> From: Gavin Smith <[email protected]> > Date: Thu, 22 Oct 2026 10:19:26 +0100 > Cc: [email protected] > > > Even though <entry text> is followed by a colon, it may itself > > contain colons, so Info readers should try to include as much of the > > line as possible in the <entry text>. (However, <entry text> may not > > contain the ‘DEL’ characters that may occur in <node-spec>.) > > > > This seems to say that "<entry text>" does NOT follow the conventions > > of menus, where menu entry text is quoted if it includes colons (and > > other special characters). But then an Index is not "just a menu", > > and so the "* Menu:" heading at the beginning of Index nodes is at > > least misleading. > > Yes, index entries should be treated differently. > > I reported an Emacs "bug" about this: > https://lists.gnu.org/archive/html/bug-gnu-emacs/2019-01/msg00235.html
Is it possible to remove this special formatting of menus in Index nodes? It makes the job of detecting and parsing menu entries significantly more complex than it already is, at least in Emacs. And what is the reason for this special formatting? why cannot the menus in Index nodes use the same conventions as in other menus? > > Am I missing something, or is the above Index entry incorrect due to > > some bug in texi2any? If such entries are correct, then it is > > impossible for an Info reader to treat Index nodes as menus, and some > > other strategy (but which one?) is needed for Index nodes. And if > > this is indeed the case, what is the reason for the different > > formatting of the menus in Index nodes? > > The recommendation was that Info resaders should use the last colon in > the line, not the first. This was to support index entries with colons > in them, assuming the node name doesn't contain a colon. Yes, but now node names _can_ include colons, as the example I posted clearly shows. So looking for a colon is no longer reliable, and quoting solves this problem cleanly.
