On Tue, Jun 03, 2014 at 10:35:01AM +0200, Frederik Kriewitz wrote: > Hello Ondrej, > > On Thu, May 29, 2014 at 11:13 PM, Ondrej Zajicek <santi...@crfreenet.org> > wrote: > > Thanks, merged. Strictly speaking they should not accept symbols as > > value, but string constants denoted by symbols. > > Thank you. > > I missed one option, in interface patterns string constants denoted by > symbols aren't allowed yet. > But it's a little more complicated. > > The relevant grammar: > iface_patt_node_body: > TEXT { this_ipn->pattern = $1; this_ipn->prefix = IPA_NONE; > this_ipn->pxlen = 0; } > | prefix_or_ipa { this_ipn->pattern = NULL; this_ipn->prefix = > $1.addr; this_ipn->pxlen = $1.len; } > | TEXT prefix_or_ipa { this_ipn->pattern = $1; this_ipn->prefix = > $2.addr; this_ipn->pxlen = $2.len; } > > > The last alternative "TEXT prefix_or_ipa" can be eaily converted to > "text prefix_or_ipa". But simply changing the first alternative (just > "TEXT") to "text" causes a conflict as both "text" and "prefix_or_ipa" > might be a symbol.
Yes, i also noticed that. But it seems that whole prefix_or_ipa is a bit problematic (you could use symbol for ipa but not for prefix, although we support prefix constants) and it would need some further changes to fix it. I will look at that. -- Elen sila lumenn' omentielvo Ondrej 'Santiago' Zajicek (email: santi...@crfreenet.org) OpenPGP encrypted e-mails preferred (KeyID 0x11DEADC3, wwwkeys.pgp.net) "To err is human -- to blame it on a computer is even more so."
signature.asc
Description: Digital signature