From mboxrd@z Thu Jan 1 00:00:00 1970 From: Johannes Schindelin Subject: Re: What's cooking in git.git (Jun 2015, #06; Wed, 24) Date: Thu, 25 Jun 2015 20:14:06 +0200 Organization: gmx Message-ID: <20853ca02f86d3e524fa961eef1f1d2e@www.dscho.org> References: Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Cc: Stefan Beller , git@vger.kernel.org To: Junio C Hamano X-From: git-owner@vger.kernel.org Thu Jun 25 20:14:31 2015 Return-path: Envelope-to: gcvg-git-2@plane.gmane.org Received: from vger.kernel.org ([209.132.180.67]) by plane.gmane.org with esmtp (Exim 4.69) (envelope-from ) id 1Z8BfF-0002U0-UT for gcvg-git-2@plane.gmane.org; Thu, 25 Jun 2015 20:14:18 +0200 Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1751448AbbFYSOO (ORCPT ); Thu, 25 Jun 2015 14:14:14 -0400 Received: from mout.gmx.net ([212.227.17.22]:59709 "EHLO mout.gmx.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750993AbbFYSOM (ORCPT ); Thu, 25 Jun 2015 14:14:12 -0400 Received: from www.dscho.org ([87.106.4.80]) by mail.gmx.com (mrgmx103) with ESMTPSA (Nemesis) id 0MWTSA-1ZbKQH2Dow-00XfUn; Thu, 25 Jun 2015 20:14:07 +0200 In-Reply-To: X-Sender: johannes.schindelin@gmx.de User-Agent: Roundcube Webmail/1.1.0 X-Provags-ID: V03:K0:BDDodpnBFs7duaP7u8iU93oEj8+inm1gNXdqo8nTC0ewp21I1t0 vOf+dhkLV73dMx/LJpCiYapLX4Qf2qGRJzX4vt7zvm7Hq6IvUmsbCo1zh20deAtPnC7NUvj l85EP5/y+lepHSWWiKJtJOnKVp8XyhklXHa/g64ET7vUmjPEHb409BfOY+mFM5ER/uqRdd9 uu3YAwi06WCfjOLBPgMDQ== X-UI-Out-Filterresults: notjunk:1;V01:K0:Be0x74ityA0=:wl03Nm6oUUswLaDevvVyoD 5pREpIzMMpfISpH8sqeaWlBYEjR36/PWa8ghDX24VrkEs4qvrYLhtzp6qFDWRzS8H70GyQOB/ LJj2QO9CW2eUcQOQVn/UPkaDLR6fQP33OY+Oyb6zNsmTOV9OrCFoC4/fLIn8cmHeeJ5D40kNr +Gl+JbJBNThoE6Hu+5qYHZD1rSNDMgxbg8EX6OQcQP/y1z1q3Bjkqx3eUmk6lGVtusRSVn/Gg Av4mOqXQHBIxKoKniqDVhT9VhM5HKsui/YqKkjOOESCp25GIWLtzOOR3OSItiTfNLl33W1pNt tj06HVZJuapd/X84WU1PsmhN9UtSQ4myZaM2OS1zAkvmanN1zU0KLvcfcPFdBf83uBEX6pIz7 dk5vtDQBKdzxrwy/VkOgy/Wf3NgErJ53TxMQdC0V3zb7j8FJSb3umlwtcduXdW9Zk3Gjwe9gf qqsenBML2aXnUSqH7XGIVe2hJPCANro2A8ojeDlYUOo23i7RQQZ97uS2wodkSvtk0YRJQHstB q6zOY23u15JB7/BHFNREY3Hiilr+x+PjakXH3SXsSC9tTCo3D7w38nLqwF6x4YqybILCGgqal D0IFv7ni+boEoJcpY+8JmCh7FPzSzjBWxdXK2CsgFcX4oaQlpG5e5/+yXtNqIG/WBxXz5npbP W2f3p1ZvTQd9XfptpS26HwMeyHnLXc7AKPV8NBc1tWk2Ufta5jilUpjtkJZjeZKjtyGQ= Sender: git-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: git@vger.kernel.org Archived-At: Hi Junio, On 2015-06-25 15:14, Junio C Hamano wrote: > Johannes Schindelin writes: > >> - I *think* that the current `recreate_opt()` is fine, but if it >> is used more, we will have to quote the arguments to guard against >> spaces and quotes and whatnot > > I admit that that was the reaction I had when I saw it for the first > time, until I realized how it is used. And this one I have to > disagree. recreate-opt is used to munge a single argument, suitable > to be placed into argv[] for execv() like thing---there is no need > for shell quoting there. It is a job for the caller that gets its > result to shell quote if it wants to pass the result to a shell. Ah yes, no single command-line is constructed from those reconstructed options. So yes, you're right, all is good! Ciao, Dscho