Terminate this thread 
>
> However, 
>
> (with-open-file (out #P"//tmp/vzntemp.cpp"
>                      :direction :output
>                      :if-exists :supersede)
>   (format out "~a~%" "this is a test"))
>
> works just fine.
>
> What is going on?

It may be that "//tmp/vzntemp.cpp" is being parsed into some different
pathname structure than the one you explicitly specified. For
instance, what is:

  (pathname-device #p"//tmp/vzntemp.cpp") ==> ???

On Allegro on OS X, it returns :UNSPECIFIC. And

  (equal #p"//tmp/vzntemp.cpp" #p"/tmp/vzntemp.cpp") ==> T

which suggests the extra leading #\/ is being ignored. My guess is the
same thing is happening in SBCL.

-Peter

-- 
Peter Seibel           * ·····@gigamonkeys.com
Gigamonkeys Consulting * http://www.gigamonkeys.com/
Practical Common Lisp  * http://www.gigamonkeys.com/book/
From: Matthew D Swank
Subject: Re: pathname weirdness
Date: 
Message-ID: <pan.2006.02.10.14.40.12.107049@c.net>
On Fri, 10 Feb 2006 14:31:45 +0000, Peter Seibel wrote:

> (pathname-device #p"//tmp/vzntemp.cpp")

This is nil in sbcl, and using that as the device, (or omitting the device
argument altogether) seems to work.

Thanks,

Matt

-- 
"You do not really understand something unless you can
 explain it to your grandmother." — Albert Einstein.
From: Pascal Bourguignon
Subject: Re: pathname weirdness
Date: 
Message-ID: <87d5hvhyak.fsf@thalassa.informatimago.com>
Peter Seibel <·····@gigamonkeys.com> writes:
> On Allegro on OS X, it returns :UNSPECIFIC. And
>
>   (equal #p"//tmp/vzntemp.cpp" #p"/tmp/vzntemp.cpp") ==> T
>
> which suggests the extra leading #\/ is being ignored. My guess is the
> same thing is happening in SBCL.

Don't some systems (perhaps POSIX systems) give different sematics
to "//path" vs. "/path"?

-- 
__Pascal Bourguignon__                     http://www.informatimago.com/

THIS IS A 100% MATTER PRODUCT: In the unlikely event that this
merchandise should contact antimatter in any form, a catastrophic
explosion will result.
comp.lang.lisp / 2006 / feb