Exit status of Sage

[Jernej Azarija]

Hello,

I have to use Sage from an external comand and in order to do so I'll need to rely on the exit status given by Sage. Considering a trivial example

=============
$ cat foo.sage
exit(0)
=============


I get the following behaviour


=============
$ sage la.sage
0
$ echo $?
1
=============

There are two things I am confused with here.

1. Why do we print 0?

2. Why is the exit status 1 - indicating an error by UNIX standards?


Is there any reason behind this? If yes , what would be the best way to force my own exit status so that I can interpret the execution of Sage from an external program?

Best,

Jernej

[Jernej Azarija]

I suppose there is a bug in the sage bash script?

===============
101
102     # Return $out
103     echo "$out"
104 }
===============

should that be exit actually?

[John H Palmieri]

The underlying problem is that the "exit" function in Python doesn't accept any arguments, so "exit(0)" raises an error when you run it in Sage. This is why the exit status is nonzero. If your script had the line "exit()", it would run as expected. Given that, I'm not sure why it prints 0. If you do "exit(3)", it will print 3 instead.

--
John

[Robert Bradshaw]

You probably want sys.exit, which does take an exit status as an integer.

[Jernej Azarija]

The thing is that sys.exit works pretty much the same way

====
azi@goodegg:~$ cat foo.sage
import sys
sys.exit(42)
azi@goodegg:~$ sage foo.sage
42
azi@goodegg:~$ echo $?
1

====

And the print itself is extremely annoying. I need to call Sage 10^6 times from an external program and get 10^6 lines of non-needed output.

--
You received this message because you are subscribed to a topic in the Google Groups "sage-devel" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/sage-devel/CriQpiwWRLA/unsubscribe.
To unsubscribe from this group and all its topics, send an email to sage-devel+...@googlegroups.com.
To post to this group, send email to sage-...@googlegroups.com.
Visit this group at http://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

[Dima Pasechnik]

On 2015-01-18, Jernej Azarija <azi.s...@gmail.com> wrote:
> The thing is that sys.exit works pretty much the same way
>
>====
> azi@goodegg:~$ cat foo.sage
> import sys
> sys.exit(42)
> azi@goodegg:~$ sage foo.sage
> 42
> azi@goodegg:~$ echo $?
> 1
>====
>
> And the print itself is extremely annoying. I need to call Sage 10^6 times
> from an external program and get 10^6 lines of non-needed output.

this looks pretty bad, given that Sage's startup+exit time is 1 or 2 seconds
on a reasoably fast computer.
(unless your computations are still much slower, of course).

There should be ways to communicate with Sage without quitting in.

[Jernej Azarija]

On Sun, Jan 18, 2015 at 5:02 PM, Dima Pasechnik <dim...@gmail.com> wrote:

On 2015-01-18, Jernej Azarija <azi.s...@gmail.com> wrote:
> The thing is that sys.exit works pretty much the same way
>
>====
> azi@goodegg:~$ cat foo.sage
> import sys
> sys.exit(42)
> azi@goodegg:~$ sage foo.sage
> 42
> azi@goodegg:~$ echo $?
> 1
>====
>
> And the print itself is extremely annoying. I need to call Sage 10^6 times
> from an external program and get 10^6 lines of non-needed output.

this looks pretty bad, given that Sage's startup+exit time is 1 or 2 seconds
on a reasoably fast computer.
(unless your computations are still much slower, of course).

There should be ways to communicate with Sage without quitting in.

I agree. I ended up writing the output to a file and parsing it within the caller. Still pretty  ugly and inefficient but it works.

Though this still does not answer the question - why is Sage printing the exit status and not conforming to the standard behaviour of unix programs?

[John H Palmieri]

On Sunday, January 18, 2015 at 8:09:00 AM UTC-8, Jernej Azarija wrote:

On Sun, Jan 18, 2015 at 5:02 PM, Dima Pasechnik <dim...@gmail.com> wrote:

On 2015-01-18, Jernej Azarija <azi.s...@gmail.com> wrote:
> The thing is that sys.exit works pretty much the same way
>
>====
> azi@goodegg:~$ cat foo.sage
> import sys
> sys.exit(42)
> azi@goodegg:~$ sage foo.sage
> 42
> azi@goodegg:~$ echo $?
> 1
>====
>
> And the print itself is extremely annoying. I need to call Sage 10^6 times
> from an external program and get 10^6 lines of non-needed output.

this looks pretty bad, given that Sage's startup+exit time is 1 or 2 seconds
on a reasoably fast computer.
(unless your computations are still much slower, of course).

There should be ways to communicate with Sage without quitting in.

I agree. I ended up writing the output to a file and parsing it within the caller. Still pretty  ugly and inefficient but it works.

Though this still does not answer the question - why is Sage printing the exit status and not conforming to the standard behaviour of unix programs?

Oh, I see: it's the preparsing. When you do "sys.exit(42)", the number 42 is a Sage integer, not a Python integer. So sys.exit doesn't know what to do with it. (I still don't know why that causes the number to be printed.)

So use sys.exit(int(42)) instead.

[Vincent Delecroix]

2015-01-18 17:32 UTC+01:00, John H Palmieri <jhpalm...@gmail.com>:

> Oh, I see: it's the preparsing. When you do "sys.exit(42)", the number 42
> is a Sage integer, not a Python integer. So sys.exit doesn't know what to
> do with it. (I still don't know why that causes the number to be printed.)
>
> So use sys.exit(int(42)) instead.

Nice catch! Alternatively:

* use a .py extension (that way there will not be any preparsing)
* write sys.exit(int(42r))

Vincent

[Samuel Lelievre]

Nice catch! Alternatively:

 * use a .py extension (that way there will not be any preparsing)
 * write sys.exit(int(42r))

You mean

    sys.exit(42r)

(42r is the raw Python integer 42, so you don't need to apply int to it).

: )