|
From: | Paul Eggert |
Subject: | Re: segmentation fault writing cyclic term |
Date: | Sat, 4 Dec 2021 08:55:53 -0800 |
User-agent: | Mozilla/5.0 (X11; Linux x86_64; rv:91.0) Gecko/20100101 Thunderbird/91.3.1 |
On 12/4/21 03:21, Paulo Moura wrote:
| ?- X = X+1, write(X). X = X+1, write(X). Segmentation fault (core dumped) Shouldn't write/1 be protected against cyclic terms?Note that support for cyclic terms is not a ISO Prolog standard requirement.
Quite right, this is not a standards violation. Still, gprolog already defends against cyclic terms when it outputs them:
| ?- X=X+1. cannot display cyclic term for Xso presumably it could defend against them in 'write' if this was considered useful (which it would be, for my students...).
[Prev in Thread] | Current Thread | [Next in Thread] |