GNU bug report logs - #30063
26.0.90; Silent fail with `rst-compile-pdf-preview'

Previous Next

Full log

Message #17 received at 30063 <at> debbugs.gnu.org (full text, mbox):

From: Eli Zaretskii <eliz <at> gnu.org>
To: Noam Postavsky <npostavs <at> users.sourceforge.net>
Cc: stefan <at> merten-home.de, 30063 <at> debbugs.gnu.org, simenheg <at> gmail.com
Subject: Re: bug#30063: 26.0.90; Silent fail with `rst-compile-pdf-preview'
Date: Tue, 30 Jan 2018 15:49:47 +0200

> From: Noam Postavsky <npostavs <at> users.sourceforge.net>
> Date: Sun, 28 Jan 2018 13:11:34 -0500
> Cc: stefan <at> merten-home.de, 30063 <at> debbugs.gnu.org
> 
> Simen Heggestøyl <simenheg <at> gmail.com> writes:
> 
> >  	 (command (format "%s %s %s && %s %s ; rm %s"
> > +			  pdf-compile-program
> >  			  buffer-file-name tmp-filename
> >  			  rst-pdf-program tmp-filename tmp-filename)))
> > +    (unless (executable-find pdf-compile-program)
> > +      (error "Cannot find executable `%s'" pdf-compile-program))
> > +    (unless (executable-find rst-pdf-program)
> > +      (error "Cannot find executable `%s'" rst-pdf-program))
> 
> It's possible to have PATH and exec-path desynchronized, such that the
> above code could throw an error even though the
> start-process-shell-command call later would succeed.  Maybe we should
> should just consider that a misconfiguration on the user's part though.

Yes, we could do that.  But can we signal an error only when the
command fails?  That should minimize false negatives.

Also please note that the shell command as written is unportable: the
";" part will not work on MS-Windows, we need to use "&" instead.  But
that's a separate issue.

Previous Next

GNU bug tracking system
Copyright (C) 1999 Darren O. Benham, 1997,2003 nCipher Corporation Ltd, 1994-97 Ian Jackson.