Re: [nft PATCH v2] parser_bison: Accept arbitrary user-defined names by quoting

[Phil Sutter <phil@nwl.cc>]

Hi Pablo,

On Mon, Jul 01, 2019 at 08:19:24PM +0200, Pablo Neira Ayuso wrote:
> On Mon, Jul 01, 2019 at 08:13:41PM +0200, Pablo Neira Ayuso wrote:
> > On Mon, Jul 01, 2019 at 06:11:39PM +0200, Phil Sutter wrote:
> > > Hi Pablo,
> > > 
> > > On Fri, Jun 28, 2019 at 08:00:51PM +0200, Pablo Neira Ayuso wrote:
> > > > On Mon, Jun 24, 2019 at 06:36:08PM +0200, Phil Sutter wrote:
> > > > > Parser already allows to quote user-defined strings in some places to
> > > > > avoid clashing with defined keywords, but not everywhere. Extend this
> > > > > support further and add a test case for it.
> > > > > 
> > > > > Signed-off-by: Phil Sutter <phil@nwl.cc>
> > > > > ---
> > > > > Changes since v1:
> > > > > - Fix testcase, I forgot to commit adjustments done to it.
> > > > > 
> > > > > Note: This is a reduced variant of "src: Quote user-defined names" sent
> > > > >       back in January. Discussion was not conclusive regarding whether
> > > > >       to quote these names on output or not, but I assume allowing for
> > > > >       users to specify them by adding quotes is a step forward without
> > > > >       drawbacks.
> > > > 
> > > > So this will fail later on, right?
> > > > 
> > > >         nft list ruleset > file.nft
> > > >         nft -f file.nft
> > > 
> > > Yes, that's right. I sent a complete version which does the necessary
> > > quoting on output in January[1], but discussion wasn't conclusive. You
> > > had a different approach which accepts the quotes as part of the name
> > > but you weren't happy with it, either. I *think* you wanted to search
> > > for ways to solve this from within bison but we never got back to it
> > > anymore.
> > > 
> > > This simplified patch is merely trying to make things consistent
> > > regarding user-defined names. IIRC, I can already have an interface
> > > named "month", use that in a netdev family chain declaration (quoted)
> > > and 'nft list ruleset' will print it unquoted, so it can't be applied
> > > anymore. Without my patch, it is simply impossible to use certain
> > > recognized keywords as names for tables, chains, etc., even if one
> > > accepted the implications it has.
> > 
> > I'm not arguing there's something to fix.
> > 
> > I'm telling this is still incomplete.
> > 
> > Would you allocate a bit of time to discuss this during the NFWS?
> 
> I mean, a quick summary of the different options for a complete
> solution for this, and we decide there.
> 
> Unless you tell me this is very urgent :-)

That's a great idea! I'll check my inbox for more of those unfinished
topics. :)

Thanks, Phil