clisp-devel

fresh-line on mingw

[Sam S. <sd...@gn...>]

if grep 'Line1.*Line2' fresh-line.out > /dev/null; then exit 1; fi
make: *** [check-fresh-line] Error 1

fresh-line.out:

*TERMINAL-IO* = #<IO TERMINAL-STREAM>
*STANDARD-OUTPUT* = #<IO TERMINAL-STREAM>
*ERROR-OUTPUT* = #<OUTPUT UNBUFFERED FILE-STREAM CHARACTER>
*QUERY-IO* = #<IO SYNONYM-STREAM *TERMINAL-IO*>
*DEBUG-IO* = #<IO TWO-WAY-STREAM #<INPUT CONCATENATED-STREAM> #<IO TERMINAL-STREAM>>
*TRACE-OUTPUT* = #<IO SYNONYM-STREAM *TERMINAL-IO*>

Line1 to *terminal-io*
Line2 to *terminal-io*

Line1 to *terminal-io*
Line2 to *standard-output*

Line1 to *terminal-io*Line2 to *error-output*



Line1 to *terminal-io*
Line2 to *query-io*

Line1 to *terminal-io*
Line2 to *debug-io*

Line1 to *terminal-io*
Line2 to *trace-output*

Line1 to *standard-output*
Line2 to *terminal-io*

Line1 to *standard-output*
Line2 to *standard-output*

Line1 to *standard-output*Line2 to *error-output*



Line1 to *standard-output*
Line2 to *query-io*

Line1 to *standard-output*
Line2 to *debug-io*

Line1 to *standard-output*
Line2 to *trace-output*

Line1 to *error-output*Line2 to *terminal-io*



Line1 to *error-output*Line2 to *standard-output*



Line1 to *error-output*

Line2 to *error-output*


Line1 to *error-output*Line2 to *query-io*



Line1 to *error-output*Line2 to *debug-io*



Line1 to *error-output*Line2 to *trace-output*



Line1 to *query-io*
Line2 to *terminal-io*

Line1 to *query-io*
Line2 to *standard-output*

Line1 to *query-io*Line2 to *error-output*



Line1 to *query-io*
Line2 to *query-io*

Line1 to *query-io*
Line2 to *debug-io*

Line1 to *query-io*
Line2 to *trace-output*

Line1 to *debug-io*
Line2 to *terminal-io*

Line1 to *debug-io*
Line2 to *standard-output*

Line1 to *debug-io*Line2 to *error-output*



Line1 to *debug-io*
Line2 to *query-io*

Line1 to *debug-io*
Line2 to *debug-io*

Line1 to *debug-io*
Line2 to *trace-output*

Line1 to *trace-output*
Line2 to *terminal-io*

Line1 to *trace-output*
Line2 to *standard-output*

Line1 to *trace-output*Line2 to *error-output*



Line1 to *trace-output*
Line2 to *query-io*

Line1 to *trace-output*
Line2 to *debug-io*

Line1 to *trace-output*
Line2 to *trace-output*



-- 
Sam Steingold (http://www.podval.org/~sds) running w2k
<http://www.camera.org> <http://www.iris.org.il> <http://www.memri.org/>
<http://www.mideasttruth.com/> <http://www.honestreporting.com>
He who laughs last did not get the joke.

Re: fresh-line on mingw

[Bruno H. <br...@cl...>]

Sam wrote:
> if grep 'Line1.*Line2' fresh-line.out > /dev/null; then exit 1; fi
> make: *** [check-fresh-line] Error 1

Good! The test pays off! :-)

> *TERMINAL-IO* = #<IO TERMINAL-STREAM>
> *STANDARD-OUTPUT* = #<IO TERMINAL-STREAM>
> *ERROR-OUTPUT* = #<OUTPUT UNBUFFERED FILE-STREAM CHARACTER>
> 
> Line1 to *terminal-io*Line2 to *error-output*

It seems that *error-output* is the problem. Can you add some debugging
output to same_handle_p (print details about handle1, handle2, filetype1,
filetype2, etc.)?

Bruno

Re: fresh-line on mingw

[Sam S. <sd...@gn...>]

> * Bruno Haible <oe...@py...t> [2005-01-25 20:34:59 +0100]:
>
>> *TERMINAL-IO* = #<IO TERMINAL-STREAM>
>> *STANDARD-OUTPUT* = #<IO TERMINAL-STREAM>
>> *ERROR-OUTPUT* = #<OUTPUT UNBUFFERED FILE-STREAM CHARACTER>
>> 
>> Line1 to *terminal-io*Line2 to *error-output*
>
> It seems that *error-output* is the problem. Can you add some debugging
> output to same_handle_p (print details about handle1, handle2, filetype1,
> filetype2, etc.)?

same_handle_p is not called!  (dunno why...)

