From mboxrd@z Thu Jan 1 00:00:00 1970 From: julia.lawall@lip6.fr (Julia Lawall) Date: Wed, 14 Oct 2015 23:48:18 +0200 (CEST) Subject: [Cocci] Compiling latest coccinelle with opensuse In-Reply-To: <20151014214454.GB2920@pema> References: <20151013211451.GD14464@wotan.suse.de> <20151013220512.GE14464@wotan.suse.de> <20151014071143.GA6178@pema> <20151014085350.GB1543@pl-59055.rocqadm.inria.fr> <20151014205913.GH14464@wotan.suse.de> <20151014214454.GB2920@pema> Message-ID: To: cocci@systeme.lip6.fr List-Id: cocci@systeme.lip6.fr On Wed, 14 Oct 2015, S?bastien Hinderer wrote: > Dear Luis, > > Great you could finally compile coccinelle. > And many thanks for having explained all the procedure to get docker > working, it's really great and will be very helpful. > > I won't be able to look into the two problems (menhir and > Common.Union_set) before next week but will do it as soonas possible. > > Regarding your suggestion to distribute the files generated by Menhir: I > don't think this will solve the problem; Indeed, the problem does not > occur when generating these files but when compiling them. > > Actually, there are two things that are used from Menhir. First, Menhir > is used as a commandline tool. It takes a .mly file as inputandproduces > corresponding .ml and .mli files. Thisis roughtly speaking equivalentto > what yacc or bison do: taking a .y file as input and producing a .c as > output. Second, the .ml file that has been generated by Menhir uses a > library whichis called MenhirLib and to which the .cmo or .cmx files > obtained by compilin the .ml file will have to be linked. That is where > the problem you described appears andthat's why I don't think including > the generated .ml files woudlhelpin any way. > > I hope I managed to be clear. :) I have the impression that he managed to have menhirLib without installing menhir when he used the github version. Is that possible? Is there a bundle for menhirLib? julia