> I see the sense but this is an incompatible change.  Csound originally
> took two files -- the first an orchestra and the second a score.  Names
> did/does not matter.  One file is thus an error.  This check allowed one
> file as long as it was a .csd file.
> 
> As long as we are happy with this minor lack of backwards....

One file is an error but we have a segfault (all parsers) with

  csound orcfile

caused by `strlen(csound->scorename)' in main.c. The change also fixes it.
There is not a segfault with a csdfile without extension because the
parser signals an error

  error: syntax error, unexpected S_LT  (token "<") line 0:

* Some free thoughts

Discarded solution: a xml file begins with a tag, so the '<' char is
the first char and it is an error in an orc file. Therefore it is
enough a check for the first non-blank character; if it is '<', the
file is 'csd'. But it is too late because there are some csd files
that begin with

  some text bla bla
  bla bla ...
  <CsoundSynthesizer>

so I think that a `--orc' flag is useful when there is only one file

  csound text         (it is a csd file or error)
  csound --orc text   (it is an orc file with INF score or error)

I use tmpfs for temporary csd files, but I like

  csound <(myscript)

or (without `-i stdin')

  myscript | csound /dev/stdin

because I avoid to set/write/remove a tmpfile and to use a "trap" for
SIGINT in my scripts.

Now we can use it only with (it is a workaround but `file.sco' is immutable)

  csound <(make_orc_file_script) file.sco
  make_orc_file_script | csound /dev/stdin file.sco

only because `/dev/fd/N' and `/dev/stdin' are without the csd extension.

Let me know what is good for you.

tito

------------------------------------------------------------------------------
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and 
threat landscape has changed and how IT managers can respond. Discussions 
will include endpoint security, mobile security and the latest in malware 
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
_______________________________________________
Csound-devel mailing list
Csound-devel@lists.sourceforge.net