Path: csiph.com!optima2.xanadu-bbs.net!xanadu-bbs.net!news.glorb.com!usenet.stanford.edu!not-for-mail From: Chet Ramey Newsgroups: gnu.bash.bug Subject: Re: command substitution is stripping set -e from options Date: Thu, 8 Oct 2015 11:26:08 -0400 Organization: ITS, Case Western Reserve University Lines: 44 Approved: bug-bash@gnu.org Message-ID: References: <560D83DA.9020405@redhat.com> <20151002122925.GK25574@eeg.ccf.org> <20151002132221.GL25574@eeg.ccf.org> <56103208.30406@case.edu> Reply-To: chet.ramey@case.edu NNTP-Posting-Host: lists.gnu.org Mime-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: 7bit X-Trace: usenet.stanford.edu 1444318106 19580 208.118.235.17 (8 Oct 2015 15:28:26 GMT) X-Complaints-To: action@cs.stanford.edu Cc: Greg Wooledge , bug-bash@gnu.org, chet.ramey@case.edu To: Christoph Gysin Envelope-to: bug-bash@gnu.org User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10.11; rv:38.0) Gecko/20100101 Thunderbird/38.3.0 In-Reply-To: X-Junkmail-Whitelist: YES (by domain whitelist at mpv2.tis.cwru.edu) X-detected-operating-system: by eggs.gnu.org: GNU/Linux 2.4.x-2.6.x [generic] X-Received-From: 129.22.105.37 X-BeenThere: bug-bash@gnu.org X-Mailman-Version: 2.1.14 Precedence: list List-Id: Bug reports for the GNU Bourne Again SHell List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Xref: csiph.com gnu.bash.bug:11589 On 10/5/15 5:37 PM, Christoph Gysin wrote: >> The parent shell (the one that enabled -e) should be the one to make the >> decision about whether or not the shell exits. The exit status of the >> command substitution doesn't make a difference except in one special case, >> so inheriting errexit and exiting (possibly prematurely) doesn't really >> help the parent decide whether or not to exit. > > I'm not sure I fully understand. > > The parent shell should be the one to decide if the script is supposed > to abort on any unsuccessful exit status. Command substitution should > not change that. The parent shell decided via set -e that it wants to > exit immediately on error. Sure, the parent shell should make this decision. Consider that, other than one special case, the exit status of a command substitution has no effect on whether a command succeeds or fails, and whether or not the parent shell exits. If you have something like set -e command -f $(command that generates a filename) -opts other arguments and the command substitution inherits the -e option and exits prematurely, the parent shell will not inspect its exit status and will happily execute the command, possibly with faulty data. This is simply how command substitution works. > If you don't want to fix this for backwards compatibility, is there > anyway we could change that behaviour explicitly? I.e. with another > option? Avoiding command substitution isn't really an option, and this > essentially disables the whole point of set -e. I think you're overlooking what I referred to above: that the exit status of a command substitution doesn't have any effect on whether the parent's command succeeds or fails except in one case: the right-hand-side of an assignment statement that is the last assignment in a command consisting only of assignment statements. To say that it `disables the whole point of set -e' is a considerable overstatement. -- ``The lyf so short, the craft so long to lerne.'' - Chaucer ``Ars longa, vita brevis'' - Hippocrates Chet Ramey, ITS, CWRU chet@case.edu http://cnswww.cns.cwru.edu/~chet/