-- 
Sam Steingold (http://www.podval.org/~sds) running w2k
<http://www.camera.org> <http://www.iris.org.il> <http://www.memri.org/>
<http://www.mideasttruth.com/> <http://www.honestreporting.com>
Your mouse pad is incompatible with MS Windows - your HD will be reformatted.

Re: fresh-line on mingw

[Bruno H. <br...@cl...>]

Sam wrote:
> >> *TERMINAL-IO* = #<IO TERMINAL-STREAM>
> >> *STANDARD-OUTPUT* = #<IO TERMINAL-STREAM>
> >> *ERROR-OUTPUT* = #<OUTPUT UNBUFFERED FILE-STREAM CHARACTER>

> same_handle_p is not called!  (dunno why...)

When clisp is called in batch mode, i.e. option -x is present,
init_streamvars(true) is called, which calls get_standard_error_file_stream(),
which _must_ call same_handle_p(2,1).

And when clisp is not called in batch mode, init_streamvars(false)
sets *ERROR-OUTPUT* to the same value as *STANDARD-OUTPUT*.

Can you retry debugging this?

Bruno

Re: fresh-line on mingw

[Sam S. <sd...@gn...>]

> * Bruno Haible <oe...@py...t> [2005-01-26 15:39:33 +0100]:
>
>> >> *TERMINAL-IO* = #<IO TERMINAL-STREAM>
>> >> *STANDARD-OUTPUT* = #<IO TERMINAL-STREAM>
>> >> *ERROR-OUTPUT* = #<OUTPUT UNBUFFERED FILE-STREAM CHARACTER>


same_handle_p(0x280,0x248)
GetFileType: 0x3  0x3
neither FILE_TYPE_CHAR nor FILE_TYPE_DISK


-- 
Sam Steingold (http://www.podval.org/~sds) running w2k
<http://www.camera.org> <http://www.iris.org.il> <http://www.memri.org/>
<http://www.mideasttruth.com/> <http://www.honestreporting.com>
Those who can laugh at themselves will never cease to be amused.

Re: fresh-line on mingw

[Bruno H. <br...@cl...>]

Sam wrote:
> same_handle_p(0x280,0x248)
> GetFileType: 0x3  0x3
> neither FILE_TYPE_CHAR nor FILE_TYPE_DISK

This is FILE_TYPE_PIPE, as expected.

Is there some API about pipes that could be used to get details about the
pipes?

Bruno

Re: fresh-line on mingw

[Sam S. <sd...@gn...>]

> * Bruno Haible <oe...@py...t> [2005-01-26 23:00:26 +0100]:
>
> Sam wrote:
>> same_handle_p(0x280,0x248)
>> GetFileType: 0x3  0x3
>> neither FILE_TYPE_CHAR nor FILE_TYPE_DISK
>
> This is FILE_TYPE_PIPE, as expected.
yep.

> Is there some API about pipes that could be used to get details about
> the pipes?
nope:
http://msdn.microsoft.com/library/default.asp?url=/library/en-us/ipc/base/anonymous_pipe_operations.asp

-- 
Sam Steingold (http://www.podval.org/~sds) running w2k
<http://www.camera.org> <http://www.iris.org.il> <http://www.memri.org/>
<http://www.mideasttruth.com/> <http://www.honestreporting.com>
Experience always comes right after it would have been useful.

Re: fresh-line on mingw

[Bruno H. <br...@cl...>]

Sam wrote:
> > Is there some API about pipes that could be used to get details about
> > the pipes?
>
> nope:
> http://msdn.microsoft.com/library/default.asp?url=/library/en-us/ipc/base/anonymous_pipe_operations.asp

Oh well. So we cannot fix it. I've changed the test to something that should
work on mingw.

Bruno

Re: fresh-line on mingw

[Sam S. <sd...@gn...>]

> * Bruno Haible <oe...@py...t> [2005-01-28 18:16:22 +0100]:
>
> Sam wrote:
>> > Is there some API about pipes that could be used to get details about
>> > the pipes?
>>
>> nope:
>> http://msdn.microsoft.com/library/default.asp?url=/library/en-us/ipc/base/anonymous_pipe_operations.asp
>
> Oh well. So we cannot fix it. I've changed the test to something that
> should work on mingw.

lisp.exe -B . -Efile UTF-8 -Eterminal UTF-8 -norc -q -M lispinit.mem -x '(progn
(format *terminal-io* "~&Line1 to *terminal-io*") (format *terminal-io* "~&Line2
 to *terminal-io*") (values))' 2>&1 >> fresh-line.out
*** - WRITE-CHAR on #<CLOSED IO TERMINAL-STREAM> is illegal


-- 
Sam Steingold (http://www.podval.org/~sds) running w2k
<http://www.camera.org> <http://www.iris.org.il> <http://www.memri.org/>
<http://www.mideasttruth.com/> <http://www.honestreporting.com>
There are 10 kinds of people: those who count in binary and those who do not.